Build if you wish--your call.
Principle of operation will become apparent as we go along--some will work it out sooner.
Part A--the drive coil configuration.
1x PM of rod type. 2 inches long 10-12mm diameter.
NEO or ceramic type ok.
1x PVC or plastic tube to fit over PM snug,and of same length.
1x Steel tube-black pipe or iron pipe,2-3mm thick,with non magnetic keepers at each end.
To be same length as magnet and PVC/plastic tube,and to fit snug over PVC/plastic tube.
A coil of 200 turns is to be wound over the steel pipe,with wire size of between .6mm to .7mm
That is the drive coil.
P.S
I would suggest using the weaker ceramic magnets,unless you are confident in being able to make a very sturdy rig,as the two attracting magnetic poles will be passing within 1mm of each other,and using neo's,this will create a very strong pull on the rotor shaft at the TDC position of each rotor+stator magnet alignment.
A heavy flywheel/rotor is recommended .
Part 2-the rotor electric/magnetic torque amplifiers.
4 to 6 of these will need to be built--6 is better for smoother operation.
4-6x 2 inch x 10-12mm rod magnets.
Each magnet has a plastic/plexiglass keeper at one end(so as in attraction mode to that of the drive coil),and a second keeper at the center of the magnets length.
A coil of 400 turns of wire size .4mm - .5mm is to be wound around the magnet,between the keepers.
Heat srink or electrical tape may be wound around the magnet if you are concerned about shorting when using neo's--ceramic magnets of course are non conductive.
That is it for the rotor torque amplifiers.
You will notice that in my diagrams,the stator coil has the south pole facing !what will be! the rotor magnets north pole.
Part3-the rotor assembly
The rotor should be between 6 and 10 inches in diameter,and 20-25mm thick.
To be made of a non magnetic material--E.G-plexiglass.
The larger the diameter,the heavier the rotor,the less vibration.
I recommend one of size 10 inches-250mm in diameter.
You are now to drill 4 or 6 holes in symmetry ,of diameter of your rotor magnets-->10 or 12mm,around the circumference and middle of thickness of your rotor.
You then glue your electromagnetic torque amplifiers (half without coil) into your rotor with a strong glue.
Please see diagram below,and think about what you are looking at.
Next post is of rotor electrical component setup.
Nice step by step instructions. Should be interesting.
Starting to round up the needed pm's etc. Will try to use my Prusa 3D printer for the rotor and stator brackets, etc.
Pm
Hi Brad.
Could you provide a link to the magnets please. 12mm diameter by 50mm long don't seem to be available via eBay UK.
Cheers Graham.
Quote from: Grumage on 2018.12.18, 19:01:21
Hi Brad.
Could you provide a link to the magnets please. 12mm diameter by 50mm long don't seem to be available via eBay UK.
Cheers Graham.
In Canada and the US, a 1/2 x2 inch magnet is only $4,18 US at...
http://www.magnet4less.com/product_info.php?cPath=1_11&products_id=1045 (http://www.magnet4less.com/product_info.php?cPath=1_11&products_id=1045)
Ron
Hi Graham,
If Brad agrees with 8mm dia x 30mm magnet sizes, then this may be a good overall offer (considering shipping):
https://www.first4magnets.com/cylindrical-rod-c39/8mm-dia-x-30mm-thick-n42-neodymium-magnet-2-7kg-pull-p3013#ps_1-1022 (https://www.first4magnets.com/cylindrical-rod-c39/8mm-dia-x-30mm-thick-n42-neodymium-magnet-2-7kg-pull-p3013#ps_1-1022)
In this slightly smaller size respect, perhaps other seller options may be found in the UK.
Gyula
Quote from: partzman on 2018.12.18, 18:45:20
Starting to round up the needed pm's etc. Will try to use my Prusa 3D printer for the rotor and stator brackets, etc.
Pm
Hi Graham
12x50s are avaliable on ebay au,so not sure why not in uk?
Anyway,i get most of my magnets from Frenergy Magnets.
There is also nothing stopping you using shorter magnets glued together-say like 2 of 12mm x 25mm to make your 12x50 magnet,or even a heap of 12mm button magnets glued together,as the resulting field will be that of one long magnet.
The size is really up to you.
I just scaled down another device to a 1/3rd scale.
As long as you can build to the design-fit coils on 1/2 of the rotor magnets,then it should work.
Dig up what you have lying around,and lets look to see if they will work.
I am using 12mm × 10mm ceramic magnets glued together.
For those using neo's,consider this__
2 of 12x50 n52s will have an attraction force of around 25kg's
This is the sideways force that will be placed on your rotor and shaft--hence my warning in the previous post about being able to build a very solid device,where there will be less than 1mm flex between the stator and rotor magnets.
Do not use alnico magnets,as the poles are easily flipped.
Cheap ceramic magnets will make for a good pulse motor O0
Brad
Neo's of this size are quite expensive here in Oz Grum
https://www.ebay.com.au/itm/2x-Strong-Rare-Earth-Cylinder-Magnets-12mm-x-50mm-N48-Super-Rod-Neodymium-Arts/230909636982?epid=630574474&hash=item35c3497176:g:CdwAAOSwX61ZCqZc:rk:1:pf:0
I am using these glued together.'
https://magnet.com.au/ferrite-disc-magnet-12mm-x-6mm.html
Below is the circuit for each of the torque amplifiers.
The output from each circuit(voltage doubler) will run to two slip rings on the rotor--drawing yet to come.
Brad
Hi Brad.
It would seem those specific magnets are only available in Oz!! :)
Cheers Graham.
On the magnets...
In the USA, Dollar Tree do the 'dancing flowers', likely everywhere else has such things as novelty solar rockers. Nearly all have a ceramic in the base.
I have decided to follow Brad's advise to start with ceramic magnets. I found that Hobby Lobby has some .5" dia. by .2" length magnets. So stacking 10 of them will give you a 1/2 inch by 2 inch magnet. They come in a package of 15 for $2.98 USD. I got 5 packages for enough to do seven magnets with 5 left over.
Brad, I have a way to measure how many feet of wire I am putting in a coil but I don't have a good way to measure the number of turns. Do you know how many feet are required or do you have a good way to measure the number of turns?
Thanks for starting this project.
Carroll
Quote from: CITFTA on 2018.12.19, 19:44:41
I have decided to follow Brad's advise to start with ceramic magnets. I found that Hobby Lobby has some .5" dia. by .2" length magnets. So stacking 10 of them will give you a 1/2 inch by 2 inch magnet. They come in a package of 15 for $2.98 USD. I got 5 packages for enough to do seven magnets with 5 left over.
Brad, I have a way to measure how many feet of wire I am putting in a coil but I don't have a good way to measure the number of turns. Do you know how many feet are required or do you have a good way to measure the number of turns?
Thanks for starting this project.
Carroll
Hi Carroll
I simply count the number of turns as i wind.
Calculating feet would be a task,as there are a few variables,like length of core,starting diameter to finnishing diameter,how neat you wind the wire on.
Just easier to count turns.
Brad
A quote from Smudge on OU.com
It is possible to get more energy out than that put in if the non-linearity changes between current rise and current fall. For instance if during current rise we have something like the second image, then during current fall it changes to a linear fall as in the first image, then we put in the small green area and get out the large green area. This requires some additional feature to change the flux v, mmf line. Stearn demonstrated something in Dublin that did just that, they used toroidal coils wound onto ring cores on their stator and had magnets brought close to the cores by the rotor. While the magnet was away from the core the flux and current build up was linear, but when the magnet got close it cross-saturated the core whence the coil flux dropped to near zero inducing voltage into a load resistor. That produces a clockwise hysteresis loop yielding an electrical energy gain. For this to be OU it must be shown that the differing forces on the magnet for approach and recede do not account for that energy gain.
Here is the explanation for the energy gain when a toroidal inductor is charged with current, the flux is held constant by having the coil shorted, a magnet is brought close to cross saturate the core and the current in the coil increases, then the inductor is discharged into a load. I think this is what Steorn tried to demonstrate in Dublin some time ago. But there will be some assymmetry of the magnet pull-in and pull-out forces so it may not be OU.
Smudge
Quote from: Smudge on 2018.12.20, 16:28:11
Here is the explanation for the energy gain when a toroidal inductor is charged with current, the flux is held constant by having the coil shorted, a magnet is brought close to cross saturate the core and the current in the coil increases, then the inductor is discharged into a load. I think this is what Steorn tried to demonstrate in Dublin some time ago. But there will be some assymmetry of the magnet pull-in and pull-out forces so it may not be OU.
Smudge
Same basic principle I was trying to accomplish with my IET or inductance energy transfer circuitry which was inspired by your papers Smudge. A coil/core was charged to a point just below saturation, and then additional energy was dumped into same coil/core from another source in a short period of time driving the core into saturation. The coil was then discharged into a load but the results I seemed to achieve were all conservative but very high efficiency.
I could see that using an approaching source of flux to a similar coil/core situation might possibly yield OU results.
Pm
In the Steorn system the magnets were on the outer edge of the rotor with the magnet axis pointed outward. The stator ring cores had their axes (passing through the hole) pointing inwards Thus the magnet flux at nearest approach was everywhere at right angles to the core flux.
Smudge
Quote from: Grumage on 2018.12.19, 18:56:22
Hi Brad.
It would seem those specific magnets are only available in Oz!! :)
Cheers Graham.
Only if neo's were not so expensive C.C
Thanks for sharing Brad. ;)
A glue that i have been using for years is Goop. I prefer Marine Goop of the few different ones they have. E6000 is similar but a bit runny. Let set up for a day but after, one will see it is a very good product.
Mags
Quote from: Magluvin on 2018.12.22, 00:38:51
Thanks for sharing Brad. ;)
A glue that i have been using for years is Goop. I prefer Marine Goop of the few different ones they have. E6000 is similar but a bit runny. Let set up for a day but after, one will see it is a very good product.
Mags
Thanks Mags for the info,it is good to have you here O0
Yes,the glue used must be very strong,as there is a lot of weight spinning around on the rotor,and the centrifugal forces will be high.
Perhaps some sort of non magnetic ring should be put around the outside circumference of the torque multipliers.
Brad
Well i only got a small portion of the test rig build today,as the temp hit the mid 40s in the workshop by 10am.
Anyway, i will keep at it as time permits,and also the temp C.C
Also being xmas,most of my free time will be spent with family of course.
Brad
+1 for the E6000.
I use it to fix soles on shoes.
The rest falls apart before the sole comes off again.
Nice ;) Dont need much. Use small amounts to attach and dry quickly, then apply an even amount around as to not unbalance the wheel.
Mags
Quote from: TinMan on 2018.12.18, 14:05:38
Please see diagram below,and think about what you are looking at.
Using changes in the transition zone between the rotor and stator magnets and resulting changes to the neutral zone within the rotor magnet to induce a current in the rotor coil, the magnitude of which is independent from the current supplied to the stator coil. ???
Cheers
Quote from: Hoptoad on 2018.12.26, 04:30:09
Using changes in the transition zone between the rotor and stator magnets and resulting changes to the neutral zone within the rotor magnet to induce a current in the rotor coil, the magnitude of which is independent from the current supplied to the stator coil. ???
Cheers
Excellent, i will buy that
Long time no see, my friend.
Ron
Quote from: Hoptoad on 2018.12.26, 04:30:09
Using changes in the transition zone between the rotor and stator magnets and resulting changes to the neutral zone within the rotor magnet to induce a current in the rotor coil, the magnitude of which is independent from the current supplied to the stator coil. ???
Cheers
Indeed
Quote from: ronee on 2018.12.26, 16:31:23
Excellent, i will buy that
Long time no see, my friend.
Ron
LOL. As usual Ron, I'm not selling! :P
Hope you've had a merry christmas festive season, all the best for the coming new year.
Cheers and KneeDeep
Well now that xmas is out of the way,i can get back into the workshop.
Should have the testbed done by sat night.
Brad
Thanks for sharing Brad. Did you just have the iron pipe lying around? I don't suppose you got it from Reece or Bunnings :)
Never mind printing with Iron PLA. Mine will be scaled down so I can print parts.
Iron PLA?
What kind of temperature do you have to set the extruder to for that stuff? I mean, 200C wouldn't melt it, would it? Have just never heard of it, have heard of conductive filaments...unless...that's what a conductive filament is made with !
:o
*sound in the background of a penny dropping*
Quote from: Slider2732 on 2018.12.30, 02:27:20
Iron PLA?
What kind of temperature do you have to set the extruder to for that stuff? I mean, 200C wouldn't melt it, would it? Have just never heard of it, have heard of conductive filaments...unless...that's what a conductive filament is made with !
:o
*sound in the background of a penny dropping*
conductive is usually graphene so resistance sucks. https://www.proto-pasta.com/pages/magnetic-iron-pla drive coil done in time for NYE :)
Ok,so i was thinking about this some more,and i believe there would be a benefit in putting a slot along the length of the steel tube,so as it is not seen as a shorted turn.
At these low RPM,it may not have much effect,but every bit helps.
So if you have not built your drive coil yet,it may be something to consider ?
Brad
A couple of pictures taken during the build.
As i have chosen to share this with you all,rather than going down the private !hope to find investors! route,i would hope that you follow and stick to the build instructions i have given you.
This includes making sure that the quality of the build is high,so as it has the best chance of giving the results we need.
This includes making sure your windings are as neat as possible.
I dont expect to see machine wound quality,but you can get the coil windings quite neat when wound by hand.
Remember,this is not a race,so give the build the time it deserves.
Brad
Hey Brad
Is the iron tube used to not demagnetize the magnets with the pulse of the drive coil?
Mags
Quote from: TinMan on 2018.12.31, 02:55:58
This includes making sure your windings are as neat as possible.
I dont expect to see machine wound quality,but you can get the coil windings quite neat when wound by hand.
Brad
FAIL on my 1st drive coil ...... :) Do you straighten your F&P rotor wire? For my rotor the print will take around 13h then I've designed hollow threaded bobbins to house the mags and screw into the rotor. The six will take a further 10 hours to print I think. Still to design a robust housing for it all.
Quote from: Magluvin on 2018.12.31, 03:50:57
Hey Brad
Is the iron tube used to not demagnetize the magnets with the pulse of the drive coil?
Mags
Hi Mags
The iron tube is to concentrate the magnetic field produced by the drive coil.
Brad
Quote from: JimBoot on 2018.12.31, 03:59:41
FAIL on my 1st drive coil ...... :) Do you straighten your F&P rotor wire? For my rotor the print will take around 13h then I've designed hollow threaded bobbins to house the mags and screw into the rotor. The six will take a further 10 hours to print I think. Still to design a robust housing for it all.
Oh wish i had a 3D printer.
Would make life so easy :)
Great idea there Jim O0
Brad
Quote from: TinMan on 2018.12.31, 06:52:37
Oh wish i had a 3D printer.
Would make life so easy :)
Great idea there Jim O0
Brad
Maaaaate you would love it. Mine is too small now. Got it from Aldi for $300. Good one to start with so I know what I'm missing. Spool prototype failed - thread needs to be bigger. These Neos aren't going anywhere. The outward end is concave internally, with the outer facing hole 6mm. I'm hoping that will give me 1mm or less clearance for the neo.
i'm glad I have a printer. I don't have your skills Brad.Rotor is printing. I'll be printing the poles in a flexible pla as it's tight fit for the mag. I'm thinking a horiz mount for this unless we're relying on gravity :)
Quote from: JimBoot on 2018.12.31, 13:35:59
i'm glad I have a printer. I don't have your skills Brad.Rotor is printing. I'll be printing the poles in a flexible pla as it's tight fit for the mag. I'm thinking a horiz mount for this unless we're relying on gravity :)
Ok
I really have o get me a 3D printer.
Then i'll be needing bulk lessons Jim lol.
Brad
So which one Jim ?
https://www.ebay.com.au/itm/Anet-A8-DIY-3D-Printer-Upgrade-High-Precision-Reprap-Prusa-i3-PLA-Filament-8G/372009728690?hash=item569d81fab2:rk:5:pf:0
https://www.ebay.com.au/itm/Creality-3D-Ender-3-3D-Printer-OSHW-Certified-220X220X250mm-Removable-Plate-AU/223285106539?epid=12027081611&hash=item33fcd4576b:g:YgsAAOSwTiVcFHLV:rk:4:pf:0
https://www.ebay.com.au/itm/2018-TEVO-Tarantula-3D-Printer-Kit-Prusa-I3-Aluminium-Frame-Auto-Leveling-PLUS/252681378402?_trkparms=aid%3D555018%26algo%3DPL.SIM%26ao%3D1%26asc%3D20140106155344%26meid%3Da1322029a06b403db841ca38ca9dc833%26pid%3D100005%26rk%3D2%26rkt%3D12%26mehot%3Dpp%26sd%3D392191265796%26itm%3D252681378402&_trksid=p2047675.c100005.m1851
https://www.ebay.com.au/itm/Geeetech-Smart-high-precision-small-Size-touch-screen-3D-printer-With-Free-Tax/392191265796?epid=18019886994&hash=item5b506bcc04:g:9qIAAOSwevdZ7req:rk:12:pf:0
Brad
Ive been looking at these. Great reviews
https://www.banggood.com/search/creality.html?brand=724&sbc=1
Mags
Quote from: Magluvin on 2019.01.01, 09:16:12
Ive been looking at these. Great reviews
https://www.banggood.com/search/creality.html?brand=724&sbc=1
Mags
Yes Mags
That seems to be the best out of the ones i listed.
Cheaper on ebay through the creality seller.
Brad
Quote from: TinMan on 2019.01.01, 04:06:53
Ok
I really have o get me a 3D printer.
Then i'll be needing bulk lessons Jim lol.
Brad
Geeze I didn't have time for a kit. I just wanted one off the shelf I could start using. I now know I need a bigger print bed and heated, plus dual extrusion capability. I'm primarily using tinkercad for design as it's what my grandkids use at primary school :) Finding out it's limitations as well.
Quote from: JimBoot on 2019.01.02, 00:35:00
Geeze I didn't have time for a kit. I just wanted one off the shelf I could start using. I now know I need a bigger print bed and heated, plus dual extrusion capability. I'm primarily using tinkercad for design as it's what my grandkids use at primary school :) Finding out it's limitations as well.
There is quite a bit of free software out there for 3d and 3d printing.
https://www.nchsoftware.com/meshmagic3d/index.html?msclkid=618cb297a3d31684c248eebb98c2180b
https://www.blender.org/
https://www.lifewire.com/free-3d-software-list-2005
https://all3dp.com/1/best-free-3d-modeling-software-3d-cad-3d-design-software/
http://www.3ders.org/3d-software/3d-software-list.html
And software to help perfect the printing like Cura and Slicer..
https://all3dp.com/1/best-free-3d-printing-software-3d-printer-program/
Mags
Well i have decided on this one-->the Anet A8
https://www.ebay.com.au/itm/On-Sale-Anet-A8-DIY-3D-Printer-Kit-1-75mm-0-4mm-Support-ABS-PLA-HIPS-AU/132796055691?hash=item1eeb43008b:g:~90AAOSwQ3tbq0KM:rk:3:pf:0
Many reasons for this--some being-->
It is only $172.00 AU delivered,and the quality is quite good.
It is a kit,and i love putting things together---many many hours of joy with lego technics and mechano in my childhood years O0
So many free upgrades for it,including the fire hazard fix ;D,and most upgrades you print your self.
A lot of support groups.
And can be made fully wifi enabled--no SD card for every print.
So,this will be my first 3D printer,as never had anything to do with them before.
So lets see how i go. :-\
The hardest part will be learning how to use some form of suitable designing program,so as i can design and print my own parts.
All these years i have been building parts by hand,so once i get this all happening,the builds will be more precise and easier--maybe even cheaper O0
Brad
Excellent! Looks pretty good. Double the print volume of mine, heated bed as well. Love the wifi onboard. I'll have to start digging for a dual extruder version.
Well i have been testing my drive coil,as has JimBoot i believe.
Seems our results are very similar,where we are getting either very large voltage spikes from the inductive kickback,or large current flow values when a heavy load is placed across the coil.
Our coils have only 200 turns,but the inductive kickback energy exceeds that of a coil of 400 turns plus-such as that of an SSG setup.
Maybe Jim can share his findings as well.
Brad
Quote from: TinMan on 2019.01.04, 04:59:12
Well i have been testing my drive coil,as has JimBoot i believe.
Seems our results are very similar,where we are getting either very large voltage spikes from the inductive kickback,or large current flow values when a heavy load is placed across the coil.
Our coils have only 200 turns,but the inductive kickback energy exceeds that of a coil of 400 turns plus-such as that of an SSG setup.
Maybe Jim can share his findings as well.
Brad
I'm seeing consistent spikes of 700v and the power supply goes haywire at 7 volts. My coil is 200 turns but only 33mm neo core. Burnt the circuit out so rebuilding that. I suspect the pn100. Edit: Placing a 16v 120000pf cap between +/- has it draw 1.2A, eliminate the spikes and speed up in a scary manner.
*Raises hand*
Have got an Anet A8 8)
I've had mine for about a year. Wait....am middle aged.....i've had mine for about a year and a half.
Give it a full 2 days to build it Brad. The Youtube instructions are really good, just pause and do. Some don't take the brown paper off that the parts ship in, but it looks far better in the gloss black.
For PLA, you don't need to heat the bed for each print. At your time of the year, the bloody thing would catch fire anyway if you did !
The weak points are the forward backwards of the bed all the time (oooerr missus) and the less than reliable power supply. Fix the bed wires by soldering the connections and using a good slightly flexible glue on the connections. Fix the power supply, by putting a 12V computer fan on its case. I've had no issues from either usual problem.
Try and keep the original tape on the bed for as long as possible, it seems to work best for adhesion of the print for the crucial first couple of layers. Am having adhesion troubles myself at the moment.
Software - all free
I use Tinkercad to design. It's easy, online and works:
https://www.tinkercad.com/
Thingiverse has a billion pre done 'things' that can be downloaded to print out.
https://www.thingiverse.com/
3D objects will be downloaded as .STL, so they need to be converted to .gcode for actual printing. I use CURA:
https://ultimaker.com/en/products/ultimaker-cura-software
Using 3D printable parts would ensure 100% replications were possible. I like the idea.
Speaking of which, Jim, do you have the STL for your coil form perchance ?
Quote from: Slider2732 on 2019.01.04, 19:47:06
*Raises hand*
Using 3D printable parts would ensure 100% replications were possible. I like the idea.
Speaking of which, Jim, do you have the STL for your coil form perchance ?
Think I can probably share via tinkercad. Looks like uploading of stls is verboten. It's had a few mods and I think it still may need work around thread size. Have done about 6 test prints so far tweaking each time .
Hey Mark,
I have setup a thingiverse account. https://www.thingiverse.com/thing:3334389 Apparently I have to wait 24 hours before it's published. Can always do a file transfer via skype too if you want or old school email :) Or when logged in to tcad https://www.tinkercad.com/things/iDOajWO2B2y-non-sense-rotor-poles-20mm
Quote from: Slider2732 on 2019.01.04, 19:47:06
*Raises hand*
Have got an Anet A8 8)
I've had mine for about a year. Wait....am middle aged.....i've had mine for about a year and a half.
Give it a full 2 days to build it Brad. The Youtube instructions are really good, just pause and do. Some don't take the brown paper off that the parts ship in, but it looks far better in the gloss black.
For PLA, you don't need to heat the bed for each print. At your time of the year, the bloody thing would catch fire anyway if you did !
The weak points are the forward backwards of the bed all the time (oooerr missus) and the less than reliable power supply. Fix the bed wires by soldering the connections and using a good slightly flexible glue on the connections. Fix the power supply, by putting a 12V computer fan on its case. I've had no issues from either usual problem.
Try and keep the original tape on the bed for as long as possible, it seems to work best for adhesion of the print for the crucial first couple of layers. Am having adhesion troubles myself at the moment.
Software - all free
I use Tinkercad to design. It's easy, online and works:
https://www.tinkercad.com/
Thingiverse has a billion pre done 'things' that can be downloaded to print out.
https://www.thingiverse.com/
3D objects will be downloaded as .STL, so they need to be converted to .gcode for actual printing. I use CURA:
https://ultimaker.com/en/products/ultimaker-cura-software
Using 3D printable parts would ensure 100% replications were possible. I like the idea.
Speaking of which, Jim, do you have the STL for your coil form perchance ?
Yes,it looks like a good litle printer.
I was going to put a fan on the power supply,and upgrade the bed wiring--have been watching many reviews and upgrade videos on them.
And yes,once the final device has all been designed and works,it can be replicated easily.
I should have gotten one years ago,but never to late O0.
Will be posting some interesting result here later tonight,on the drive coils performance.
Lets see it anyone can work this one out ???
Brad.
Quote from: TinMan on 2019.01.05, 04:38:22
Will be posting some interesting result here later tonight,on the drive coils performance.
Lets see it anyone can work this one out ???
Brad.
while we're waiting... :) I rebuilt the circuit. Here is a screenshot - I've put something arse backwards in this build as the spikes are all in the neg now. edit. added close up of wave with a cap. Last night it reduced the spikes considerably tonight it didn't. I was using a pn100 and are now using a 2222A in the circuit.
Just uploaded a short vid showing the effect without the motor running. https://youtu.be/k6AaKOpFrnU (first from the new shed)
Nice video, Jim! O0
Off to a good start.
Quote from: PhysicsProf on 2019.01.05, 13:55:27
Nice video, Jim! O0
Off to a good start.
so nice to have a bench again :)
Here's a shot of the wave in "self oscillation" . Have tested with air coil and it does not do it. Have not tested with a mag core only to see if it's that or a combo of it and the iron. So the 682 volt trough happens as the hall switch triggers. 6volts turns into over 600 across the coil, flipping domains in the iron PLA? Looking forward to Brad's findings. In stationery mode depending on where you hold the neo rotor, I get two diff waveforms. Both have troughs but one does not have the ringing as seen below .
Thanks for the upload Jim, it got 404'd though.
Am wanting to try the ceramic type magnets to see if the same power supply rattling spikes are there.....that supply on the video above was flickering like a good 'un !
Got a livingroom window to finish putting in with the better weather, but after that i'll see about blowing up some input circuits.
The thingiverse link you'll need to wait a few more hours, as it is my first upload so I'm sandboxed 24hrs
All my pn100s are still in the shipping container as gradually bring everything into the new shed. I've built a ferrocell over the drive coil and rotor. I suspect I won't see anything unexpected, just want to have a look at it when it's not running but oscillating
Quote from: JimBoot on 2019.01.06, 01:17:37
Here's a shot of the wave in "self oscillation" . Have tested with air coil and it does not do it. Have not tested with a mag core only to see if it's that or a combo of it and the iron. So the 682 volt trough happens as the hall switch triggers. 6volts turns into over 600 across the coil, flipping domains in the iron PLA? Looking forward to Brad's findings. In stationery mode depending on where you hold the neo rotor, I get two diff waveforms. Both have troughs but one does not have the ringing as seen below .
I would say so.
The field through the windings is being flipped very fast,and that results in a very high voltage across the coil.
I have been triggering mine with a transistor,and at a very low 2 volts from my power supply,i can blow out 240 volt neon's within seconds. It looks like lightning going on in those neon's,but this also results in transistor smoke as well ;D
My video is uploading ATM--it is long,and will be boring to most.
It;s late due to being a tad warm over here--was still just on 50*C in the workshop at 4pm.
So did not start bench work and video until around 10pm last night. Once finished,it was to late to upload.
Will post link to video as soon as it has uploaded O0
Brad
Quote from: TinMan on 2019.01.06, 09:02:04
I would say so.
The field through the windings is being flipped very fast,and that results in a very high voltage across the coil.
I have been triggering mine with a transistor,and at a very low 2 volts from my power supply,i can blow out 240 volt neon's within seconds. It looks like lightning going on in those neon's,but this also results in transistor smoke as well ;D
My video is uploading ATM--it is long,and will be boring to most.
It;s late due to being a tad warm over here--was still just on 50*C in the workshop at 4pm.
So did not start bench work and video until around 10pm last night. Once finished,it was to late to upload.
Will post link to video as soon as it has uploaded O0
Brad
Excellent look forward to watching the whole thing through. Yes soo many transistors. Nice to know it's doing it to the neons as well :) In self oscillation I get 1khz in cycles but in motor on mode only 400hz . No load to move, SSD potential? My new shed was only a balmy 36 on our recent 42.
The video on the coil testing so far.
https://www.youtube.com/watch?v=LyNCcnXbwdY
Will watch the rest tmrw. Nice vid so far beautiful coil. Is that new wire? My pipe is solid and I'm not seeing any diff in spikes by swapping coil polarity. Want to build some new coils now.
Your video is very well done TinMan. Your explanation is clear
and easy to follow as you go along.
Could we perhaps say that for purposes of driving your coil that
the Power Source is principally the Capacitor? Then your Power
Supply is simply restoring charge to the Capacitor to replenish
energy dissipated during the Transistor On Time while the Coil
was being energized?
It seems that the Capacitor is in a sense isolating the Power
Supply from the heavy pulse current demand of the Coil when
the Transistor energizes it.
But then there is the very interesting fact that the measured
RMS currents in the two portions of your analysis are somewhat
different. Hmmmm. Wonder why that is?
The Inductive Kickback Pulse is apparently quite effective in
restoring Charge to the Capacitor also?
If, after having started your circuit, you turn off your Power Supply.
How long would the circuit continue to operate with just the Charge
remaining in the Capacitor?
Quote from: Slider2732 on 2019.01.04, 19:47:06
Speaking of which, Jim, do you have the STL for your coil form perchance ?
I've printed my 6 poles now. https://www.thingiverse.com/thing:3334389
if someone designs a better version please share :)
Quote from: muDped on 2019.01.06, 14:52:37
Your video is very well done TinMan. Your explanation is clear
and easy to follow as you go along.
Could we perhaps say that for purposes of driving your coil that
the Power Source is principally the Capacitor? Then your Power
Supply is simply restoring charge to the Capacitor to replenish
energy dissipated during the Transistor On Time while the Coil
was being energized?
It seems that the Capacitor is in a sense isolating the Power
Supply from the heavy pulse current demand of the Coil when
the Transistor energizes it.
But then there is the very interesting fact that the measured
RMS currents in the two portions of your analysis are somewhat
different. Hmmmm. Wonder why that is?
The Inductive Kickback Pulse is apparently quite effective in
restoring Charge to the Capacitor also?
If, after having started your circuit, you turn off your Power Supply.
How long would the circuit continue to operate with just the Charge
remaining in the Capacitor?
The circuit continues to run for about 20 seconds after the supply is removed.
Brad
Quote from: JimBoot on 2019.01.06, 20:56:46
I've printed my 6 poles now. https://www.thingiverse.com/thing:3334389
if someone designs a better version please share :)
Great job on the spools Jim.
Brad
Great test vid Brad, enjoyed that, especially the swap over so the scope would have to be bull crapping twice!
So it looks like neo's are they way then, ok. How many need to be ordered?
Is it just 1 for the drive coil, then 4 ceramic types on the rotor, per build?
Am off to print Jim's spool design :)
Btw - USA TIP3055's are $2 minimum ($9.99 for 5) on Ebay, anyone got a cheap as chips alternative ? Can foresee some going pop.
Quote from: Slider2732 on 2019.01.07, 01:25:16
Great test vid Brad, enjoyed that, especially the swap over so the scope would have to be bull crapping twice!
So it looks like neo's are they way then, ok. How many need to be ordered?
Is it just 1 for the drive coil, then 4 ceramic types on the rotor, per build?
Am off to print Jim's spool design :)
Btw - USA TIP3055's are $2 minimum ($9.99 for 5) on Ebay, anyone got a cheap as chips alternative ? Can foresee some going pop.
I get my 3055s from ebay.
I buy a bulk lot of 20 for $11.00AU delivered.
Downside is that they come from china,so 3 to 4 weeks delivery time.
Brad
Quote from: Slider2732 on 2019.01.07, 01:25:16
Great test vid Brad, enjoyed that, especially the swap over so the scope would have to be bull crapping twice!
So it looks like neo's are they way then, ok. How many need to be ordered?
Is it just 1 for the drive coil, then 4 ceramic types on the rotor, per build?
Am off to print Jim's spool design :)
Btw - USA TIP3055's are $2 minimum ($9.99 for 5) on Ebay, anyone got a cheap as chips alternative ? Can foresee some going pop.
If you want to adjust I think you should still be able to ungroup the stl in tinkercad.
So i was thinking a little more about the coil typology.
What are the effects of having the PM as the core?
Well my thoughts are as follows--but may also be incorrect.
In a coil with a steel or ferrite core--
When we send a current through a coil,a magnetic field starts to build from 0. This creates an EMF that apposes the current flow from the source. Second-when that current flow is interupted,the magnetic field collapses back through the windings,back to 0. This creates our inductive kickback,where the voltage polarity inverts across the coil,but the current continues to flow in the same direction through the coil.
Now our coil with the PM as the core--
We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?
Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.
Any thoughts on this is welcome.
Brad
In regards to my last post,i have decided to build another coil the same as the one with the magnet core,but with a steel core,and do a side by side comparison.
Brad
Quote from: TinMan on 2019.01.07, 05:16:52
In regards to my last post,i have decided to build another coil the same as the one with the magnet core,but with a steel core,and do a side by side comparison.
Brad
That would be surprising to me and bloody annoying if you're right. My limited knowledge thought that our iron held the polarity of the mag and when saturated by the coil, switched polarity and then switched back. If it holds no polarity from the start don't we miss something? And I have no idea on the effect of my massive neo rotor on what I'm seeing.
It will be annoying because of the amount of times I tested motors with ferrite cores back in the day. I had even less of an idea back then and no scope.
Quote from: TinMan on 2019.01.06, 22:37:28
Great job on the spools Jim.
Brad
Thanks, they're a little rough. I designed them that way for my tiny printer. It's probably an easier job to print a rotor with hollow threaded poles and separate threaded capped spools. I'm still learning the tricks but tinkercad had been the most intuitive for me. That's where you design then you get in print ready in Cura. Then you print from your posh wifi connection.
Quote from: JimBoot on 2019.01.07, 11:25:16
Thanks, they're a little rough. I designed them that way for my tiny printer. It's probably an easier job to print a rotor with hollow threaded poles and separate threaded capped spools. I'm still learning the tricks but tinkercad had been the most intuitive for me. That's where you design then you get in print ready in Cura. Then you print from your posh wifi connection.
I am going o have to learn to use tinker cad,or some form of design program.
You may have to run me through it Jim lol.
Ok,one for the power measuring guru's
Do we use RMS values,or average values when calculating P/in on a DC system,but where the current has a sawtooth waveform?.
Below i have supplied the circuit with scope probe placement,along with the scope shots that include-RMS value's,average values,and math calculated by the scope.
Can one of the EE guys here please calculate the correct values for each.
What i want to know is the power flowing into the cap from the source,and the power flowing through the coil during the on time.
Thanks in advance O0
P.S
I have added the full size scope shot's,as the resized ones are not very clear.
Brad
Hi Brad,
neither being a Guru nor an EE, but i think all power measurements via the scope math require the input
voltage / current to be in rms, and the outcome in mean (average).
Itsu
Brad,
I'm with Itsu but I will recommend what is needed but not being familiar with your scope, this will possibly require help from someone who has the same model as yours to set it up.
To determine the accurate Pin let's say in your circuit, every instantaneous value of voltage v and current i must be sampled and multiplied together with each product stored in a table. After a known period of time t, all the products accumulated in the table during t are added together and averaged which will represent the average power consumed at the input. Fortunately, the Math channel on your scope does this process for us but the result must be displayed in Mean or average output. In your particular case, this process can be repeated for the output measurement.
The time period t your Math channel will use is created by taking measurements of your waveforms between two vertical cursors that are precisely positioned so that a complete cycle or number of cycles are included between them. The time period you are now using is from screen edge to screen edge which will get you close but is not the most accurate.
I'll assume your scope has 8 bit vertical resolution so the maximum accuracy in this axis is with the waveform peaks reaching as close to full screen deflection as possible without clipping.
For maximum accuracy in the horizontal axis, select the highest number of horizontal samples or sample rate available which will give the highest number of products to average.
Hope this makes sense and is helpful.
Pm
Hi Brad,
I come from the age of analogue oscilloscopes, but I do appreciate that modern digital scopes take samples of the analogue waveforms. Thus you have the capability of doing math on individual samples. The correct method is to multiply each voltage sample by the current sample occurring at the same point in time and then to take the average of all those multiplications over one or more complete cycles. I think the scope should be able to do that. If you wish to multiply the rms value of current by the rms value of voltage and then take the average this must also be done over complete cycles. Not having a digital scope I can't advise any further.
Smudge.
Edit PM posted while I wrote this.
Good news on the Anet A8, it seems to be printing fine now, on the 4th try.
DIsregard my Youtube question Jim, as the forum came back up :)
I added 'supports' in Cura and the printer seems to be fine. Will find out in about another 30 minutes. 51 minute print btw, on a 3mm rough/fast setting.
Can't seem to edit the Tinkercad STL though...I might well be missing something.
You only seem to get grief from me lately :-*
Update: Printed fine
You might be able to see how support has been added under the left hand side threads and between the coil ends. It just breaks off, as it's only 1 layer. Not a bad print for 3mm, now to clean it up.
Quote from: partzman on 2019.01.07, 19:13:42
Brad,
I'm with Itsu but I will recommend what is needed but not being familiar with your scope, this will possibly require help from someone who has the same model as yours to set it up.
To determine the accurate Pin let's say in your circuit, every instantaneous value of voltage v and current i must be sampled and multiplied together with each product stored in a table. After a known period of time t, all the products accumulated in the table during t are added together and averaged which will represent the average power consumed at the input. Fortunately, the Math channel on your scope does this process for us but the result must be displayed in Mean or average output. In your particular case, this process can be repeated for the output measurement.
The time period t your Math channel will use is created by taking measurements of your waveforms between two vertical cursors that are precisely positioned so that a complete cycle or number of cycles are included between them. The time period you are now using is from screen edge to screen edge which will get you close but is not the most accurate.
I'll assume your scope has 8 bit vertical resolution so the maximum accuracy in this axis is with the waveform peaks reaching as close to full screen deflection as possible without clipping.
For maximum accuracy in the horizontal axis, select the highest number of horizontal samples or sample rate available which will give the highest number of products to average.
Hope this makes sense and is helpful.
Pm
Mmm
Well this is why i asked.
I assumed that the math function on the scope was designed to do all this for us,and calculate the actual (real) power being used,taking phase angle also into account.
None the less,i also measured the temperature across both CVR 1 and CVR2. CVR2s temperature is higher than that of CVR1s temperature. This is another indication that the average current flowing through the coil is higher than that flowing in from the source,where the voltage remains almost constant.
I would have posted the video last night,but the forum went down--something about maximum band width been exceeded.
But i will post that video here tonight,or you can just go watch my latest video.
In the last video,i also measure the power being dissipated by the 12 volt LED,which for memory was another 128mW.
Unless some one can point out any mistake in the power measurements,i would say that this particular coil arrangment and simple circuit is running at an efficiency of 200% +.
Brad
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.
I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled?
Quote from: Slider2732 on 2019.01.08, 03:20:46
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.
I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled?
The stakes on the cheap garden solar lights over here are SS-304 id say.
Yes,the PVC i am using is class 6 13mm--or 1/2 inch if you prefer.
Brad
@JimBoot
I have been using a Robo 3D and design on Solid Works. I have never been able to edit an .stl with it. Might be good to upload the dwg file if you are ok with edits.
By the way, DO NOT GET THIS PRINTER. Support is nonexistent and am sure I received a refurb for a new one. Case broke in transit and will not even talk to me. I am ordering a new one, Creality CR-10S to replace it.
@TinMan
How are you calculating the winds necessary to offset the magnet insert? I am trying to figure out how to offset armature reaction in a gen head just enough to null it. Would be a great help.
thay
Quote from: TinMan on 2019.01.08, 03:46:02
The stakes on the cheap garden solar lights over here are SS-304 id say.
Yes,the PVC i am using is class 6 13mm--or 1/2 inch if you prefer.
Brad
So i had an idea in regards to this power measurement issue.
My claim is that there is more power flowing through the coil during each pulse than there is being delivered by the source-that being my power supply.
So far i have shown this by way of scope measurements,and heat dissipated by the CVRs,where CVR1 dissipates less heat than CVR2,meaning more current flowing through CVR2,and thus the coil.
Now,what if i was to put a small HHO cell in series with the coil,so as it is between the positive of the cap and input to the coil. I then run the system,and calculate the amount of HHO produced in 5 minutes. I then drive that HHO cell with my power supply only,at the input power it was when driving the circuit,and once again calculate the volume of HHO produced in 5 minutes. What would it mean if more HHO was produced by the coil circuit than that produced by running the HHO cell straight of the power supply at the same amount of power?.
Would this convince you that the coil and circuit is producing its own energy ?
If not,then why are we getting the results we have ?
Brad
Quote from: Thaelin on 2019.01.08, 04:46:53
@JimBoot
I have been using a Robo 3D and design on Solid Works. I have never been able to edit an .stl with it. Might be good to upload the dwg file if you are ok with edits.
By the way, DO NOT GET THIS PRINTER. Support is nonexistent and am sure I received a refurb for a new one. Case broke in transit and will not even talk to me. I am ordering a new one, Creality CR-10S to replace it.
@TinMan
How are you calculating the winds necessary to offset the magnet insert? I am trying to figure out how to offset armature reaction in a gen head just enough to null it. Would be a great help.
thay
It is not so much-how many turns,it's more-how much current required to flow through the coil with a set amount of turns before a neutral field is had.
Brad
OK,here is a little something else for you.
This test at 24 volts,4% duty cycle,@ 86 Hz
Below is a scope shot(with circuit and scope probe positions attached) showing the inductive kickback cycle,and also shows that when the transistor is on,there is 24 volts across the coil(minus the V drop across CVR2). CH1 is on the positive side of the coil,and CH3 is on the negative side. As soon as the transistor conducts,we see CH3 drop to 0 volts-minus the voltage drop across CVR2. This tells us we have close to 23 volts across the coil during the current flow period.
The second thing it shows us,that stands out,is the duration of the inductive kickback cycle O0
Have you ever seen the duty cycle of the inductive kickback exceed that of the input pulse by this much--or at all ?.
Remember my theory i posted-->reply 75 Quote:
We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?
Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.
Well lets have a look at the second scope shot i have posted below.
We are looking at CH4-the blue chanel.
This is the waveform created across a second coil placed face on to our drive coil-->our coil under test.
The second coil has a 220 ohm resistor across it as a load.
What do you see ? 8)
Brad
Quote from: Slider2732 on 2019.01.08, 03:20:46
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.
I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled?
I picked up some black iron pipe nipples from Lowes. They call it black steel pipe but I am pretty sure it is identical to the black iron pipe we used to sell at the hardware store I once worked in. It seems to be working ok with the tests I have done so far. I turned the inside and outside on a lathe to true it up for a better fit over the plastic sleeve and a smoother surface for wrapping tape on before adding the windings. I guess it was made in China because the inside and outside were pretty irregular.
Carroll
Quote from: CITFTA on 2019.01.08, 13:26:42
I picked up some black iron pipe nipples from Lowes. They call it black steel pipe but I am pretty sure it is identical to the black iron pipe we used to sell at the hardware store I once worked in. It seems to be working ok with the tests I have done so far. I turned the inside and outside on a lathe to true it up for a better fit over the plastic sleeve and a smoother surface for wrapping tape on before adding the windings. I guess it was made in China because the inside and outside were pretty irregular.
Carroll
Hi Carroll
When you get the time,can you set up a circuit the same as mine(posted above),and see if you get the same result's--just for confirmation O0
Thanks
Brad
Quote from: TinMan on 2019.01.08, 13:20:39
OK,here is a little something else for you.
This test at 24 volts,4% duty cycle,@ 86 Hz
Below is a scope shot(with circuit and scope probe positions attached) showing the inductive kickback cycle,and also shows that when the transistor is on,there is 24 volts across the coil(minus the V drop across CVR2). CH1 is on the positive side of the coil,and CH3 is on the negative side. As soon as the transistor conducts,we see CH3 drop to 0 volts-minus the voltage drop across CVR2. This tells us we have close to 23 volts across the coil during the current flow period.
The second thing it shows us,that stands out,is the duration of the inductive kickback cycle O0
Have you ever seen the duty cycle of the inductive kickback exceed that of the input pulse by this much--or at all ?.
Remember my theory i posted-->reply 75 Quote:
We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?
Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.
Well lets have a look at the second scope shot i have posted below.
We are looking at CH4-the blue chanel.
This is the waveform created across a second coil placed face on to our drive coil-->our coil under test.
The second coil has a 220 ohm resistor across it as a load.
What do you see ? 8)
Brad
This reminds me of the Robert Adams Pulse Motor in the sense that he found a sweet spot of operation that sounds similar to what you are describing.
I purchased all of the Adams material from Nexus if you are interested in PDF copies.
Dave
Hi Brad
At first look I thought I was looking at a boost circuit, then I realised how you have wired it. ???
My take on this is :-
The transistor switches on, the coil charges compressing the fixed (PM) field into the coil. The transistor switches off and the charged current in the coil wants to run in the same direction as the charge, but this time the charge of the coil is going against the incoming supply, no ground. So what is creating the current? Well my guess is the decompression of the PM magnetic field is the only power source adding to the battery voltage. It is a magnetic spring but without resonance, one shot compress and expand, then back to normal.
That is my first thought, and may not be right.
Regards
Mike 8)
PS. doing something similar but with only coils.
Just looked again, no all the power is coming from the compression of the PM, no power can go beyond the coil unless it comes from the coil, due to the diode being where it is :D
Thanks Brad, Carroll.
Have just checked their website and may have to go in person to our local store. A search for black pipe seems to come up with PVC products.
Am wondering if they think people will use real iron for car lot/road rage fights.
Got a coil wound, using the steel sheet. I had to use foam for the ends and they kept coming off, even with glue, but it has 200 turns. Ceramics inside, while I wait for the 1/2" x 2" N48 neo's. Have ordered 2. Just a test coil, something to get started with.
Quote from: TinMan on 2019.01.08, 13:37:54
Hi Carroll
When you get the time,can you set up a circuit the same as mine(posted above),and see if you get the same result's--just for confirmation O0
Thanks
Brad
Hi Brad,
I did put together the test circuit like you have shown. My test coil is NOT exactly the same as yours. When you posted you thought we should probably cut a slit down the side of the pipe I had already finished my coil. So my pipe does not have a slit down the side to prevent it from acting like a shorted turn. When I did some tests to see how much current it would take to neutralize the magnetic attraction it took several amps. I am thinking now it is because of the pipe not having the slit cut in it. So I rewound the coil with 275 feet of wire. By the way 200 turns comes out to be 51 feet of wire if you wind it on the pipe of the size you recommended. So I have increased the length of the wire by a factor of over 5 times. I now believe you probably wanted the shorter wire so that you could hit the coil with a very short high amperage pulse. You can clarify that if you wish. Anyway I am now planning to make a new coil with the split in the side of the iron pipe and the correct number of turns of wire.
Okay, now for some test results. I have not had time to take pictures or a video of what I did but hopefully anyone that has been following your tests will understand my results.
I connected my scope leads just as you have shown in your schematic. I was using my Rigol DS1102E. Channel one across the supply and channel 2 first across CR1 and then across CR2. I did not bother to use the math functions for two reasons. I haven't learned how yet. LOL And since channel one across the supply never changed I realized all I needed to do was look at what I could see on channel 2.
I measured both CR1 and CR2 with a precision meter for measuring ohms into the milliohm range. Neither of my resistors were rated as precision resistors but amazing they both measured the same at 10.18 ohms.
Since my coil has a lot more wire on it I had to tune the function generator to find the sweet spot for the frequency. I kept the pulse width at the 5% you recommended. I found the best transfer of energy to the second cap to be when my frequency was 22khz.
I measured the voltage across CR1 and CR2 a couple of ways and got some strange results. When I measured the RMS values I got 1.44 volts across CR1 and 1.88 volts across CR2. I was using a regulated power supply set to 9.02 volts instead of a battery as I wanted to keep the input power steady. I believe these results are similar to what you were getting with a higher RMS value across CR2 than across CR1.
However what seems really strange is that my scope showed the average voltage across the resistors to be the opposite with the drop across CR1 being 550 mv and the average across CR2 being 235 mv. Of course the much lower value for the average voltage across the resistors is because of the very short on time of the pulses. But why the average value shows the opposite of the RMS value is something someone much smarter than me will have to figure out. Maybe you can do the test again and see what you scope shows for the average value of voltage drop across the resistors.
Take care,
Carroll
Quote from: Slider2732 on 2019.01.07, 19:53:50
Good news on the Anet A8, it seems to be printing fine now, on the 4th try.
DIsregard my Youtube question Jim, as the forum came back up :)
I added 'supports' in Cura and the printer seems to be fine. Will find out in about another 30 minutes. 51 minute print btw, on a 3mm rough/fast setting.
Can't seem to edit the Tinkercad STL though...I might well be missing something.
You only seem to get grief from me lately :-*
Update: Printed fine
You might be able to see how support has been added under the left hand side threads and between the coil ends. It just breaks off, as it's only 1 layer. Not a bad print for 3mm, now to clean it up.
Ok here you go - https://www.tinkercad.com/things/iDOajWO2B2y-non-sense-rotor-poles-20mm you should be able to tinker directly and ungroup. the thread support sucks to get off which is why I printed lying them down. Got 4 grandkids here for the week plus back at work, so spread a bit thin this week. :)
Well I decided to do the tests again. I realized after watching Brad's video again that I was using way too high of a frequency. So I redid the test starting out with a frequency of 10 hz. I also raised my power supply voltage to 12.41 volts. When I got my frequency up to 200 hz that seemed to give the best transfer of energy to the second cap with the Led across it.
Checking my scope readings again I got 1.85 volts RMS across CR1 and 8.95 volts RMS across CR2! A big change in difference between the two. When I checked the average voltage across CR1 I got 1.75 volts and across CR2 I got 2.42 volts. So this time the scope agreed that more energy was going through CR2 than CR1 with both the RMS values and the average values.
My only conclusion from all this is that these digital scopes have problems with short pulse duration pulses. Anyone else have any ideas?
Carroll
Ok here's a draft of my rotor for the poles to screw into. I have not printed this yet. The learning curve was long for me to work out how to place threaded holes evenly in a 3D disc. The axle hole was designed for my 6mm ex Fax machine axle. I wanted a snug fit. The size is easily changed in in tinkercad using the Tire parameters. https://www.tinkercad.com/things/3lMv1XNqPID-non-sense-rotor-25mm
Quote from: Thaelin on 2019.01.08, 04:46:53
@JimBoot
I have been using a Robo 3D and design on Solid Works. I have never been able to edit an .stl with it. Might be good to upload the dwg file if you are ok with edits.
By the way, DO NOT GET THIS PRINTER. Support is nonexistent and am sure I received a refurb for a new one. Case broke in transit and will not even talk to me. I am ordering a new one, Creality CR-10S to replace it.
I've only done around 45 projects in tinkercad so I am green. It will only allow export of svg, stl or obj files. But you cand edit direct in tinkercad. I have not used solidworks but I believe it's the ducks nuts. https://www.tinkercad.com/things/iDOajWO2B2y-non-sense-rotor-poles-20mm
Quote from: CITFTA on 2019.01.08, 22:38:26
Hi Brad,
I did put together the test circuit like you have shown. My test coil is NOT exactly the same as yours. When you posted you thought we should probably cut a slit down the side of the pipe I had already finished my coil. So my pipe does not have a slit down the side to prevent it from acting like a shorted turn. When I did some tests to see how much current it would take to neutralize the magnetic attraction it took several amps. I am thinking now it is because of the pipe not having the slit cut in it. So I rewound the coil with 275 feet of wire. By the way 200 turns comes out to be 51 feet of wire if you wind it on the pipe of the size you recommended. So I have increased the length of the wire by a factor of over 5 times. You can clarify that if you wish. Anyway I am now planning to make a new coil with the split in the side of the iron pipe and the correct number of turns of wire.
Okay, now for some test results. I have not had time to take pictures or a video of what I did but hopefully anyone that has been following your tests will understand my results.
I connected my scope leads just as you have shown in your schematic. I was using my Rigol DS1102E. Channel one across the supply and channel 2 first across CR1 and then across CR2. I did not bother to use the math functions for two reasons. I haven't learned how yet. LOL And since channel one across the supply never changed I realized all I needed to do was look at what I could see on channel 2.
I measured both CR1 and CR2 with a precision meter for measuring ohms into the milliohm range. Neither of my resistors were rated as precision resistors but amazing they both measured the same at 10.18 ohms.
Since my coil has a lot more wire on it I had to tune the function generator to find the sweet spot for the frequency. I kept the pulse width at the 5% you recommended. I found the best transfer of energy to the second cap to be when my frequency was 22khz.
I measured the voltage across CR1 and CR2 a couple of ways and got some strange results. When I measured the RMS values I got 1.44 volts across CR1 and 1.88 volts across CR2. I was using a regulated power supply set to 9.02 volts instead of a battery as I wanted to keep the input power steady. I believe these results are similar to what you were getting with a higher RMS value across CR2 than across CR1.
However what seems really strange is that my scope showed the average voltage across the resistors to be the opposite with the drop across CR1 being 550 mv and the average across CR2 being 235 mv. Of course the much lower value for the average voltage across the resistors is because of the very short on time of the pulses. But why the average value shows the opposite of the RMS value is something someone much smarter than me will have to figure out.
Take care,
Carroll
Hi Carroll
I only operate in a frequency range between 50 and 100 Hz.
This is around the frequency we can obtain from the RPM of our rotor,with the 4 to 6 pulses per revolution.
QuoteMaybe you can do the test again and see what you scope shows for the average value of voltage drop across the resistors.
Yes,did that some posts back,but i have attached the results below.
As you can see,my average voltage across CVR 2 is also close to double to that of CVR1
Confirmation this is correct is--CVR2s temperature is around 5*C hotter than CVR1
QuoteI now believe you probably wanted the shorter wire so that you could hit the coil with a very short high amperage pulse.
Yes,high current ,short pulses for several reasons--
1-low I/R losses
2-will run fine on low voltages
3-produces a strong magnetic field to both pust the rotor magnet away,and induce rotor coil with strong magnetic field.
Brad
Hey Jim:
Every link you gave just sends me to AutoDesk instead of a d/l. Just me?
Ok,lets see this one explained away.
I spent most of last night thinking (instead of sleeping)about how i could go about proving this another way.
It then hit me-->show in a physical way that there is more energy flowing through CVR2 than there is flowing through CVR1.
So here is what i did--
I removed both CVRs,and replaced each with a capacitor/DC motor combo.
Both caps,motors,and props are identical,and i have swapped each around to make sure the results are the same.
Even though we can see the results physically,i also placed my scope across each cap,so as we could see the average voltage across each motor/cap combo.
The only possible way one motor can start up before the other,is if more current is flowing through it,resulting in more power being delivered to it.
Take a good look at the circuit.
The source is not in series with the inductor,it is in parallel with C1.
This means that the source supplies the energy to C1,which supplies the energy to the circuit.
The voltage across the coil is the voltage across C1,minus V/drop across CVR2 during the on time.
Below is the circuit and scope probe placements.
And the video of the visual test O0
https://www.youtube.com/watch?v=vSODFYCqQH8
Brad
Quote from: Thaelin on 2019.01.09, 04:18:30
Hey Jim:
Every link you gave just sends me to AutoDesk instead of a d/l. Just me?
Ok yeah - that's if you want to edit it with their software. It's where I created it. I'm not sure how to give you a downloadable editable file with all the component parts.
Quote from: TinMan on 2019.01.09, 04:23:46
Ok,lets see this one explained away.
I spent most of last night thinking (instead of sleeping)about how i could go about proving this another way.
It then hit me-->show in a physical way that there is more energy flowing through CVR2 than there is flowing through CVR1.
So here is what i did--
I removed both CVRs,and replaced each with a capacitor/DC motor combo.
Both caps,motors,and props are identical,and i have swapped each around to make sure the results are the same.
Even though we can see the results physically,i also placed my scope across each cap,so as we could see the average voltage across each motor/cap combo.
The only possible way one motor can start up before the other,is if more current is flowing through it,resulting in more power being delivered to it.
Take a good look at the circuit.
The source is not in series with the inductor,it is in parallel with C1.
This means that the source supplies the energy to C1,which supplies the energy to the circuit.
The voltage across the coil is the voltage across C1,minus V/drop across CVR2 during the on time.
Below is the circuit and scope probe placements.
And the video of the visual test O0
https://www.youtube.com/watch?v=vSODFYCqQH8
Brad
Great video Brad. O0
I've just designed this solid spool in Tinkercad . I'll be printing it in iron PLA (minus the keepers) to use as my iron core coil. It's dimensions are easily adjustable in tinkercad. https://www.tinkercad.com/things/arwolzR7kv8-solid-spool
Quote from: JimBoot on 2019.01.09, 07:44:01
I've just designed this solid spool in Tinkercad . I'll be printing it in iron PLA to use as my iron core coil. It's dimensions are easily adjustable in tinkercad. https://www.tinkercad.com/things/arwolzR7kv8-solid-spool
What is the solid spool for?
And you do not want the two ends to be magnetic.
Brad
Quote from: TinMan on 2019.01.09, 10:07:42
What is the solid spool for?
And you do not want the two ends to be magnetic.
Brad
- Yeah thought you'd say that. You can edit the keepers out and have a nicely flared core, then simply glue some keepers on. I'm also always looking for spools, so at least now I can print one in whatever material I choose. Post edited.
Quote from: JimBoot on 2019.01.09, 07:44:01
I've just designed this solid spool in Tinkercad . I'll be printing it in iron PLA (minus the keepers) to use as my iron core coil. It's dimensions are easily adjustable in tinkercad. https://www.tinkercad.com/things/arwolzR7kv8-solid-spool
What is the iron cored coil for? ???
Brad
Quote from: TinMan on 2019.01.09, 11:51:18
What is the iron cored coil for? ???
Brad
sorry just for experimentation not for actual motor
Quote from: TinMan on 2019.01.09, 04:23:46
Ok,lets see this one explained away.
I spent most of last night thinking (instead of sleeping)about how i could go about proving this another way.
It then hit me-->show in a physical way that there is more energy flowing through CVR2 than there is flowing through CVR1.
So here is what i did--
I removed both CVRs,and replaced each with a capacitor/DC motor combo.
Both caps,motors,and props are identical,and i have swapped each around to make sure the results are the same.
Even though we can see the results physically,i also placed my scope across each cap,so as we could see the average voltage across each motor/cap combo.
The only possible way one motor can start up before the other,is if more current is flowing through it,resulting in more power being delivered to it.
Take a good look at the circuit.
The source is not in series with the inductor,it is in parallel with C1.
This means that the source supplies the energy to C1,which supplies the energy to the circuit.
The voltage across the coil is the voltage across C1,minus V/drop across CVR2 during the on time.
Below is the circuit and scope probe placements.
And the video of the visual test O0
https://www.youtube.com/watch?v=vSODFYCqQH8
Brad
Ha now I see what you are getting at.
There is a time difference on your measurements, CVR1 is 95% on to charge the C1 and 5% on through the CVR2 when the transistor is on. The thing is when the transistor is on power is also coming from the C1 at a higher current, as you know caps give up current near instantly, this adds to the current from the supply. So you need to put the CVR2 between the C1 negative and the ground connection so as to measure the power from the cap C1 via CVR2, and the power from the supply via CVR1.
I think I have that right
Regards
Mike 8)
PS the other post I made was looking at another part of it which does seem an anomaly ;)
Corrected the cap polarity
Quote from: Centraflow on 2019.01.09, 12:16:36
Ha now I see what you are getting at.
There is a time difference on your measurements, CVR1 is 95% on to charge the C1 and 5% on through the CVR2 when the transistor is on. The thing is when the transistor is on power is also coming from the C1 at a higher current, as you know caps give up current near instantly, this adds to the current from the supply. So you need to put the CVR2 between the C1 positive and the ground connection so as to measure the power from the cap C1 via CVR2, and the power from the supply via CVR1.
I think I have that right
Regards
Mike 8)
PS the other post I made was looking at another part of it which does seem an anomaly ;)
Err--so you want to put CVR2 from positive of C1(which is also the positive of the supply),to ground,which is the negative of the power supply--> ???
At 12 volts,and the CVR being 1 ohm,that would result in 12 amps flowing through CVR2 continuously.
I think you have something mixed up there Mike :D
Brad
Quote from: TinMan on 2019.01.09, 12:25:07
Err--so you want to put CVR2 from positive of C1(which is also the positive of the supply),to ground,which is the negative of the power supply--> ???
At 12 volts,and the CVR being 1 ohm,that would result in 12 amps flowing through CVR2 continuously.
I think you have something mixed up there Mike :D
Brad
Sorry Brad, I corrected my mistake, it is the negative of the c1 cap to ground :-[
Quote from: Centraflow on 2019.01.09, 12:31:06
Sorry Brad, I corrected my mistake, it is the negative of the c1 cap to ground :-[
But CVR1 is from the negative of the cap to ground.
See circuit below.
CVR1 measures input current from source.
CVR2 measures current flowing through the coil.
Brad
Hi Brad,
Do you plan to remove the permanent magnet from the coil so that you are left with the iron sleeve as the core? I believe that would be a step forward to find out where the 'extra' the measurements show comes from.
Or go on the direction to utilize the 'extra' in the 2nd coil's output you kindly showed as a much longer time duration than the input pulse width was. Of course it is possible that if the 2nd coil is moving and not stationary, then thing change either for worse or better.
Just thinking loudly, I do not mean to direct you in any way.
Thanks for sharing.
Gyula
Well it would seem i made a mistake--i think.
After looking at my own circuit,it appears to me that C1 is not receiving any of the inductive kickback energy-->unless some how current can still flow through the transistor when it is open ?
If we look at the circuit,when the transistor is open,then the inductive kickback current can only flow through the coil/cap/LED loop,as depicted by the red arrows.
So,this brings even more confusion-->where is the extra current coming from that is flowing through CVR2 ?
As i did not get a definitive answer on how to correctly calculate the power measurements using the scope,i went on the hunt for a thread that covered this very thing.
I found this one by Poynt,but Smudge's post seem to make the most sense to me
http://www.overunityresearch.com/index.php?topic=704.0
Quote Smudge :For simple continuous waveforms we don't need to take samples and do the math, there are standard values relating peak or peak-to-peak to rms and we use those. But modern digital 'scopes can do math and they do not use those simple values, they calculate it properly from the train of samples so they can give correct rms values for arbitrary waveforms. But there is a catch here, the averaging has to be done over an integral number of cycles else the result is wrong. For a continuous waveform where the samples just keep coming the 'scope math averaging effectively takes place over many many cycles so the error for not using an integral number becomes insignificant. But this is not so for single shot or a small number of cycles, so there you need care in using the scope. Some scopes allow you to set cursors to select integral cycles.
Some scopes can multiply the voltage and current waveforms to display the power waveform. Its "DC" value is the average. But again the average must be taken over an integral number of cycles. Just because scopes can give you rms or mean there appears to be confusion over the use of the rms function applied to the power waveform. For us engineers this has no real meaning and is useless. It should not be used. In the electrical world the RMS function gives you an effective DC value for the power capability of a voltage or current waveform and the M stands for Mean. That is the Mean of a power waveform, not the RMS of a power waveform.
Ok,so if i understand this correctly,then CH1(voltage) and CH2(current) are multiplied together using there RMS values.
The math function must then be set to display the average value(not the RMS value)of that multiplication result,and this is our DC value of power--is that correct?.
Brad
Quote from: gyula on 2019.01.09, 13:03:41
Hi Brad,
Do you plan to remove the permanent magnet from the coil so that you are left with the iron sleeve as the core? I believe that would be a step forward to find out where the 'extra' the measurements show comes from.
Or go on the direction to utilize the 'extra' in the 2nd coil's output you kindly showed as a much longer time duration than the input pulse width was. Of course it is possible that if the 2nd coil is moving and not stationary, then thing change either for worse or better.
Just thinking loudly, I do not mean to direct you in any way.
Thanks for sharing.
Gyula
I believe that with the moving magnet/coil on the rotor,the effect will be amplified some what due to the motion of the rotor magnets. We will see a generating effect as the rotor coil/magnet approaches the drive coil/magnet combo. This we dont have ATM,due to there being no motion of the fixed second coil.
Brad
Quote from: gyula on 2019.01.09, 13:03:41
Hi Brad,
Do you plan to remove the permanent magnet from the coil so that you are left with the iron sleeve as the core? I believe that would be a step forward to find out where the 'extra' the measurements show comes from.
Thanks for sharing.
Gyula
My magnets are glued in,but i should be able to get them out,as not glued in that well.
I will give that a try,and see what happens O0
Brad
Quote from: TinMan on 2019.01.09, 13:21:02
...
Ok,so if i understand this correctly,then CH1(voltage) and CH2(current) are multiplied together using there RMS values.
The math function must then be set to display the average value(not the RMS value)of that multiplication result,and this is our DC value of power--is that correct?.
Brad
Yes it sounds correct but it is a question if a scope has such feature to choose that. And add to that the use of the integral number of cycles, assigned even between vertical cursors if needed.
Gyula
EDIT: Thanks for the other reply you posted in the meantime a minute ago.
Quote from: gyula on 2019.01.09, 13:35:40
Yes it sounds correct but it is a question if a scope has such feature to choose that. And add to that the use of the integral number of cycles, assigned even between vertical cursors if needed.
Gyula
EDIT: Thanks for the other reply you posted in the meantime a minute ago.
Well after reading this thread,im even more confused
http://www.overunityresearch.com/index.php?topic=786.25
To quote poynt
Using a ~3kHz square wave (~50% duty) drive from a 555 circuit (more or less the original Ainslie circuit built using Groundloop's PCB), I was able to confirm what the Pin measurement was with ~15VDC supply input through the switching circuitry on a 10 Ohm inductive resistor.
The 100W load resistor got up to about 64ºC, and the measured Pin's using 4 different methods were as follows:
1) PSU meters: 0.68A x 15.2V = 10.3W
2) Scope: MEAN[vbat(t) x 4icsr(t)] (4i to account for the 0.25 Ohm CSR) = 9.88W
3) Scope: MEAN[vbat(t)] x MEAN[4icsr(t)] = 9.91W
4) DMM: 15.18VDC x 158.2mVDC(Vcsr) x 4 = 9.6W
No external filtering was used for the DMM measurement, and the signals were fairly "clean" with little ringing. So far, so good.
This is the confusing part-->
2) Scope: MEAN[vbat(t) x 4icsr(t)] (4i to account for the 0.25 Ohm CSR) = 9.88W
3) Scope: MEAN[vbat(t)] x MEAN[4icsr(t)] = 9.91W
Mean is average-right?,as my scope only has average,and not mean.
Now,im sure Poynt told me that you do not average both voltage and current to get power measurements.
Is it not suppose to be max voltage x average current,or max current by average voltage to get power?
Im sure you do not average out both values to get your real power value ???
Poynt says,using the scope,we multiply the average(mean) voltage by the average(mean) current to get the real power value.
If anyone could clear this measuring with scope mess up,that would be great O0.
Brad
Brad,
just use CH1 x CH2 for your scope (it does not matter how you display those 2 signals (rms, average, p2p) as
the scope Math function does it its own thing with them (like Partzman said: "every instantaneous value of
voltage v and current i must be sampled and multiplied together with each product stored in a table etc.).
The important thing is to have the output (power) in average (not rms as you now have it).
You then have the average output power.
Itsu
Quote from: TinMan on 2019.01.09, 13:59:10
Well after reading this thread,im even more confused
This is the confusing part-->
2) Scope: MEAN[vbat(t) x 4icsr(t)] (4i to account for the 0.25 Ohm CSR) = 9.88W
3) Scope: MEAN[vbat(t)] x MEAN[4icsr(t)] = 9.91W
Mean is average-right?,as my scope only has average,and not mean.
Now,im sure Poynt told me that you do not average both voltage and current to get power measurements.
Is it not suppose to be max voltage x average current,or max current by average voltage to get power?
Im sure you do not average out both values to get your real power value ???
Poynt says,using the scope,we multiply the average(mean) voltage by the average(mean) current to get the real power value.
If anyone could clear this measuring with scope mess up,that would be great O0.
Brad
If you have a 50% duty cycle square wave from a switching circuit ito is OK to use the average value of current multiplied by the DC supply voltage. That is a particular case and the protocol does not apply to other cases. Of course to tell the scope to use the DC voltage you can ask it to smooth out any ripple by demanding the average value be used. I would say just do as pm or itsu said and just multiply the two channels to get the power waveform then ask for the average (mean) value of that power waveform over full cycles. That works for any waveform and any duty cycle.
Smudge
Quote from: TinMan on 2019.01.09, 12:41:32
But CVR1 is from the negative of the cap to ground.
See circuit below.
CVR1 measures input current from source.
CVR2 measures current flowing through the coil.
Brad
Like I have altered below.
How you have it CVR1 is measuring to C1 when you have a low duty cycle, as you increase the duty it will start to measure to the coil as well because the C1 can't supply sufficient.
When it is a short duty the power comes just from the C1, C1 gets charged when the transistor is off and not supplying to the coil.
As I have said, it is a timing (duty cycle) problem. If you move CVR2 to the new position then it will read what is coming from C1 "only".
As you increase the duty cycle, C1 will not be able to supply enough current, the voltage will drop and more will be taken direct from the source to charge the coil.
The new position of CVR2 will be able to measure what goes in and out of C1, and as you change the duty you will see the difference.
Going back to my first post, what was interesting for me was the L1 and magnet (they do not charge C1), they charge C2. C2 can't be charged directly from the source as you have it wired. So all the charge in C2 comes from the coil and magnet setup.
What seems to be happening is that the PM field is being compressed, and then released, which then generates a current greater than what was used to compress the field in the first place by the coil, hence my reference to spring, compress and release. But here the release is two fold, the coil + the added field change of the magnet on the coil.
Now you need to measure all the energy to the load (LED)
Regards
Mike 8)
Thinking further, the only thing C1 does is supply an instant high current at a very low duty cycle,in this case it is a good thing as it is current that creates the field in the coil, but at longer duty cycles it may as well be removed.
Regards
Mike 8)
I believe you're very, very close to finding the answer.
How would reducing the capacitance of C1 affect the
current magnitude distribution? If C1 were to be quite
a lot smaller for example?
It certainly does look like all of the Inductive Kickback
is dedicated to illuminating the LED array.
I made a quick replication of Tinman his magnet/coil setup in order to take some measurements.
Using a 2.7mH / 2.6 Ohm coil i had laying around.
Screenshot shows:
yellow: battery voltage (12.35V, not shown due to scopes limitation of 4 measurement points only).
Purple: CSR1 (10 ohm) signal 15.68mA mean
Red: math function (yellow x purple) 189.2mW input mean.
Blue: CSR2 signal
green: current probe signal in C1 minus lead.
Video also shows input power calculated by using DMM's for voltage and current (196mW).
Video here: https://www.youtube.com/watch?v=xTDG_xs8FNU
Itsu
Thanks for the files Jim.
Can confirm the rotor file is editable, when logged in :)
Quote from: Centraflow on 2019.01.09, 16:38:30
Like I have altered below.
How you have it CVR1 is measuring to C1 when you have a low duty cycle, as you increase the duty it will start to measure to the coil as well because the C1 can't supply sufficient.
When it is a short duty the power comes just from the C1, C1 gets charged when the transistor is off and not supplying to the coil.
As I have said, it is a timing (duty cycle) problem. If you move CVR2 to the new position then it will read what is coming from C1 "only".
As you increase the duty cycle, C1 will not be able to supply enough current, the voltage will drop and more will be taken direct from the source to charge the coil.
The new position of CVR2 will be able to measure what goes in and out of C1, and as you change the duty you will see the difference.
Going back to my first post, what was interesting for me was the L1 and magnet (they do not charge C1), they charge C2. C2 can't be charged directly from the source as you have it wired. So all the charge in C2 comes from the coil and magnet setup.
What seems to be happening is that the PM field is being compressed, and then released, which then generates a current greater than what was used to compress the field in the first place by the coil, hence my reference to spring, compress and release. But here the release is two fold, the coil + the added field change of the magnet on the coil.
Now you need to measure all the energy to the load (LED)
Regards
Mike 8)
Where you have placed CVR2 Mike,means that both CVRs will have current flowing through them during both the off and on time of the transistor
CVR1 will show a DC current flow,and CVR2 will show an alternating current flow.
I want to see the current flow in from the source,and the current flow from the cap and source during the on time seperately.
Brad
Quote from: Itsu on 2019.01.09, 21:24:04
I made a quick replication of Tinman his magnet/coil setup in order to take some measurements.
Using a 2.7mH / 2.6 Ohm coil i had laying around.
Screenshot shows:
yellow: battery voltage (12.35V, not shown due to scopes limitation of 4 measurement points only).
Purple: CSR1 (10 ohm) signal 15.68mA mean
Red: math function (yellow x purple) 189.2mW input mean.
Blue: CSR2 signal
green: current probe signal in C1 minus lead.
Video also shows input power calculated by using DMM's for voltage and current (196mW).
Video here: https://www.youtube.com/watch?v=xTDG_xs8FNU
Itsu
Hi Itsu
Thanks for taking the time to experiment with this.
May i ask if you put the steel sleve inside the coil as well?.
You seem to have the same current flow values across both CVRs ?
That seems to be different to my results.
As you can see in my last video,where i swap out the CVRs for small DC motors with parallel caps,the one in CVR 2 position starts running well before the one in CVR1s position--why would that be?.
In the previous video to that,i show the temperature of each CVR,where CVR2s temperature is some 4 to 5 degrees C hotter--why would that be?
I will carry out the math test using average tonight,and see what happens.
Brad
Quote from: Slider2732 on 2019.01.09, 22:23:31
Thanks for the files Jim.
Can confirm the rotor file is editable, when logged in :)
No worries, Just realised I can't make these individual components editable in a export because they're using tinkercad scripting. duh!
As axles etc will vary depending on what people have access to, I'm thinking we need a better option for those that don't use Tinkercad. Otherwise it will be a pain the bum to resize for what you have at hand.
Quote from: TinMan on 2019.01.09, 23:28:18
As you can see in my last video,where i swap out the CVRs for small DC motors with parallel caps,the one in CVR 2 position starts running well before the one in CVR1s position--why would that be?.
In the previous video to that,i show the temperature of each CVR,where CVR2s temperature is some 4 to 5 degrees C hotter--why would that be?
Hey brad
I see it like this.... CVR1 never sees a lot of current. Mostly due to very little voltage ever across cvr1. Once C1 nearly equals the source, then its source minus C1 equals Vcvr1. Cvr2 on the other hand gets to see higher voltages. Cvr1 never sees anything higher than source minus C1 voltage thus not enough voltage to get the motor turning, even as you increase the on time and cvr2 turns faster, cvr1 and its cap never really sees any spikes, etc, like cvr2. Thinkin what effect a larger or smaller cap on motor 2.
Heck, consider the average currents in cvr1 and how much heat is dissipated, and now imagine the heat generated by the spikes of current in cvr2. cvr1 never gets those inrushes of current and has not chance of cvr heat potentials. Maybe use smaller value resistors. Just the mention of the heat from them alone is too much energy not getting to the coil. :) Especially if those losses could be wasting some of the rainbow. ;)
Mags
So your C1 is a good isolator. Try smaller caps for C1?
One more thing.. Try to measure the voltage and possibly note the brightness of the led and try the circuit without cvrs in the circuit to see the possible loss happening in that hot resistor. So if we have a low loss resistor for a cvr1 and measure the output of the circuit(cap and led), and we can see very insignificant losses, well then thats a nice band in the rainbow. ;)
Mags
Quote from: TinMan on 2019.01.09, 23:28:18
Hi Itsu
Thanks for taking the time to experiment with this.
May i ask if you put the steel sleve inside the coil as well?.
You seem to have the same current flow values across both CVRs ?
That seems to be different to my results.
As you can see in my last video,where i swap out the CVRs for small DC motors with parallel caps,the one in CVR 2 position starts running well before the one in CVR1s position--why would that be?.
In the previous video to that,i show the temperature of each CVR,where CVR2s temperature is some 4 to 5 degrees C hotter--why would that be?
I will carry out the math test using average tonight,and see what happens.
Brad
Brad,
QuoteMay i ask if you put the steel sleve inside the coil as well?.
sorry, i do not have any steel sleeved coil, just this normal coil.
QuoteYou seem to have the same current flow values across both CVRs ?
No, see screenshot below, there is 4x more current (rms) flowing through CSR2 (70mA) compared to csr1 (18mA).
QuoteAs you can see in my last video,where i swap out the CVRs for small DC motors with parallel caps,the one in CVR 2 position starts running well before the one in CVR1s position--why would that be?.
Because there is (like you said) more current flowing through the csr2 position.
QuoteIn the previous video to that,i show the temperature of each CVR,where CVR2s temperature is some 4 to 5 degrees C hotter--why would that be?
Because of the higher current through csr2 (mine has 0.3°C higher temp., but these are 20W csr's).
Itsu
If we calculate the power dissipated by the both csr's we get:
P=U²/R
Pcsr1= 0.185²/10 = 3.4mW
Pcsr2= 0.778²/10 = 60.5mW
Itsu
Quote from: Itsu on 2019.01.10, 09:52:07
Brad,
No, see screenshot below, there is 4x more current (rms) flowing through CSR2 (70mA) compared to csr1 (18mA).
Ok
So if the voltage is a constant,then the power is calculated using that constant voltage value by the current.
So if more current is flowing through CVR2,then more power is being delivered to the coil from C1 than the power supply is delivering to the circuit--correct?.
Brad
Thought experiment.
I have removed the inductor from the circuit
Should we still get the same results,where more current passes through CVR 2 than that of CVR1,or will both CVRs have the same value of current flowing through them ?.
Brad
Quote from: TinMan on 2019.01.10, 12:29:37
Ok
So if the voltage is a constant,then the power is calculated using that constant voltage value by the current.
So if more current is flowing through CVR2,then more power is being delivered to the coil from C1 than the power supply is delivering to the circuit--correct?.
Brad
Yes, at first glance that seem to be correct, however i think it has something to do like Centraflow mentioned
in his post #124, that C1 is fully charged to 12V at the start and being maintained to be fully charged by the
battery while the transistor only at 5% duty cycle pulls power from this C1.
Itsu
Quote from: TinMan on 2019.01.10, 12:59:26
Thought experiment.
I have removed the inductor from the circuit
Should we still get the same results,where more current passes through CVR 2 than that of CVR1,or will both CVRs have the same value of current flowing through them ?.
Brad
No, no more current flows as we have no more 12V on the Collector (i cheated as i tried it).
Itsu
Quote from: Itsu on 2019.01.10, 13:06:33
Yes, at first glance that seem to be correct, however i think it has something to do like Centraflow mentioned
in his post #124, that C1 is fully charged to 12V at the start and being maintained to be fully charged by the
battery while the transistor only at 5% duty cycle pulls power from this C1.
Itsu
But the average current through CVR1 still should be the same as the average current through CVR2,as a capacitor is not an energy source.
When the voltage is a constant,then the same amount of current has to flow into the cap (CVR1)as what flows out of the cap (CVR2).
Here are the test results from having the scope math give us the average power,instead of the RMS power.
First scope shot is P/in from the power supply--CH2 across CVR1.
Second scope shot is CH2 acoss CVR2.
Brad
Quote from: Itsu on 2019.01.10, 13:09:14
No, no more current flows as we have no more 12V on the Collector (i cheated as i tried it).
Itsu
Ah yes,my mistake.
Retry
How about in the below circuit?.
Quote from: TinMan on 2019.01.10, 13:14:36
But the average current through CVR1 still should be the same as the average current through CVR2,as a capacitor is not an energy source.
When the voltage is a constant,then the same amount of current has to flow into the cap (CVR1)as what flows out of the cap (CVR2).
Here are the test results from having the scope math give us the average power,instead of the RMS power.
First scope shot is P/in from the power supply--CH2 across CVR1.
Second scope shot is CH2 acoss CVR2.
Brad
Brad,
i agree with you there, but obviously this cannot be true, so another explaination must be found.
Thanks for the screenshots, it shows you now are running at 24V.
It also shows in screenshot 1 that your input power according to the Math trace is 1.09W average.
This is confirmed by calculating the 24.1V x 46mA = 1.1W.
No comment on screenshot 2 as i am not sure how to interpret that data, as we might be comparing apples with pears.
Itsu
I also took a couple of scope shots of just the current traces,as the voltage is a constant.
Scope shot one across CVR1
Scope shot 2 across CVR2
Both show RMS and average values.
Brad
Quote from: gyula on 2019.01.09, 13:03:41
Hi Brad,
Do you plan to remove the permanent magnet from the coil so that you are left with the iron sleeve as the core? I believe that would be a step forward to find out where the 'extra' the measurements show comes from.
Thanks for sharing.
Gyula
Hi Gyula.
I tried to get the magnets out,but seems i glued them a bit better than i thought,as they will not come out without breaking them or the coil former.
I will make one the same,but leave out the magnets this time,and see how we go.
I did try another coil,and CVR had a very slightly higher current value,but were talking only about 4 mV difference.
So nothing like the results we are getting with the magnet core coil.
Brad.
Quote from: TinMan on 2019.01.10, 13:23:56
Ah yes,my mistake.
Retry
How about in the below circuit?.
Same story there, still no 12V on the collector.
Itsu
Quote from: Itsu on 2019.01.10, 13:33:46
Same story there, still no 12V on the collector.
Itsu
I am not sure what 12v on the collector has to do with it,as when the transistor closes,current will flow through the circuit via the LED(now up the right way),and the capacitor.
But anyway,how about this circuit?
Quote from: Itsu on 2019.01.10, 13:28:35
Brad,
i agree with you there, but obviously this cannot be true, so another explaination must be found.
Itsu
That is what we are told.
It is odd that no matter what i do,whether i use the scope to show that CVR2 has a higher current flowing through it than CVR1,or i show CVR2 having a higher temperature than CVR1,or i use two identical DC motor/cap combo's in place of CVR 1&2,and show that the motor/cap combo in the CVR2 position start spinning up way before that of the motor/cap combo in the CVR1 position-->or even you your self see it in your own test setup,that what we are told/taught should be the case,seems to trump all bench tests that show the opposite.
I do have a theory about why we are seeing what we are seeing,but i have a feeling that it will only fall on deaf ears.
Anyway,thanks for taking the time to run the test O0
Brad.
@Jim
Found the rotor coil .stl on thingiverse but not the rotor. Do not need to edit as I can use any axle size you set. Chance you can upload to there as well?
New printer here next week and spools of #20 and #22 available. Magnets here local to me.
thay
Quote from: TinMan on 2019.01.10, 13:32:35
Hi Gyula.
I tried to get the magnets out,but seems i glued them a bit better than i thought,as they will not come out without breaking them or the coil former.
I will make one the same,but leave out the magnets this time,and see how we go.
I did try another coil,and CVR had a very slightly higher current value,but were talking only about 4 mV difference.
So nothing like the results we are getting with the magnet core coil.
Brad.
Hi Brad,
Thanks for trying the no magnet core in the sleeve coil too. Hopefully you can find a sleeve from the same iron material you use in the first coil, I believe the iron may also have an influence on the results.
Re on your modified (coil removed) drawing shown in Reply #141: Itsu probably meant diode D1 in series with the collector, for it still has a reverse connection and block any collector current as the LED did before you flipped it.
But now that you completely left out both diodes from the collector circuit and show only a resistor (Reply #146), does the circuit still show higher average current across CVR2 versus CVR1? As per conventional theory, the voltage drop across CVR1 should be higher than across CVR2 because the voltage drop across it should increase whenever capacitor C1 is (slightly) recharged from the 12V source to make up for the the transistor current that takes out charge first of all from C1. MAybe I am wrong with this.
Gyula
Quote from: TinMan on 2019.01.10, 13:47:51
I am not sure what 12v on the collector has to do with it,as when the transistor closes,current will flow through the circuit via the LED(now up the right way),and the capacitor.
But anyway,how about this circuit?
Hmmm, like Gyula mentioned, D1 will prevent any current to flow.
Anyway, your new circuit with the load resistor (1K in my case), i got the below screenshot.
Yellow voltage across the battery/C1 (12.5V)
purple current through csr1 = 718uA rms (0.718mA) / 633uA average
Blue voltage current through csr2 = 8.3mA rms / 2.12mA average
The input current DMM shows 0.57mA = 570uA (average).
So also with this circuit we measure more current through csr2 then csr1.
Itsu
Quote from: TinMan on 2019.01.10, 13:14:36
But the average current through CVR1 still should be the same as the average current through CVR2,as a capacitor is not an energy source.
When the voltage is a constant,then the same amount of current has to flow into the cap (CVR1)as what flows out of the cap (CVR2).
Here are the test results from having the scope math give us the average power,instead of the RMS power.
First scope shot is P/in from the power supply--CH2 across CVR1.
Second scope shot is CH2 acoss CVR2.
Brad
Brad,
Your Pin and Pout measurements in these scope shots are correct and would indicate an apparent COP = 1.81/1.09 = 1.66. This does not include any power in the load from the field collapse.
Regards,
Pm
I removed the csr's which enabled me to move the FG ground lead to the battery minus lead.
The current probe now shows a different signal for the input (ex csr1 position).
The ex csr2 position shows still a similar signal as before.
Still the ex csr2 signal (current) is higer then ex csr1 signal.
Green is ex csr1 signal
white is ex csr2 signal
Will restore the full circuit and repeat this csr-less measurement.
Itsu
Original circuit restored without the csr's
leds are much brighter now without csr's, currents also stronger.
Green signal again ex csr1 current
White signal ex csr2 current.
input current DMM shows 50mA (average), input voltage still 12.5V.
So in any case i have tested (with or without csr's, with or without coil, etc.) i find that the csr2 position
always shows more current compared to csr1 position.
I know current alone is not power, but i cannot explain this.
Itsu
wow O0 Thank you for trying such things, Brad! clever, I must say.
A very straightforward circuit now, with the inductor removed... surprised you still get more current through CVR2.
At what frequency does the circuit operate (resonate) with this change?
Quote from: Itsu on 2019.01.10, 16:37:49
Original circuit restored without the csr's
leds are much brighter now without csr's, currents also stronger.
Green signal again ex csr1 current
White signal ex csr2 current.
input current DMM shows 50mA (average), input voltage still 12.5V.
So in any case i have tested (with or without csr's, with or without coil, etc.) i find that the csr2 position
always shows more current compared to csr1 position.
I know current alone is not power, but i cannot explain this.
Itsu
Hi Itsu
With the 5% duty cycle you only need to use one CSR to see power from the battery to the cap and from the cap to the coil. Why? well a cap does not work the same as a battery as we know, a cap will give up whatever we put across it much quicker than a battery.
So lets turn % into seconds :) The cap is being charged for 95 seconds to replace the power taken out of the cap in 5 seconds. So if we measure on the CSR1 1mA = 95mA in 95 seconds, the CSR2 would measure 19mA, 19 X 5=95mA to keep the balance. Naturally this is all without a load, so real tests will vary a bit. This is why CSR2 is a lot more.
As I have said, it is all a question of TIME FRAME (@ 50% duty things will be a lot different, but you would have to increase the frequency or blow your transistor).
Regards
Mike 8)
Mike,
i get what you are saying, and it sound logical, however, the currents shown are averaged over a long timeframe,
so like Brad said, there comes a time that the high average current through csr2 can not anymore being topped
of by the lower averaged current through csr1.
Anyway, to be complete, my setup has:
C1 = C2 = 330uF / 450V electrolytical
L1 = 2.7mH / 2.6 Ohm ferrite cored coil
LED = 4x 3W power leds in series
D1 = UF4007
Frequency = 86Hz (not in resonance).
Source = 12V battery
FG = 3.5V DC square wave @ 5% duty cycle
Itsu
Quote from: Itsu on 2019.01.10, 18:25:09
Mike,
i get what you are saying, and it sound logical, however, the currents shown are averaged over a long timeframe,
so like Brad said, there comes a time that the high average current through csr2 can not anymore being topped
of by the lower averaged current through csr1.
Anyway, to be complete, my setup has:
C1 = C2 = 330uF / 450V electrolytical
L1 = 2.7mH / 2.6 Ohm ferrite cored coil
LED = 4x 3W power leds in series
D1 = UF4007
Frequency = 86Hz (not in resonance).
Source = 12V battery
FG = 3.5V DC square wave @ 5% duty cycle
Itsu
Thanks for the reply Itsu.
What if you replaced C1 with the same size battery as the source, both batteries being exactly equal, what would you see?
The number of samples does not make any difference, the duty cycle is what makes the difference, CSR1 is on all the time "apart from when C1 is full", CSR2 is only on 5% of the time (but dumping a lot of current in that 5%, it is a cap).
Regards
Mike 8)
Quote from: Itsu on 2019.01.10, 16:37:49
Original circuit restored without the csr's
leds are much brighter now without csr's, currents also stronger.
Green signal again ex csr1 current
White signal ex csr2 current.
input current DMM shows 50mA (average), input voltage still 12.5V.
So in any case i have tested (with or without csr's, with or without coil, etc.) i find that the csr2 position
always shows more current compared to csr1 position.
I know current alone is not power, but i cannot explain this.
Itsu
I would say your DMM current measurement is probably the most reliable and accurate at this point. If you display 50 current pulses rather than 9, you might find the scope mean computation of 37mA approaches 50mA. So your Pin is indicating about 625mW. What is your Pout (and which component(s) is your load?).
I would only be worried that CSR1 and CSR2 aren't showing the same current if C1 wasn't present.
Poynt,
indeed, when i increase the number of pulses in the screen, then the mean current approaches 50mA.
I normally do that (more pulses), but that is not the issue here, i know that the csr1 together with
the voltage across the battery is the real input to this circuit (as you mention about 625mW).
The issue is that the current through csr2 is constantly higher (factor 2 by 4) then through csr1.
If we then calculate the power using this 4x higher current together with the same 12.5V battery voltage
we get a much higher power there.
The question is why do we see such a higher current (power?) at csr2 compared to csr1?
Pout is not measured / calculated (yet).
Itsu
I build something in LTspice XVII, is it useable?
Itsu
Quote from: Centraflow on 2019.01.10, 18:44:12
Thanks for the reply Itsu.
What if you replaced C1 with the same size battery as the source, both batteries being exactly equal, what would you see?
The number of samples does not make any difference, the duty cycle is what makes the difference, CSR1 is on all the time "apart from when C1 is full", CSR2 is only on 5% of the time (but dumping a lot of current in that 5%, it is a cap).
Regards
Mike 8)
Mike
Regardless of that,the average current over every full cycle should be the same.
If you look at my circuit,and where i have the ground leads for the scope,you will see that i have CH1 across the cap-not the power supply. We are measuring power delivered to the cap from the supply when ch2 is across CVR1,and then measuring power being delivered to the circuit by the cap when CH2 is across CVR2.
How can the average current through CVR2 be greater than the average current through CVR1,when the voltage is a constant 12 volts--in my case,24 volts.
Average current X peak voltage--we do not average twice to calculate power.
Brad
Thank you Itsu for confirming the scope lines up with the DMM, that is a good thing!
In regards to the CSR currents, and as I already alluded, why is it surprising to anyone that the currents aren't equal?
What is the current in the bottom leg of C1?
Quote from: Itsu on 2019.01.10, 21:23:44
I build something in LTspice XVII, is it useable?
Itsu
Yes. What have you found with it so far?
Hello Tinman, good investigation work so far.
I guess the best way to mesure OU is bypass whats happning in the circuit. Insted. Just mesure input and output.
I have a simple request to mesure the input power supply, and the CVR1. This is the total input.
And the total output. Load. In this case its the LED. So mesure across the LED, and place the CVR2 in line with the LED.
If there is OU this will sum it up. Input vs output.
Please see attached drawing.
If the circuit was making energy. This will sum it up quite quickly.
Blessings! ~Russ
:o 8) :P O0
Then of course we must not overlook Transistor Base Drive Current.
This drive current will appear in the Emitter Circuit of the Transistor.
If the Transistor is being driven into Hard Saturation then the Base
Current could be quite substantial.
Quote from: poynt99 on 2019.01.10, 23:42:38
Thank you Itsu for confirming the scope lines up with the DMM, that is a good thing!
In regards to the CSR currents, and as I already alluded, why is it surprising to anyone that the currents aren't equal?
What is the current in the bottom leg of C1?
Why would the average current through each be different ?,when the voltage is a constant.
So, there is nothing out of the ordinary when the source is delivering 12v @ 36mA to the cap,and the cap is delivering 12v @ 78mA to the circuit,where there is no energy from the circuit being returned to C1 ???
Brad
Quote from: ~Russ on 2019.01.11, 01:19:01
Hello Tinman, good investigation work so far.
I guess the best way to mesure OU is bypass whats happning in the circuit. Insted. Just mesure input and output.
I have a simple request to mesure the input power supply, and the CVR1. This is the total input.
And the total output. Load. In this case its the LED. So mesure across the LED, and place the CVR2 in line with the LED.
If there is OU this will sum it up. Input vs output.
Please see attached drawing.
If the circuit was making energy. This will sum it up quite quickly.
Blessings! ~Russ
:o 8) :P O0
Hey Russ,great to see you here.
Were not to interested in the electrical output ATM.
What wete interested in is the current through the coil,and the voltage across the coil.
This is because the coil is the drive coil of a pulse motor with a difference--read whole thread for explanation.
So,the higher the current through the coil,the stronger the magnetic field.
If we use the CVR1 value for average current of 36mA,the coil will build a magnetic field of a certain strength.
If we use CVR2s average current of 78mA,then the field built by the coil will be much greater.
So that is what we are looking at here,as well as why the 2 CVRs show different values in !average! current values.
Brad
Quote from: muDped on 2019.01.11, 02:09:31
Then of course we must not overlook Transistor Base Drive Current.
This drive current will appear in the Emitter Circuit of the Transistor.
If the Transistor is being driven into Hard Saturation then the Base
Current could be quite substantial.
Covered this in my first video.
VP from the SG is 3v.
Base resistor is 100 0hms.
The tip3055 starts to conduct with a base voltage of about .7v.
So max voltage across the 100 ohm base resistor will be about 2-2.3v. Max current is there for 23mA at a 5%duty cycle.
Average current is therefore 1.15mA from the SG
Brad
Ok, i understand your motivation.
However is there any chance you can take the measurement? Post the screen shot here ?
Im trying to wrap my head around things here and the more data you can post the better we can analyze the system and understand it.
If you don't have time so be it. But it sure would be a good thing to measure.
Thanks. ~Russ
Quote from: ~Russ on 2019.01.11, 06:12:39
Ok, i understand your motivation.
However is there any chance you can take the measurement? Post the screen shot here ?
Im trying to wrap my head around things here and the more data you can post the better we can analyze the system and understand it.
If you don't have time so be it. But it sure would be a good thing to measure.
Thanks. ~Russ
Sorry Russ,i cannot test your circuit,as my scope and SG share a common ground,so CVR2 in your schematic is a !no can do!.
I must also add that the LED is not total power out as described in your schematic.
Total power out is power dissipated by L1 + LED
Power dissipated by L1 is VxI(CVR2)-LED.
We are of course disregarding the small amount of power dissipated by CVR2,transistor,and D1.
Brad
Quote from: poynt99 on 2019.01.10, 23:42:38
Thank you Itsu for confirming the scope lines up with the DMM, that is a good thing!
In regards to the CSR currents, and as I already alluded, why is it surprising to anyone that the currents aren't equal?
What is the current in the bottom leg of C1?
Poynt,
its not surprising to me they are not equal, but its surprising to me the csr2 current is constanly higher.
The current in the bottom leg of C1 is similar as through csr2, see this point in my earlier video:
https://youtu.be/xTDG_xs8FNU?t=280
Concerning the LTspice sim, i just put it together, i was/am not able to have it run, probably some
run parameters are not specified correctly.
Not sure about the FG parameters either (3V square, 86Hz, 5% duty cycle, DC).
Will try lateron today to get it run.
Itsu
Below is the black box scenario,and is what needs explaining.
CVRs 1 & 2 average current as calculated by the scope.
Brad
Quote from: Itsu on 2019.01.11, 09:28:20
Poynt,
its not surprising to me they are not equal, but its surprising to me the csr2 current is constanly higher.
The current in the bottom leg of C1 is similar as through csr2, see this point in my earlier video:
https://youtu.be/xTDG_xs8FNU?t=280
Itsu
Exactly as it should be O0
CSR2 will have a constant current at this 5% duty, it is the CSR2 which is giving the real current from both sources (battery+C1 in parallel).
The scope can't measure the CSR1 current to the C1 because the battery is 100% of the time across C1, CSR1 sees a current when C1 drops in voltage compared to the battery
only.
As I have said T (time) is not taken into consideration when two sources (battery+ cap) do not work the same on discharge and the battery is 100% all the time connected to the cap.
Regards
Mike 8)
PS. The very interesting thing for me is what is happening on the other side of the transistor, the coil and magnet, magnetic compression of a magnetic field, one field to compress and two fields springing back :D
Brad
Look at the top of your CSR1 waveform in relation to CSR2, You can see the wave where the battery is putting power back into the C1, over a longer period of time.
You might think that current flows in two directions at the same time :D but we know it does not.
Regards
Mike 8)
Quote from: Centraflow on 2019.01.11, 12:21:32
Exactly as it should be O0
CSR2 will have a constant current at this 5% duty, it is the CSR2 which is giving the real current from both sources (battery+C1 in parallel).
The scope can't measure the CSR1 current to the C1 because the battery is 100% of the time across C1, CSR1 sees a current when C1 drops in voltage compared to the battery only.
As I have said T (time) is not taken into consideration when two sources (battery+ cap) do not work the same on discharge and the battery is 100% all the time connected to the cap.
Regards
Mike 8)
PS. The very interesting thing for me is what is happening on the other side of the transistor, the coil and magnet, magnetic compression of a magnetic field, one field to compress and two fields springing back :D
Mike,
i think you are right, but i just cannot see it (yet).
I tried your earlier suggestion by putting a similar battery at C1.
Having the both 10 Ohm csr's in place again.
Yellow is Voltage at the input battery
Blue is csr2 current
purple is csr1 current.
First screenshot is without any capacitor or battery in the c1 position, note the both csr currents are the same
of cource, but notice also the dip in the 12V supply.
second screenshot is when i connect a 2th similar battery in the c1 position.
This 2th battery takes over almost all the supply to the circuit.
No more dip in the 12V supply line.
Itsu
Quote from: Centraflow on 2019.01.11, 12:21:32
Exactly as it should be O0
CSR2 will have a constant current at this 5% duty, it is the CSR2 which is giving the real current from both sources (battery+C1 in parallel).
The scope can't measure the CSR1 current to the C1 because the battery is 100% of the time across C1, CSR1 sees a current when C1 drops in voltage compared to the battery only.
As I have said T (time) is not taken into consideration when two sources (battery+ cap) do not work the same on discharge and the battery is 100% all the time connected to the cap.
Regards
Mike 8)
PS. The very interesting thing for me is what is happening on the other side of the transistor, the coil and magnet, magnetic compression of a magnetic field, one field to compress and two fields springing back :D
Actually that makes sense even to me. Thanks Mike.
The circuit itself is quite interesting.
Reminds me of a buck converter but on the negative rail.
Played around with it a bit
D1 may not be needed, just an after thought.
Keep on keepen on looks promising
Quote from: Thaelin on 2019.01.10, 14:39:14
@Jim
Found the rotor coil .stl on thingiverse but not the rotor. Do not need to edit as I can use any axle size you set. Chance you can upload to there as well?
New printer here next week and spools of #20 and #22 available. Magnets here local to me.
thay
It's in an earlier post - sorry battery about to go. Spent the night hacking together a new hotend. Will put on thingiverse tmrw
Quote from: Itsu on 2019.01.11, 09:28:20
Poynt,
its not surprising to me they are not equal, but its surprising to me the csr2 current is constanly higher.
The current in the bottom leg of C1 is similar as through csr2, see this point in my earlier video:
https://youtu.be/xTDG_xs8FNU?t=280
Concerning the LTspice sim, i just put it together, i was/am not able to have it run, probably some
run parameters are not specified correctly.
Not sure about the FG parameters either (3V square, 86Hz, 5% duty cycle, DC).
Will try lateron today to get it run.
Itsu
Istu,
The sim ran for me, although I did not spend too much time with it.
Mike is close if not bang on with his explanation I think, but I encourage you to experiment with LTSpice to see if you can get the same result. If you can, then it should help in explaining the apparent discrepancy in currents.
Poynt,
ok, i got it running i guess and it shows similar results (higher csr2 current), see picture below.
I now need to refine some things like visualize the probe points etc.
thanks for now.
Itsu
Quote from: Centraflow on 2019.01.11, 12:33:56
Brad
Look at the top of your CSR1 waveform in relation to CSR2, You can see the wave where the battery is putting power back into the C1, over a longer period of time.
Regards
Mike 8)
OK Mike,as hard as i try, i do not see what you are saying.
The current has to pass through CVR1 to get to the cap--there is no other way around that CVR1.
All current has to pass through CVR2 to get to the circuit.
The total power being delivered to the cap by the source is less than the cap is delivering to the circuit.
Average current means just that--the average over 1 cycle,and where we put many cycles together to get a very accurate average.
The fact that CVR2 see's a very high current spike over a very short period of each cycle means nothing,as the scope averages that current spike out over the whole cycle--thats what average is.
The fact that the average current flowing through CVR2 is higher than that of CVR1 stands.
QuoteYou might think that current flows in two directions at the same time :D but we know it does not.
No,we dont know that is dose not.
For every action there is an equal and opposite reaction.
Thus,for every electron that moves forward in a wire,then something has to be moving backwards.
There are many things that make no sense,like the magnetic field of a permanent magnet--what is it?.
This garbage about virtual photons is just rubbish.
The permanent magnets field is made of the very thing we are missing here--a charge of some sort.
One end emits a positive charge,and the other a negative charge. So,a positive charge repel's a positive charge,a negative charge repel's a negative charge,but a positive charge is attracted to a negative charge,and both are attracted to something that has a neutral charge--E.G iron O0. When a PM is brought close to the iron,the two charges within the iron start to seperate,and we see most of the negative charges within that iron migrate toward the face of the magnets end that is of a positive charge,thus the iron becomes attracted to the PM.
Here is what we need to find out-->what we need to know--->what is the magnetic field made of?.
Once we know this,we can design solar panel like devices,where instead of a photon knocking an electron through the PN junction,we have the magnetic field knocking !what ever it is we need! through a junction,thus creating power like the solar panel dose.
Anyway,enough rambling from me
Back on subject now.
Brad
Quote from: Itsu on 2019.01.11, 13:03:27
Mike,
i think you are right, but i just cannot see it (yet).
I tried your earlier suggestion by putting a similar battery at C1.
Having the both 10 Ohm csr's in place again.
Yellow is Voltage at the input battery
Blue is csr2 current
purple is csr1 current.
First screenshot is without any capacitor or battery in the c1 position, note the both csr currents are the same
of cource, but notice also the dip in the 12V supply.
second screenshot is when i connect a 2th similar battery in the c1 position.
This 2th battery takes over almost all the supply to the circuit.
No more dip in the 12V supply line.
Itsu
Maybe this will help you to 100% what I see.
Your second shot shows that nearly all the power comes from the second battery, why?, for two reasons, 1. the chemistry is the same for both sources, 2. there is a resistance(CSR1) extra for a path of source to the load. Current will take the path of less resistance in this case (both sources being the same). vola.
Thanks Darren, sorry to all that at my age I do not explain things very well ;D today is my birthday 68yrs :'(
Regards
mike 8)
Quote from: poynt99 on 2019.01.11, 14:15:39
Mike is close if not bang on with his explanation I think,
OK,hang on a minute here.
Quote Mike-- The scope can't measure the CSR1 current to the C1 because the battery is 100% of the time across C1, CSR1 sees a current when C1 drops in voltage compared to the battery only.
Quote Poynt--Mike is close if not bang on with his explanation I think
Quote Poynt--I would say your DMM current measurement is probably the most reliable and accurate at this point. If you display 50 current pulses rather than 9, you might find the scope mean computation of 37mA approaches 50mA. So your Pin is indicating about 625mW.
Ok,so now this is confusing.
Are we trusting both Itsu's and my DMM current reading for P/in,which also agree's with the calculated average current from the scope across CVR1 ?
Or is Mike correct when he states that the scope cannot measure the average current from CVR1 ?
You seem to be agreeing with both here poynt,and each are the exact opposite ???
So either the DMMs cannot give an accurate average current as you have taught us they can,or Mike is incorrect in saying that the scope cannot get an accurate current average across CVR1.
Now,in all my testing,the scopes calculated average current across CVR1 is exactly the same as the DMMs average value of current--all my video's on this DUT show this.
There is simply no way that P/in cannot,and is not measured across CVR1s current and CH1s voltage.
I limited my power supply's current to 37mA @ 12v,and the voltage never dropped. At 40mA draw from the power supply,the power supplies voltage will drop below 12 volts--this i can show with no problem at all by increasing the duty cycle of the FG.
So my scope says 36mA as the current in across CVR1
My DMM says exactly 36mA as the current value also.
My power supply limited to 37mA(as seen in the video's),also says the current is below 37mA
This tells me--no,in fact this is proof that the P/in is 36mA @ 12 volts,and Mike is incorrect in that the scope cannot measure the current in across CVR1.
If there is a problem,then it is with CVR2.
But here is what you are up against.
The scope shows an average current through CVR2 of near double that of CVR1.
CVR2s temperature runs 5 degree's C hotter than that of CVR1.
With my motor/cap combo's in place of the CVRs,the motor in CVR2s position fire's up long before that of the motor in CVR1s position.
Video here-->
https://www.youtube.com/watch?v=vSODFYCqQH8
So,3 confirmation tests for P/in via CVR1s current,and 3 confirmation tests to show CVR2s current value is higher than that of CVR1.
Brad
Quote from: Centraflow on 2019.01.11, 15:26:19
D today is my birthday 68yrs :'(
Regards
mike 8)
Happy birthday Mike.
Many happy returns.
Brad
Quote from: TinMan on 2019.01.11, 15:31:31
Happy birthday Mike.
Many happy returns.
Brad
Thanks Brad O0
Regards
Mike 8)
Quote from: Itsu on 2019.01.11, 13:03:27
Mike,
i think you are right, but i just cannot see it (yet).
I tried your earlier suggestion by putting a similar battery at C1.
Having the both 10 Ohm csr's in place again.
Yellow is Voltage at the input battery
Blue is csr2 current
purple is csr1 current.
First screenshot is without any capacitor or battery in the c1 position, note the both csr currents are the same
of cource, but notice also the dip in the 12V supply.
second screenshot is when i connect a 2th similar battery in the c1 position.
This 2th battery takes over almost all the supply to the circuit.
No more dip in the 12V supply line.
Itsu
Nothing out of the ordinary there Itsu,as i have done this test my self.
If you leave it running long enough,the current from the first battery to the second will increase.
you are better off using a power supply in place of your first battery,as your first batteries voltage will drop with your second batteries voltage.
Brad
Penblwydd hapus Mike. O0
Iechyd da Graham.
Quote from: Grumage on 2019.01.11, 16:48:16
Penblwydd hapus Mike. O0
Iechyd da Graham.
O0
Happy Birthday Mike! O0
Happy Birthday Mike! O0 O0
In five more days I will be 5 years older than you. I have heard that people born in January have very logical minds and are deep thinkers.
I have a technical mind, but not so sure about the deep thinking part. C.C
I actually met you at the first energy conference in Idaho. My son and I spent some time chatting with you after the conference was over for the day. We both are Ham radio operators and chatted about mixing frequencies to separate water. Maybe you remember the small group that met with you in the lobby.
Take care,
Carroll
Quote from: Itsu on 2019.01.11, 14:28:11
Poynt,
ok, i got it running i guess and it shows similar results (higher csr2 current), see picture below.
I now need to refine some things like visualize the probe points etc.
thanks for now.
Itsu
Nice work Itsu! O0
I'm using your new circuit as well. I would run it out to 200ms rather than 100ms.
CTL-Click on the trace lablel to bring up a small window that tells you the AVG and RMS values of the trace. R2 is 8.8mA, and R3 is 9mA AVG values. Interestingly, the AVG current through C1 is only 51uA. The AVG battery current is of course also 8.8mA. :o
So I'm not so sure we are getting the same result (or not to the same degree anyway) you are with your test setup.
Poynt,
great, i have very similar results.
by the way, i made some changes, see attached .asc file in my post #180, like having the FG return between csr1 and csr2, and change the L1 resistance to 2.6 Ohm
I have csr1 current: 8.8mA, csr2 current: 9.08mA, C1 current: -46uA and Battery current also: 8.8mA, all average @ 200ms interval,
see picture (added some 0.01 Ohm resistors for current measurents).
The signals look creepy similar as with the real wordl circuit, see screenshot below, only the values (mean) are different.
Blue is csr2 current (still more then csr1 / battery)
purple is csr1 current
white is C1 current (when i increase the number of pulses the current drops into uA range)
green is battery current (inverted to match the sim).
Could it be that the sim C1 is more perfect then my real life C1? (Sim C1 = 0.09 Ohm Esr and 0 H Esi)
Itsu
Whats the chances of carrying out the measurements on a real device,using the equipment we have to do that very job,instead of using a simulator that dose not simulate the results we are seeing on the real world device--is there any chance of that ?
I know people that will hop in there car,drive 2km to the exercises club,only to jump on a tread mill to go for a walk.
So i ask this only once--
Can we leave the car at home,and take a real walk around the block,where the scenery is completely different to that seen on the tread mill to no where.
Brad
Quote from: JimBoot on 2019.01.11, 13:33:35
It's in an earlier post - sorry battery about to go. Spent the night hacking together a new hotend. Will put on thingiverse tmrw
Here you go @Thaelin. My rotor on thingiverse. I have not printed this yet. https://www.thingiverse.com/thing:3349938
Quote from: TinMan on 2019.01.11, 15:30:08
OK,hang on a minute here.
Quote Mike-- The scope can't measure the CSR1 current to the C1 because the battery is 100% of the time across C1, CSR1 sees a current when C1 drops in voltage compared to the battery only.
Quote Poynt--Mike is close if not bang on with his explanation I think
Quote Poynt--I would say your DMM current measurement is probably the most reliable and accurate at this point. If you display 50 current pulses rather than 9, you might find the scope mean computation of 37mA approaches 50mA. So your Pin is indicating about 625mW.
Ok,so now this is confusing.
Are we trusting both Itsu's and my DMM current reading for P/in,which also agree's with the calculated average current from the scope across CVR1 ?
Or is Mike correct when he states that the scope cannot measure the average current from CVR1 ?
You seem to be agreeing with both here poynt,and each are the exact opposite ???
So either the DMMs cannot give an accurate average current as you have taught us they can,or Mike is incorrect in saying that the scope cannot get an accurate current average across CVR1.
Now,in all my testing,the scopes calculated average current across CVR1 is exactly the same as the DMMs average value of current--all my video's on this DUT show this.
There is simply no way that P/in cannot,and is not measured across CVR1s current and CH1s voltage.
I limited my power supply's current to 37mA @ 12v,and the voltage never dropped. At 40mA draw from the power supply,the power supplies voltage will drop below 12 volts--this i can show with no problem at all by increasing the duty cycle of the FG.
So my scope says 36mA as the current in across CVR1
My DMM says exactly 36mA as the current value also.
My power supply limited to 37mA(as seen in the video's),also says the current is below 37mA
This tells me--no,in fact this is proof that the P/in is 36mA @ 12 volts,and Mike is incorrect in that the scope cannot measure the current in across CVR1.
If there is a problem,then it is with CVR2.
But here is what you are up against.
The scope shows an average current through CVR2 of near double that of CVR1.
CVR2s temperature runs 5 degree's C hotter than that of CVR1.
With my motor/cap combo's in place of the CVRs,the motor in CVR2s position fire's up long before that of the motor in CVR1s position.
Video here-->
https://www.youtube.com/watch?v=vSODFYCqQH8
So,3 confirmation tests for P/in via CVR1s current,and 3 confirmation tests to show CVR2s current value is higher than that of CVR1.
Brad
Hey Brad. How about just using large, equal value caps in place of cvr1 and cvr2? Just run for a short amount of time and measure which cap charges faster. ??? Naturally it will fill up and cause a variation in the circuit as it charges, so this is why I suggest large caps so they dont affect the circuit in the short test time? just put a short across cvr1 cap till c1 is fully charged then remove the short at the beginning of the test. Or short both cvr caps and the test begins when those shorts are removed. Dunno. Just thinking
Mags
Quote from: Magluvin on 2019.01.12, 01:40:27
Hey Brad. How about just using large, equal value caps in place of cvr1 and cvr2? Just run for a short amount of time and measure which cap charges faster. ??? Naturally it will fill up and cause a variation in the circuit as it charges, so this is why I suggest large caps so they dont affect the circuit in the short test time? just put a short across cvr1 cap till c1 is fully charged then remove the short at the beginning of the test. Or short both cvr caps and the test begins when those shorts are removed. Dunno. Just thinking
Mags
Good idea Mags
I will try that
Brad
And another Happy Birthday Mike from over here O0
Jim, Cura reports your rotor would take 7hrs to print.
Am thinking of using wood, because my printer has a bit of a habit of dropping larger prints in the colder weather. Speaking of which, either I bring the scope and components into the bedroom or i'll have to wait a little again...got the opposite problem to the Aussies here.
The 1/2" x 2" pair of magnets have arrived today though :)
Quote from: Slider2732 on 2019.01.12, 02:57:34
And another Happy Birthday Mike from over here O0
Jim, Cura reports your rotor would take 7hrs to print.
Am thinking of using wood, because my printer has a bit of a habit of dropping larger prints in the colder weather. Speaking of which, either I bring the scope and components into the bedroom or i'll have to wait a little again...got the opposite problem to the Aussies here.
The 1/2" x 2" pair of magnets have arrived today though :)
Yeah I know. :) Nearly finished repairs on my printer. Testing now. Alrready printed the smaller version, which took 15 hours as I did it on 0.15mm
Quote from: TinMan on 2019.01.12, 02:30:15
Good idea Mags
I will try that
Brad
Thanks for the offer to try. Was thinking... It may not be fruitfull... Ccrv1 will never see a higher voltage than Ccvr2. That may be a problem. Ccvr1 may punk out once it reaches the voltage difference between source and C1. But thats just thinking that it will all act similar to using the 1ohm resistors, but its not. But, then something interesting happens... C1 begins depleting to the coil, thus more of a voltage offset for Ccvr1 to charge to higher voltage, but I think C1 would keep going down in voltage as that happens causing a mess with measurements. UGGG. Well, I try. ^-^
Anyway... I suppose that using the current value from cvr2 we can...
1 simply calc the power dissipated for the resistor
2 use scope function to get power delivered by the cap
3 use the scope function to get power dissipated in the coil
Once you have all 3, see if the cvr2 + coil add up to the total power were seeing delivered by C1.
Did someone say at some point to just be rid of C1 and cvr1? Maybe. But, it does take away a baseline measurement that should be very trustworthy. Id measure the coil on time zoomed in to just 1 shot. The transistor is only on for such a short time, and a zoom in on 1 pulse should give an accurate power measurement and if we know the time it took for the pulse to happen and figure that with the time interval between on times, you should have a good average watts used between off and next on time. if it all ends the same as what has been tested, then Id say we have to measure the power dissipated by cvr2 and the coil, to see where all that power is going for sure.
Weird stuff
Mags
Mags
Looks like a ring magnet. I wouldn't mind trying a neo tube with an iron core now.
Quote from: Magluvin on 2019.01.12, 04:36:56
Thanks for the offer to try. Was thinking... It may not be fruitfull... Ccrv1 will never see a higher voltage than Ccvr2. That may be a problem. Ccvr1 may punk out once it reaches the voltage difference between source and C1. But thats just thinking that it will all act similar to using the 1ohm resistors, but its not. But, then something interesting happens... C1 begins depleting to the coil, thus more of a voltage offset for Ccvr1 to charge to higher voltage, but I think C1 would keep going down in voltage as that happens causing a mess with measurements. UGGG. Well, I try. ^-^
Anyway... I suppose that using the current value from cvr2 we can...
1 simply calc the power dissipated for the resistor
2 use scope function to get power delivered by the cap
3 use the scope function to get power dissipated in the coil
Once you have all 3, see if the cvr2 + coil add up to the total power were seeing delivered by C1.
Did someone say at some point to just be rid of C1 and cvr1? Maybe. But, it does take away a baseline measurement that should be very trustworthy. Id measure the coil on time zoomed in to just 1 shot. The transistor is only on for such a short time, and a zoom in on 1 pulse should give an accurate power measurement and if we know the time it took for the pulse to happen and figure that with the time interval between on times, you should have a good average watts used between off and next on time. if it all ends the same as what has been tested, then Id say we have to measure the power dissipated by cvr2 and the coil, to see where all that power is going for sure.
Weird stuff
Mags
Mags
Well i did the cap test,and yes,C2 charges to a higher value than C1 by about 1/2 volt.
That may not seem like much,but here is what i have--see circuit below.
C1 and C2 are 5.5v x 1F super caps.
Now two things with this test--
1-we must be careful here,as C2 is getting some what pulsed charged,and caps can be charged more efficiently with pulse charging.
2-For some reason the circuit stops drawing current once C2 gets close to the 2 volt level,and C1 close to 1.6v.
I do not know why,even taking into account the voltage of C3,as when the transistor is on,the series connected C1 and C2 caps have the full 12 volts across them. Even when lifting the power supply to 24 volts,the results are the same.
I understand that C3 may stop charging,once the voltage between C3 and C1 get close,but why dose C1 and C2 not continue to climb in voltage?,due to them having 24 volts across then during the on time of the transistor?.
Anyway,a scope shot below showing C2 being some what pulse charged,while C1 is charged with a steady DC current.
Brad
Quote from: JimBoot on 2019.01.12, 09:12:10
Looks like a ring magnet. I wouldn't mind trying a neo tube with an iron core now.
What are we looking at there Jim?
Quote from: TinMan on 2019.01.12, 10:19:41
What are we looking at there Jim?
a non powered drive coil under the ferro cell. I can clearly see 2 concentric fields. One from the the neo the other from the pipe. Now I understand the reason for the insulator. I can see compression with the dark centre of the neo rather than usually around the edges.
Quote from: JimBoot on 2019.01.12, 10:26:55
a non powered drive coil under the ferro cell. I can clearly see 2 concentric fields. One from the the neo the other from the pipe. Now I understand the reason for the insulator. I can see compression with the dark centre of the neo rather than usually around the edges.
Jim ,that looks great O0
Now,could you send a couple of amps of current through the coil,and then see what the field looks like.
A video would be great,showing what happens when you pulse a current through the coil :)
Then maybe flip the polarity,and pulse again.
I'd love to watch and see what the field dose when you pulse the coil.\
Brad
Quote from: TinMan on 2019.01.12, 11:24:35
Jim ,that looks great O0
Now,could you send a couple of amps of current through the coil,and then see what the field looks like.
A video would be great,showing what happens when you pulse a current through the coil :)
Then maybe flip the polarity,and pulse again.
I'd love to watch and see what the field dose when you pulse the coil.\
Brad
Yep that is the plan. There is a time lag for the field to appear but I would expect to see the iron field change. I didn't weigh my pipe to calculate the iron content but we may be able to tune the density for the right field. Just need to rebuild another circuit. I suspect I'm melting the breadboard. Edit. Or do you literally mean just send a couple of amps of current through it?
Quote from: JimBoot on 2019.01.12, 11:44:14
Yep that is the plan. There is a time lag for the field to appear but I would expect to see the iron field change. I didn't weigh my pipe to calculate the iron content but we may be able to tune the density for the right field. Just need to rebuild another circuit. I suspect I'm melting the breadboard. Edit. Or do you literally mean just send a couple of amps of current through it?
What is the winding resistance of your coil?
Maybe a pic at 1 amp,two amps,3 amps-ETC.
If you have used neo's,then you will need a little more current the neutralise the field.
But i believe that even 1 amp will be enough to pull the PMs field to the pipes opposite field.
Brad
Quote from: TinMan on 2019.01.12, 12:06:17
What is the winding resistance of your coil?
Maybe a pic at 1 amp,two amps,3 amps-ETC.
If you have used neo's,then you will need a little more current the neutralise the field.
But i believe that even 1 amp will be enough to pull the PMs field to the pipes opposite field.
Brad
I'll do pics tonight but vid will have to wait. I'm going to source a Neo tube, I'm thinking iron core, insltr, neo tube, insltr, iron pipe
Quote from: TinMan on 2019.01.12, 12:06:17
What is the winding resistance of your coil?
Maybe a pic at 1 amp,two amps,3 amps-ETC.
If you have used neo's,then you will need a little more current the neutralise the field.
But i believe that even 1 amp will be enough to pull the PMs field to the pipes opposite field.
Brad
Just had a go with the iPhone but no good. There's a change but it's subtle I'll need to setup the tripod with the macro tmrw
Playing with the sim (changing the csr's to 1 Ohm instead of 10 Ohm i was wrongly using) i noticed that the
current through C1 (an alternating current) is almost the difference between the csr1 and csr2 currents.
So i added up via the math function of the scope / sim the csr1 and C1 currents (inverted C1 current) and then i
saw that the resulting math trace was almost identical as the csr2 current trace (see sim picture).
So to me this means that through csr2 not only runs the same current as through csr1, but also the alternating
C1 current and thus the higher average current through csr1 compared to the current through csr1.
Video showing the same on the real circuit here: https://www.youtube.com/watch?v=S1P13eRWqCY&feature=youtu.be
I guess i will leave it by this concerning this csr current difference.
Itsu
Itsu,
Nice work.
Think I'll get this one. The nozzle goes mental apparently
Quote from: Itsu on 2019.01.12, 21:30:28
Playing with the sim (changing the csr's to 1 Ohm instead of 10 Ohm i was wrongly using) i noticed that the
current through C1 (an alternating current) is almost the difference between the csr1 and csr2 currents.
So i added up via the math function of the scope / sim the csr1 and C1 currents (inverted C1 current) and then i
saw that the resulting math trace was almost identical as the csr2 current trace (see sim picture).
So to me this means that through csr2 not only runs the same current as through csr1, but also the alternating
C1 current and thus the higher average current through csr1 compared to the current through csr1.
Video showing the same on the real circuit here: https://www.youtube.com/watch?v=S1P13eRWqCY&feature=youtu.be
I guess i will leave it by this concerning this csr current difference.
Itsu
All you have to do now,is work out where the extra current is coming from,if not through CSR1,as all current entering the system !must! go through CSR1. Now maybe a good time to use electron current flow to get a handle on what is going on.
Brad
earlier someone wanted to see the difference with and without the magnet. If you cant get the magnet out, then reverse the coil polarity. Now we can see the difference of reversing the magnet. If there is change, then there must be a difference without the magnet. ;)
Mags
Quote from: poynt99 on 2019.01.12, 22:15:22
Itsu,
Nice work.
Still dose not explain how CSR 2 has a higher average value of current flowing through it than that of CSR1.
All current must flow through CSR1 to enter the circuit.
During the off time,a % of that average current through CSR1 is sent through C1
During the on time,the remainder of that average current through CSR1 is sent through CSR2,along with the current flow provided by C1-->but C1 must be charged by the current flowing through CSR1 during the off time.
So,how exactly dose CSR2 have a higher average current value flowing through it than that of CSR1 ?.
Brad
Quote from: TinMan on 2019.01.13, 07:38:52
...
So,how exactly dose CSR2 have a higher average current value flowing through it than that of CSR1 ?.
...
What is the collector voltage of Q1?
If it is less than the base voltage+0.6v, then the pulses are rectified by the base-collector junction and the mean DC current goes through R7-L1-R6-R5 and can add up in R2.
[update - Correction of the quote in my answer to TinMan, it was not from him]
You may be onto something there F6FLT. If the transistor itself
is responsible for the apparent anomaly it should be possible
to narrow things down a tad.
If the transistor was to be replaced by a relay which is driven
at the same switching rate would the anomaly persist?
There surely must be an explanation. And, as is usually the case,
it eventually is made manifest by determined and creative effort.
Quote from: CITFTA on 2019.01.11, 19:01:44
Happy Birthday Mike! O0 O0
In five more days I will be 5 years older than you. I have heard that people born in January have very logical minds and are deep thinkers.
I have a technical mind, but not so sure about the deep thinking part. C.C
I actually met you at the first energy conference in Idaho. My son and I spent some time chatting with you after the conference was over for the day. We both are Ham radio operators and chatted about mixing frequencies to separate water. Maybe you remember the small group that met with you in the lobby.
Take care,
Carroll
Hi Carroll
Yes I remember that, long time ago, I have a video of that which someone took, I expect you and your son are on it, remember the security moved us?
Yes capricornians are good for working things out, they like the challenge.
Regards
Mike 8)
Good thinking guys, looks like the missing current is indeed coming from the base of the transistor Q1 (so from the FG).
See sim screenshots below (a very nice tool by the way, worth to dive into!!!).
1st screenshot shows all average currents.
At 0.0ms the transistor activates, at 0.5ms the transistor goes off.
During Q1 activate we see currents flowing through/to csr1 and C1 both making up almost the same average current as through csr2.
After shutdown of Q1 no more current through csr2, but we see current through csr1 charging C1 making up the lost charge during Q1 activate time.
Average current values are:
Csr1 152.6 mA
Csr2 158.71 mA
C1 1.3685 mA
Csr1+C1 153.97 mA
So we still miss 4.47mA between Csr2 and Csr1+C1
2th screenshot shows the average currents through the transistor
We see that the Q1 emitter and Q1 collector currents are the same as resp. Csr2 and Csr1+C1:
Q1 emitter 158.71mA (Csr2)
Q1 collector 153.97mA (Csr1 + C1)
When we now measure the Q1 base current (coming from the FG) we see:
Q1 base 4.7414mA
This 4.4714mA base current is the same as the difference between Csr2 and Csr1+C1 average currents.
So to me this means that all currents are accounted for.
Itsu
Quote from: Centraflow on 2019.01.13, 11:34:42
Hi Carroll
Yes I remember that, long time ago, I have a video of that which someone took, I expect you and your son are on it, remember the security moved us?
Yes capricornians are good for working things out, they like the challenge.
Regards
Mike 8)
Hi Mike,
I had remembered we moved as a group. I had forgotten that security was the reason we moved.
Take care,
Carroll
In the real life circuit, i used my current probe for all 3 transistor Q1 current measurements.
Q1 emitter (same as csr2)
Q1 collector (same as Csr1+C1)
Q1 Base
The difference here between emitter and collector currents (same as csr2 and csr1+C1) also matches the
base current delivered by the FG.
Itsu
Still be interesting to see how it operates as a drive coil.
Should the rotor magnets be the same diameter as the iron sleeve?
I never dreamed the 3d printers were that cheap, gotta have one.
mudPed's idea of a relay has merit if you can get it to switch fast enough and there is not too much contact bounce.
Trading the transistor for an FET would help as the holding current is zero, only switching current would bleed through, but even this can be remedied by returning the SG directly to the FET "source" bypassing the CSR, provided the SG has an isolated ground. In this way SG drive current no longer flows through the CSR. This can also be tried with the transistor.
The circuit can be tried also with isolated photo-voltaic driver and FET so no extra current flows in CSR2.
Photo-voltaic driver can be tied directly to source and gate.
Quote from: Dave45 on 2019.01.13, 16:53:26
Still be interesting to see how it operates as a drive coil.
...
Hi Folks
Yes, I also think that this setup should further be examined when it drives the rotor assembly. See here what Brad measured across a single 2nd coil facing his drive coil:
http://www.overunityresearch.com/index.php?topic=3684.msg70941#msg70941 (http://www.overunityresearch.com/index.php?topic=3684.msg70941#msg70941)
but he first wanted to clarify the higher current via CVR2 versus CVR1 and put the rotor coil steps aside I think.
Itsu, nice job on figuring out and measuring the total current balance, I think the eye opener was the comment by F6FLT.
Gyula
Quote from: ion on 2019.01.13, 17:05:53
mudPed's idea of a relay has merit if you can get it to switch fast enough and there is not too much contact bounce.
Trading the transistor for an FET would help as the holding current is zero, only switching current would bleed through, but even this can be remedied by returning the SG directly to the FET "source" bypassing the CSR, provided the SG has an isolated ground. In this way SG drive current no longer flows through the CSR. This can also be tried with the transistor.
The circuit can be tried also with isolated photo-voltaic driver and FET so no extra current flows in CSR2.
Photo-voltaic driver can be tied directly to source and gate.
ION,
thanks, i tried to tie the FG return to the collector, but the sim won't run, i guess the FG return is tied to ground.
So i inserted a PC817 optocoupler between FG plus and Q1 base (1K inbetween) tied via 1K to 12.5V source and that worked.
There is still the base current difference between collector and emitter, but it does not show up through the csr's
Csr2 and Csr1+C1 average currents are now equal.
Itsu
From Itsu:
Quotethanks, i tried to tie the FG return to the collector, but the sim won't run, i guess the FG return is tied to ground.
[/i]
In the sim, the pulse generator return should go to the emitter, and on the bench the FG ground (at the wall socket) should float unless the FG output does not tie to earth ground. In your case this would probably be dangerous so don't try it, especially with 220 VAC operated equipment.
I'm very used to working with floating equipment and am aware of the possible hazards and work arounds. I only lift grounds when other means are not available, and always with great care (not advised). Back in the day with probing 240 and 480 Volt mains circuits, we would lift the ground on the scope and make sure that no one touched the metal frame of the scope by hanging a big sign on it that said "Hotta Scopini" ...not to be confused with an Italian cousine. ;) An isolation transformer would have been a better way to go, but we didn't have one large enough for the big old tube Tek Scopes such as 500 series.
If the FG were grounded at the wall socket and the scope is also grounded at the wall socket then you must use common grounds for measurement in your circuit or ground loop currents would obfuscate your readings.
A lot of this depends on the equipment used and whether the signals tie back to earth ground at the wall or not. Even when you try to float equipment, there is capacitance coupling and other noise from the internal switching power supply that messes with readings.
That's why it's best to keep all equipment grounded and use other means of isolating signals for measurement.
Hope this helps and does not further confuse.
ION,
its crystal clear, thanks.
Tying the sim FG return to emitter helps :-[ it now also shows no more difference of average current between Csr2 and Csr1+C1.
Doing the same on my real circuit using my battery operated FG seems to also work, but there is a difference between
the current probe value and the same voltage probe value across a csr, so its kind of tricky to measure as it needs compensating.
Anyway, guess we now know where the csr's current difference comes from and how to handle.
Thanks all, Itsu
Quote from: Itsu on 2019.01.13, 21:17:18
ION,
its crystal clear, thanks.
Tying the sim FG return to emitter helps :-[ it now also shows no more difference of average current between Csr2 and Csr1+C1.
Doing the same on my real circuit using my battery operated FG seems to also work, but there is a difference between
the current probe value and the same voltage probe value across a csr, so its kind of tricky to measure as it needs compensating.
Anyway, guess we now know where the csr's current difference comes from and how to handle.
Thanks all, Itsu
Itsu
Why do you keep adding C1s current to CSR1s current?, when it was the current flowing through CSR1 for 95% of the time that filled the cap in the first place.
What you are doing is accounting for CSR1s current twice.
Change everything over to water flowing through pipes,where C1 is a storage vessel,CSR1 AND CSR2 are flow meters that measure LPM. 5 LPM is flowing through meter 1,and for 50 second it fills the storage vessel. Then for 10 seconds,a valve opens and both the water from the storage vessel and that flowing through meter 1 now flows through meter 2.
Are you telling me that more LPM (which represents our current flow)will flow through meter 2 than is flowing through meter 1?--i dont think so.
You would have to take the average of current flowing through CSR1 & C1,and add them together if you wish to do it that way.
The average current flowing through C1 will be a negative value,as there are losses in charging and discharging a cap--as you well know.
Brad
Hey Brad
It does seem very logical that the fg/transistor can affect the circuit. Back in the days of Rose Ainsley circuit, the fg did cause misreadings in the circuit. Like I said earlier, if one were to measure all the components after C1 we should be able to target the issue. If we are to think that the magnet/coil are doing something special, then we should remove the magnet and test and like I said a few posts ago, reverse the magnet by reversing the coil polarity to see the differences. The fg and the transistor seems to be an obvious culprit here.
Mags
I liked Brad's video with the two DC motors: https://www.youtube.com/watch?v=vSODFYCqQH8.
Are some of you saying that AC from the FG is sufficient to run the "output" DC motor (while the "input" DC motor remains stopped)?
That is, how do you explain Brad's observations in this video?
Quote from: PhysicsProf on 2019.01.14, 01:10:17
I liked Brad's video with the two DC motors: https://www.youtube.com/watch?v=vSODFYCqQH8.
Are some of you saying that AC from the FG is sufficient to run the "output" DC motor (while the "input" DC motor remains stopped)?
That is, how do you explain Brad's observations in this video?
I still believe that motor on in the source side of C1 never gets enough voltage across it to get it moving, but the second motor does get higher voltage via the on time of the circuit.
Mags
Quote from: Magluvin on 2019.01.14, 00:43:46
Hey Brad
It does seem very logical that the fg/transistor can affect the circuit. Back in the days of Rose Ainsley circuit, the fg did cause misreadings in the circuit. Like I said earlier, if one were to measure all the components after C1 we should be able to target the issue. If we are to think that the magnet/coil are doing something special, then we should remove the magnet and test and like I said a few posts ago, reverse the magnet by reversing the coil polarity to see the differences. The fg and the transistor seems to be an obvious culprit here.
Mags
I did calculate any possible current the FG could add to the circuit--it was in my very first video. So this is something i took into account first up. I also use a diode on the iput to the base,so no reverse current can enter the system via the emmiter/common ground via the FG.
Regarding reversing the coil--
Well that dose make a difference to the flyback output--it is much lower. It also increases the I/in a little-but not much.
The drive coil setup is not for generating extra electrical energy across that coil,but to increase the torque on the rotor,and a higher induced magnetic field through the rotor coils.
But it would seem that (to me anyway) the drive coil dose indeed increase the flyback output.
Brad
Quote from: Magluvin on 2019.01.14, 01:18:27
I still believe that motor on in the source side of C1 never gets enough voltage across it to get it moving, but the second motor does get higher voltage via the on time of the circuit.
Mags
And to increase the voltage across a resistance means what ?
Brad
Quote from: PhysicsProf on 2019.01.14, 01:10:17
I liked Brad's video with the two DC motors: https://www.youtube.com/watch?v=vSODFYCqQH8.
Are some of you saying that AC from the FG is sufficient to run the "output" DC motor (while the "input" DC motor remains stopped)?
That is, how do you explain Brad's observations in this video?
The only way to increase the voltage across a resistance is to increase the current flow.
And no,the FGs average current into the system is about 1.3mA
Brad
Quote from: TinMan on 2019.01.14, 01:27:53
The only way to increase the voltage across a resistance is to increase the current flow.
And no,the FGs average current into the system is about 1.3mA
Brad
V=I*R is something I do remember from my audio engineering days but I thought it was more complicated than that.
Quote from: JimBoot on 2019.01.14, 04:36:29
V=I*R is something I do remember from my audio engineering days but I thought it was more complicated than that.
Yes
Simple ohms law.
To increase a voltage across a resistance,the current flowing through that resistance must increase.
There is no argument that a higher average current flows through CSR2 than that of CSR1. The problem is--no one has explained why yet.
Adding the current flowing out of C1 to the average current flow through CSR1 is wrong. You have to subtract the current that was flowing into C1 from CSR1s average,as it is the current flowing through CSR1 that charges the cap during the 95% off time of the transistor.
Why few dont seem to understand this ?, well i just dont know.
Quote from: JimBoot on 2019.01.12, 12:16:20
I'll do pics tonight but vid will have to wait. I'm going to source a Neo tube, I'm thinking iron core, insltr, neo tube, insltr, iron pipe
Ok I think I have a measure I can use to show what looks like a flipping of fields through the ferrocell when current is applied. I need to tripod up at night time to get the most accurate pics. You can clearly see a 180 degree change when you look from the right angle. Will post pics later tonight.
Quote from: JimBoot on 2019.01.14, 06:04:46
Ok I think I have a measure I can use to show what looks like a flipping of fields through the ferrocell when current is applied. I need to tripod up at night time to get the most accurate pics. You can clearly see a 180 degree change when you look from the right angle. Will post pics later tonight.
Sounds good Jim.
This shows the coil doing exactly what we want it to do.
I do have another coil design for solid state opperation if anyone is interested.
Brad
Quote from: TinMan on 2019.01.14, 06:23:51
Sounds good Jim.
This shows the coil doing exactly what we want it to do.
I do have another coil design for solid state opperation if anyone is interested.
Brad
Yes please
Me too. O0
Mags
Quote from: Magluvin on 2019.01.14, 07:47:40
Me too. O0
Mags
I will draw it up tonight and post it here.
Quote from: TinMan on 2019.01.14, 00:24:16
Itsu
Why do you keep adding C1s current to CSR1s current?, when it was the current flowing through CSR1 for 95% of the time that filled the cap in the first place.
What you are doing is accounting for CSR1s current twice.
Change everything over to water flowing through pipes,where C1 is a storage vessel,CSR1 AND CSR2 are flow meters that measure LPM. 5 LPM is flowing through meter 1,and for 50 second it fills the storage vessel. Then for 10 seconds,a valve opens and both the water from the storage vessel and that flowing through meter 1 now flows through meter 2.
Are you telling me that more LPM (which represents our current flow)will flow through meter 2 than is flowing through meter 1?--i dont think so.
You would have to take the average of current flowing through CSR1 & C1,and add them together if you wish to do it that way.
The average current flowing through C1 will be a negative value,as there are losses in charging and discharging a cap--as you well know.
Brad
Brad,
does the fact alone that all currents in the circuit add up perfectly both in the sim and the real circuit
say something to you?
Does the fact that both the sim and the real circuit are giving the exact same signals not tell you that this
circuit seem to follow the well established rules of electronic circuits?
Anyway, please look at my post #209 and picture.
Remember i reversed the C1 current there for more clarity, but it flows the opposite as shown.
And yes, the first time we connect the battery to the circuit / C1 there initially flows current from the
battery through csr1 into C1 to charge it. This is the initial status.
Then the follow repeats:
# we have Q1 off, C1 is charged, no currents flow.
# When Q1 turns on, we see C1 discharge via csr2, transistor etc. and
we see current through csr1 from the battery also through csr2, transistor etc. (thus we can add up those currents)
# when Q1 shuts off, current through csr2 stops, but current through csr1 still flows into C1 to recharge it.
# etc.
This is how i see the currents through the csr's run, so yes we need to add up the currents through/to csr1/C1.
There still is a missing part after adding up, but this was explained in posts #218 and #220 by showing that
that missing part is exactly the same as the FG current.
Itsu
Quote from: TinMan on 2019.01.14, 06:23:51
Sounds good Jim.
This shows the coil doing exactly what we want it to do.
I do have another coil design for solid state opperation if anyone is interested.
Brad
I also am interested in a coil for solid state operation. Please post it when you have time.
Thanks,
Carroll
Quote from: Itsu on 2019.01.14, 10:45:33
Brad,
Itsu
Quotewe have Q1 off, C1 is charged, no currents flow.
Yes,current dose flow through CSR1,and then through C1 CSR.
QuoteWhen Q1 turns on, we see C1 discharge via csr2, transistor etc. and
we see current through csr1 from the battery also through csr2, transistor etc. (thus we can add up those currents)
You can only add up the current from CSR1 and CSR-C2 that are within the 5% on time of the transistor.
You cannot add the average of CSR1 to the output current average of CSR-c1 to arrive at the average current flow in from the source.
QuoteDoes the fact that both the sim and the real circuit are giving the exact same signals not tell you that this
circuit seem to follow the well established rules of electronic circuits?
There is nothing wrong with the wave form's--they are as they should be.
Nor is there anything wrong with the well (theoretical) established rules of electronics.
The problem here is human error--nothing more.
QuoteThere still is a missing part after adding up, but this was explained in posts #218 and #220 by showing that that missing part is exactly the same as the FG current.
Not on my bench.
Please see scope shot below,where i have taken a differential voltage measurement across the 100 ohm base resistor.
You can clearly see that the V/drop across that resistor is !at best! 500mV.
500mV across 100 ohms give us 5mA of current. And that 5mA is only for 5% of a full cycle.
What would be the average current flowing into the system from the FG ?
Peak current through CSR2 during the one time in my tests(as i have shown many times now),is 1.1 amps.
So if you wish to subtract that 5mA from that value,then we have 1.095 amps.
Quotewhen Q1 shuts off, current through csr2 stops, but current through csr1 still flows into C1 to recharge it.
That is correct,and is the very reason you cannot add the average current through CSR1 to the current being delivered by the cap during the on time to calculate your current in,as during the off time,the current through CSR 1 is being stored in the cap.
You are accounting for some of the current flowing through CSR1 twice--and you cant do that.
Please see scope shot below,where i have marked out the portion of current through CSR1 you can add to the output current of C1,and the portion of CSR1 current you cannot add to the output current of CSR-C1.
You can add the % of current between the red lines to the output current of CSR-C1
You cannot add the portion of current between the green lines to the output current of CSR-C1,as that portion is going into C1,and not through the coil.
Hope that is clear enough.
If not,then use the water flow through pipes ,with the added vessel(C1),and see if you can make more water flow out of meter 2 than is flowing in through meter 1.
Quotedoes the fact alone that all currents in the circuit add up perfectly both in the sim and the real circuit
say something to you?
Yes
It means that people will find what ever they need,so as to conform to the theories of today.
Brad
Ok, Brad,
you stick with your explanation and i will stick with mine.
I have shown in both the sim and the real circuit how i think the currents flow and that they all are accounted for,
so i am happy with that.
Lets continue the build.
Itsu
Quote from: Itsu on 2019.01.14, 12:22:30
Ok, Brad,
you stick with your explanation and i will stick with mine.
I have shown in both the sim and the real circuit how i think the currents flow and that they all are accounted for,
so i am happy with that.
Lets continue the build.
Itsu
Itsu
What is the point in not clarifying this?
Please explain as to how you can add the average current flow through CSR1 to the current flow out of C1 to get the average current flow through CSR2 ?
P/in is V x I,regardless of what the circuit is--always has been,always will be-->this is the known accepted electrical laws you were quoting a couple of posts back.
Now your saying that P/in is no longer V x I,and that P/in is taken from components within the circuit.
I urge you to rethink this.
I also do not know why Poynt has not said something about this,as it was he that has said,and tested,that DMMs do a very good job at averaging out voltage and current-->has that now flown out the window?.
I am now going to draw that water flow through pipes circuit,and lets you take a look at it.
Brad
Itsu
Here is our circuit using water flow through pipe's.
If you can tell me how more water !on average! will flow through flow meter 2 than that of flow meter 1,i would love to hear how this is possible.
Brad
Quote from: CITFTA on 2019.01.14, 11:44:04
I also am interested in a coil for solid state operation. Please post it when you have time.
Thanks,
Carroll
Ok guys,here it is.
This is my design,and has not been tested yet--still to wind L1
Dimensions are up to you.
My magnet is a neo N52 65mm long x 19mm diameter
L1 is also this long,and will be 500 turns of .77mm wire
L2 and L3 are 20mm wide each,for a total coil length of 115mm,as each end coil former is 5mm thick.
L2 and L3 are also 500 turns,using .4mm wire.
Make sure that the pvc sleeve and steel pipe continue past the magnet as far as your L2 and L3 coils do--> so L2 and L3 are not atop the magnet.
Operation.
L1 is pulsed with a DC current that creates a magnetic field opposite to that of the PMs field--both current value and duty cycle determined through testing.
L2 and L3 are induced by L1.
Flyback is caught and used from all three coils.
The flyback will be large due to the complete reversal of the magnetic field through all 3 coils.
Brad
Quote from: TinMan on 2019.01.14, 12:49:35
Itsu
What is the point in not clarifying this?
Please explain as to how you can add the average current flow through CSR1 to the current flow out of C1 to get the average current flow through CSR2 ?
P/in is V x I,regardless of what the circuit is--always has been,always will be-->this is the known accepted electrical laws you were quoting a couple of posts back.
Now your saying that P/in is no longer V x I,and that P/in is taken from components within the circuit.
I urge you to rethink this.
I also do not know why Poynt has not said something about this,as it was he that has said,and tested,that DMMs do a very good job at averaging out voltage and current-->has that now flown out the window?.
I am now going to draw that water flow through pipes circuit,and lets you take a look at it.
Brad
Brad,
I have already explained how i think the average currents are flowing, see my earlier posts.
We are looking at the current flows here, not sure why "P/in is V x I" is pulled into it as i said nothing about that.
My DMM shows indeed the same average current as i see through csr1, so nothing goes out of the window.
Concerning your water flow circuit,
With the valve closed, the water flows through csr1 untill C1 is full, this is the start position.
start:
# When the valve is opened, water starts to flow through csr1 and water starts to flow from C1, both passing
through csr2 (so we can add up the flows from C1 and csr1).
# When the valve closes, flow through csr2 stops, but the flow through csr1 still continuous to fill up C1,
then the flow stops (start position).
The water flown through csr2 when the valve was open is the same as the water that was stored in C1 together
with the flow through csr1.
Itsu
Quote from: Itsu on 2019.01.14, 14:44:17
Brad,
I have already explained how i think the average currents are flowing, see my earlier posts.
Concerning your water flow circuit,
With the valve closed, the water flows through csr1 untill C1 is full, this is the start position.
start:
# When the valve is opened, water starts to flow through csr1 and water starts to flow from C1, both passing
through csr2 (so we can add up the flows from C1 and csr1).
# When the valve closes, flow through csr2 stops, but the flow through csr1 still continuous to fill up C1,
then the flow stops (start position).
The water flown through csr2 when the valve was open is the same as the water that was stored in C1 together
with the flow through csr1.
Itsu
Regarding the water flow circuit,the question was very simple.
Flow meter 1(CSR1) has an average water flow of 5 LPM
Through continuous operation,what is the average water flow through flow meter 2 (CSR2) ?
We are not talking about peak flow through flow meter 2 here,we are talking average flow over time.
QuoteWe are looking at the current flows here, not sure why "P/in is V x I" is pulled into it as i said nothing about that.
My DMM shows indeed the same average current as i see through csr1, so nothing goes out of the window.
Because the voltage is a constant.
So,if we are using 12 volts,then P/in is 12v x (in my case) 36mA--the 1 ohm resistor has 36mV across it,and the DMM also says 36mA
P/in is 432mW
Now, we have !on average! (in my case) 72mA through CSR2,which is also at 12 volt's.
This means that the !apparent! output is 864mW.
But anyway,if you could just answer the water circuit question,then we can take it from there.
And to clarify,lets say we have a water pressure of 12 psi(our voltage),and of course our flow rate(amps) through flow sensor 1 of 5 LPM.
If anyone else would like to answer the question,please feel free.
Brad
QuoteRegarding the water flow circuit,the question was very simple.
Flow meter 1(CSR1) has an average water flow of 5 LPM
Through continuous operation,what is the average water flow through flow meter 2 (CSR2) ?
We are not talking about peak flow through flow meter 2 here,we are talking average flow over time.
I told you, its the same, so that would be 5 LPM
Itsu
Quote from: TinMan on 2019.01.14, 06:03:36
There is no argument that a higher average current flows through CSR2 than that of CSR1.
How/where has this been determined?
Quote from: TinMan on 2019.01.08, 05:04:37
So i had an idea in regards to this power measurement issue.
My claim is that there is more power flowing through the coil during each pulse than there is being delivered by the source-that being my power supply.
So far i have shown this by way of scope measurements,and heat dissipated by the CVRs,where CVR1 dissipates less heat than CVR2,meaning more current flowing through CVR2,and thus the coil.
Brad
Let's try to clear some things up regarding power measurements and current readings.
First, average current is used to find Pin when the source is DC. So with a 1R CVR/CSR we find Pin by: Iavg x VDC. Iavg can be found with a DMM across the CVR.
One can determine Iavg at other points in the circuit, but it is of little interest in terms of power measurements.
Second, if one is wanting to determine the average power dissipated in a pure resistance (such as a CVR/CSR), then we use the
RMS current through that resistor (not the average current).
The fact that CVR2 dissipates double the average power compared to CVR1 is interesting, but is otherwise quite unremarkable. It certainly does not indicate that the circuitry after the battery (not including the FG) is somehow generating extra power above and beyond what the battery is supplying.
Third, average power is what matters. Peak powers can exceed average input power, but at the end of the day it comes down to accounting for the average power in all components that can supply and dissipate power.
In summary:
Average power from a DC supply (Pin): Iavg x VDC
Average power in a pure resistance: Irms
2 x R, OR Vrms
2 / R
Average power in any device: AVG[v(t) x i(t)]
Quote from: poynt99 on 2019.01.14, 18:30:13
How/where has this been determined?
Really?
How about the last 4 to 5 pages of this thread O0
I have shown this in all my video's,as has Itsu.
I have also posted many screenshots showing this-EG below.
Brad
Quote from: poynt99 on 2019.01.14, 19:47:05
Let's try to clear some things up regarding power measurements and current readings.
One can determine Iavg at other points in the circuit, but it is of little interest in terms of power measurements.
QuoteFirst, average current is used to find Pin when the source is DC. So with a 1R CVR/CSR we find Pin by: Iavg x VDC. Iavg can be found with a DMM across the CVR.
Agreed O0
QuoteSecond, if one is wanting to determine the average power dissipated in a pure resistance (such as a CVR/CSR), then we use the RMS current through that resistor (not the average current).
Once again-agreed O0
QuoteThe fact that CVR2 dissipates double the average power compared to CVR1 is interesting, but is otherwise quite unremarkable.
Using the circuit below,where the input current from the FG is a meere 5mA for 5% of a complete cycle,can you explain as to why the !average! current flowing through CSR2 is near double that of what flows through CSR1,when the voltage is a constant .
QuoteThird, average power is what matters. Peak powers can exceed average input power, but at the end of the day it comes down to accounting for the average power in all components that can supply and dissipate power.
Agreed once again.
So once again,looking at the circuit below,from which CVR would you use the average current value from to determine the average power being sent to the coil,where the voltage is a constant ?.
QuoteIn summary:
Average power from a DC supply (Pin): Iavg x VDC
Average power in a pure resistance: Irms2 x R, OR Vrms2 / R
Average power in any device: AVG[v(t) x i(t)]
Agreed O0
Brad
I don't know what you're showing in those screen shots.
Can you describe what the signals and measurements are and how it is shown that Icsr2(avg)>Icsr1(avg)?
Thanks for posting the SS model Brad. So this one sorta misses out on rotor gen action. Does it matter much, or is the flyback in both setups the main goal? Just wondering.
Now on the rotor model, the coil is over the end of the magnet. Is there a reason why its not in the SS setup? Or does it just work better that way in each case?
Been thinking on all of this a bit. At first was thinking, hmm. Just using energy of the coil to move, to say, the magnets field and would there be any advantage to that. But then thought, are we getting a combined coil and magnet field that ends up greater than what the coil produced during on time. Anyway, nice idea.
Ive seen some ferrite cores that are thin tubes. There is a lot of ferrite shapes and sizes and was looking for them for the SS orbo, orbonbon, and wondered if that may be useful here in later developments as all goes well.
Mags
Quote from: poynt99 on 2019.01.14, 23:32:10
I don't know what you're showing in those screen shots.
Can you describe what the signals and measurements are and how it is shown that Icsr2(avg)>Icsr1(avg)?
Yes,below are the circuit and scope probe placement for each.
This was when i was running the circuit at 12VDC
Quote from: TinMan on 2019.01.14, 23:31:59
Using the circuit below,where the input current from the FG is a meere 5mA for 5% of a complete cycle,can you explain as to why the !average! current flowing through CSR2 is near double that of what flows through CSR1,when the voltage is a constant .
What do you mean by "voltage is constant"?
The coil stores some energy, then it releases the energy.
Quote
So once again,looking at the circuit below,from which CVR would you use the average current value from to determine the average power being sent to the coil,where the voltage is a constant ?.
You've already determined the "power being sent to the coil" (to the whole circuit really) when you calculated Pin, and you used CVR1.
Here is what I would be asking myself and subsequently determining if I suspected the circuit was somehow running overunity:
"Does the sum of the average power dissipation of all the circuit elements (excluding the battery) equal the average power the battery is putting out?"
If you wish to know the Volt-Amps (apparent power) in the coil itself, then you need a CVR in series with the coil (not in the emitter), and perform a AVG[v(t) x i(t)] measurement with the scope.
Quote from: poynt99 on 2019.01.15, 00:03:43
What do you mean by "voltage is constant"?
The coil stores some energy, then it releases the energy.
You've already determined the "power being sent to the coil" (to the whole circuit really) when you calculated Pin, and you used CVR1.
Here is what I would be asking myself and subsequently determining if I suspected the circuit was somehow running overunity:
"Does the sum of the average power dissipation of all the circuit elements (excluding the battery) equal the average power the battery is putting out?"
If you wish to know the Volt-Amps (apparent power) in the coil itself, then you need a CVR in series with the coil (not in the emitter), and perform a AVG[v(t) x i(t)] measurement with the scope.
Would not the power being sent to the coil be the voltage across C1 x the average current through CSR2 ?,as CSR2 shows both the current delivered by C2,and also the power supply.
The voltage is a constant 12.1 volts,both across the cap,and across the coil when the transistor is on.
I have supplied a scope shot of the differential voltage across the coil,where we see the voltage across the cap minus V/drop across CSR2.
Brad
The Volt-Amps (apparent power) IN THE COIL is:
AVG[v(t)(across the coil) x i(t)(through the coil)]
Average current is used for Pin, and only when the source is DC. Other than that, we don't use average current.
CSR2 only tells us the current in the emitter of Q1. The emitter current is made up from two other currents; the base current and the collector current. The collector current is made from two other currents; the coil current and the diode current.
So one should realize that CSR2 is far removed from providing an accurate reading of the coil current alone.
The above assumes that the circuit is per Itsu's, where the circuit common is between the two csr's.
Quote from: poynt99 on 2019.01.15, 00:03:43
If you wish to know the Volt-Amps (apparent power) in the coil itself, then you need a CVR in series with the coil (not in the emitter), and perform a AVG[v(t) x i(t)] measurement with the scope.
Not that i see any difference in having the CSR on the positive side of the coil to that having it on the emmiter,i will give that a try tonight.
Thanks for your time on this Poynt
Brad
Quote from: TinMan on 2019.01.14, 13:40:02
Ok guys,here it is.
This is my design,and has not been tested yet--still to wind L1
Dimensions are up to you.
My magnet is a neo N52 65mm long x 19mm diameter
L1 is also this long,and will be 500 turns of .77mm wire
L2 and L3 are 20mm wide each,for a total coil length of 115mm,as each end coil former is 5mm thick.
L2 and L3 are also 500 turns,using .4mm wire.
Make sure that the pvc sleeve and steel pipe continue past the magnet as far as your L2 and L3 coils do--> so L2 and L3 are not atop the magnet.
Operation.
L1 is pulsed with a DC current that creates a magnetic field opposite to that of the PMs field--both current value and duty cycle determined through testing.
L2 and L3 are induced by L1.
Flyback is caught and used from all three coils.
The flyback will be large due to the complete reversal of the magnetic field through all 3 coils.
Brad
Well that's a decent bloody neo and a big arse coil to match. Pretty easy to print something up. ONce the new printer arrives I want to see what effect a greater mass of iron will have in the drive coil as well.
Quote from: Magluvin on 2019.01.14, 23:37:26
Thanks for posting the SS model Brad. So this one sorta misses out on rotor gen action. Does it matter much, or is the flyback in both setups the main goal? Just wondering.
Now on the rotor model, the coil is over the end of the magnet. Is there a reason why its not in the SS setup? Or does it just work better that way in each case?
Been thinking on all of this a bit. At first was thinking, hmm. Just using energy of the coil to move, to say, the magnets field and would there be any advantage to that. But then thought, are we getting a combined coil and magnet field that ends up greater than what the coil produced during on time. Anyway, nice idea.
Ive seen some ferrite cores that are thin tubes. There is a lot of ferrite shapes and sizes and was looking for them for the SS orbo, orbonbon, and wondered if that may be useful here in later developments as all goes well.
Mags
I left the magnet short so as we can slide in a shorter magnet in each end to see what difference it makes during testing.
I am hoping that this design will totally remove all magnetic fields at L2 and L3 during the pulse input.
Brad
Quote from: TinMan on 2019.01.15, 04:27:34
Not that i see any difference in having the CSR on the positive side of the coil to that having it on the emmiter,i will give that a try tonight.
Thanks for your time on this Poynt
Brad
You can go through the exercise, but I don't think you will see much. In fact, I don't think the coil is doing very much in the current circuit.
The vast majority of power transfer is occurring between the battery and CSR2 (CSR2 at 10 Ohms is the largest consumer of power (80 to 90%) in your circuit).
If you are trying to get maximum power to the LED's then this circuit isn't going to do it in the current configuration.
Without any tuning or tweaking, here are a couple of simple things to maximize power transfer to the LED's:
1) remove both CSR's. (with csr1 removed, you probably no longer need C1)
2) increase C2 to about 2200u.
Quote from: poynt99 on 2019.01.15, 15:00:10
You can go through the exercise, but I don't think you will see much. In fact, I don't think the coil is doing very much in the current circuit.
The vast majority of power transfer is occurring between the battery and CSR2 (CSR2 at 10 Ohms is the largest consumer of power (80 to 90%) in your circuit).
If you are trying to get maximum power to the LED's then this circuit isn't going to do it in the current configuration.
The CSRs in my circuit are 1 ohm.
The LEDs are only to burn off the inductive kickback energy.
The main goal here is to send enough current through the coil to neutralise the magnetic field of the PM core.
This anomaly with more current flowing through CSR2 than that of CSR1 was only found during coil testing,and i thought it interesting enough to post--and it went on from there.
Brad
Quote from: TinMan on 2019.01.15, 16:23:07
The CSRs in my circuit are 1 ohm.
Brad
Could you double-check that with a meter perhaps? I don't think you should be getting a saw-tooth wave form on CSR1 with a 1R value.
Quote from: poynt99 on 2019.01.15, 22:11:05
Could you double-check that with a meter perhaps? I don't think you should be getting a saw-tooth wave form on CSR1 with a 1R value.
I always check there value before using them in a circuit O0
As C1 is 10,000uF in my circuit,and frequency is 50Hz,and a duty cycle of 5%,then i think a sawtooth wave form is quite normal.
I have swapped the resistors positions around,(which is now a habit i have),and the results are the same.
Brad
OK, with 10,000u and 1R, yes is the same time constant as 10R with 1000u. (You might want to include all component values on your schematics).
Still doesn't quite compute with the current peaks. With 1R on the emitter, you should have more than 1.2A, unless your coil has a high resistance.
btw, I'd be curious to know if a DMM measures the same average current in the CSRs as indicated by your scope.
Quote from: poynt99 on 2019.01.15, 23:28:27
OK, with 10,000u and 1R, yes is the same time constant as 10R with 1000u. (You might want to include all component values on your schematics).
Still doesn't quite compute with the current peaks. With 1R on the emitter, you should have more than 1.2A, unless your coil has a high resistance.
btw, I'd be curious to know if a DMM measures the same average current in the CSRs as indicated by your scope.
Yes,sorry the values were not on that schematic,as that was an early one. The schematic i posted a few pages back dose have all the values.
I know your time is shot,and you cant watch everything that is going on,but in my video's i do also have the DMMs reading the current,and they give the exact same value as the scope is giving in average.
Peak current through CSR2 is between 1.1 and 1.2 amps --during the 5% on time.
Brad
Good news guys--my 3D printer arrived today O0
Bad news is-the power is out,due to upgrades going on in our area.
I have taken an un-boxing video,and will upload as soon as power is back on.
So now the fun begins--time to put it together O0
Brad
Quote from: TinMan on 2019.01.16, 05:02:38
Good news guys--my 3D printer arrived today O0
Bad news is-the power is out,due to upgrades going on in our area.
I have taken an un-boxing video,and will upload as soon as power is back on.
So now the fun begins--time to put it together O0
Brad
Fun and games. Mine arrives tomorrow. Slic3r seems to be the thing they recommend over Cura for Anet. One for mac is buggy as hell. Looks like it gives you more control though.
wrapped 400 turns of .51mm and tried diff cores as a pickup for the drive coil.
An air core, a steel bolt and 18 pieces of 1.75mm iron pla. Edit: but when I put the drive coil near the neo rotor with the pu iron core coil attached It jumps to over 500v. The closer the rotor neo repelling force is to drive coil the the higher the voltage. - .
The 3D printer unboxing.
https://www.youtube.com/watch?v=eO7EcDosnCo
Quote from: TinMan on 2019.01.16, 09:42:25
The 3D printer unboxing.
https://www.youtube.com/watch?v=eO7EcDosnCo
oh gawd good luck. Mine is partially assembled. They reckon an hour to get up and running. We will see.
Quote from: JimBoot on 2019.01.16, 10:38:44
oh gawd good luck. Mine is partially assembled. They reckon an hour to get up and running. We will see.
All done Jim
Only took 4 hours O0
Now we just have to learn how to use it lol--if it works that is lol
Brad
And the build video
https://www.youtube.com/watch?v=_GYvSFZ1MGE
Concerning the anomaly found by Tinman about the higher average current through csr2 then through csr1 and to discuss the LTspice simulation of this circuit, i think its better to move to another thread as to not further disrupt this one.
I parked it here for now:
http://www.overunityresearch.com/index.php?topic=3691.msg71266#msg71266
Thanks, regards itsu
Can confirm the upgrade to the bed cable. Mine has the single red/black.
But I believe the other issue is the constant movement back and forth of the print bed. I've got gobs of glue on my cable, to spread the forces out backwards from the connection.
That 3D printed part for under the extruder. A tip is to glue it in place. Damn thing has come adrift the odd time over many uses, pushing itself through a print and ruining that print.
A fan on the outside of the power supply is another upgrade, as well as a C shaped filament guide to put at the top of the printer (plenty of those on Thingiverse).
You'll get a major buzz when it prints an upgrade for itself.
Btw, my first print was the Chinese coin example and turned out great, using the white PLA. It's fine for test prints.
Now to watch the build vid.
O0
Quote from: Slider2732 on 2019.01.16, 21:21:58
Can confirm the upgrade to the bed cable. Mine has the single red/black.
But I believe the other issue is the constant movement back and forth of the print bed. I've got gobs of glue on my cable, to spread the forces out backwards from the connection.
That 3D printed part for under the extruder. A tip is to glue it in place. Damn thing has come adrift the odd time over many uses, pushing itself through a print and ruining that print.
A fan on the outside of the power supply is another upgrade, as well as a C shaped filament guide to put at the top of the printer (plenty of those on Thingiverse).
You'll get a major buzz when it prints an upgrade for itself.
Btw, my first print was the Chinese coin example and turned out great, using the white PLA. It's fine for test prints.
Now to watch the build vid.
O0
Yes,a fan is being fitted to both power supply and mother board.
Brad
The PSU isn't anything fancy, it's the same as an 80's arcade machine brick....not that such info is very useful to most.
But also, an ATX power supply for a PC will work fine, if you ever get any bother from it.
Has anyone tried to print Jims rotor yet? I did and only got a piece of the side at the start so aborted.
Also, Jim is there a way to orient the spool to have the threads in the vertical top side? Support pieces
are turning out to be a real pain to remove.
thay
Quote from: Thaelin on 2019.01.17, 05:17:24
Has anyone tried to print Jims rotor yet? I did and only got a piece of the side at the start so aborted.
Also, Jim is there a way to orient the spool to have the threads in the vertical top side? Support pieces
are turning out to be a real pain to remove.
thay
I'm not following sorry. Not sure if you mean threads in the rotor vertical on the face instead of side? I've been fighting the good fight with thermistor insulation until my new printer arrives. So I am still yet to print it.
Quote from: TinMan on 2019.01.16, 14:44:00
And the build video
https://www.youtube.com/watch?v=_GYvSFZ1MGE
Nice build - hope I don't have to buy a tap .
Quote from: TinMan on 2019.01.15, 06:23:16
I left the magnet short so as we can slide in a shorter magnet in each end to see what difference it makes during testing.
I am hoping that this design will totally remove all magnetic fields at L2 and L3 during the pulse input.
Brad
Having messed around with coils whilst waiting for a new printer, I'd be very interested in seeing what effect those end magnets may have. Reminds me of one of Luc's experiments with like poles facing at the end of each core.
There are 2 things am a bit stumped on at the moment, the PVC sleeve and the metal tube with a slit.
I was thinking that simple electrical tape would do for the PVC, if it's just a case of shorting to the metal tube. I may well have missed if there is some capacitance/static involved.
A simple 15mm diameter tube seemed like a good idea, with thin walls, as it would allow the overhangs for the solid state coil too.
It printed fine vertically on the Anet A8 at 2mm, using a 'skirt' to stop it falling over.
Have attached the STL and Gcode files...not that it takes more than a minute to do in Tinkercad.
Quote from: Slider2732 on 2019.01.17, 21:20:12
There are 2 things am a bit stumped on at the moment, the PVC sleeve and the metal tube with a slit.
I was thinking that simple electrical tape would do for the PVC, if it's just a case of shorting to the metal tube. I may well have missed if there is some capacitance/static involved.
A simple 15mm diameter tube seemed like a good idea, with thin walls, as it would allow the overhangs for the solid state coil too.
It printed fine vertically on the Anet A8 at 2mm, using a 'skirt' to stop it falling over.
Have attached the STL and Gcode files...not that it takes more than a minute to do in Tinkercad.
I think the mag and pipe need to be separated enough so that the pipe forms an independent mag field rather than just an extension of the neo's one.
Ok up and powered on the Anet E12. Now to look for software for it. Slider you're using Cura for your Anet?
Quote from: JimBoot on 2019.01.18, 06:50:06
Ok up and powered on the Anet E12. Now to look for software for it. Slider you're using Cura for your Anet?
Using Slic3r and printing some test pieces. Will do a long run tonight on the rotor.
After a couple of hours tweaking the setting's,and getting to know my machine(learning from scratch here :D),here is my first test print.
I can believe how well my first attempt turned out--these machines are awesome O0
Now 35% through building a pyramid O0,and after a bit more tweaking,it looks even better.
Brad
Quote from: TinMan on 2019.01.18, 11:04:09
After a couple of hours tweaking the setting's,and getting to know my machine(learning from scratch here :D),here is my first test print.
I can believe how well my first attempt turned out--these machines are awesome O0
Now 35% through building a pyramid O0,and after a bit more tweaking,it looks even better.
Brad
Yeah ok the hardest part was the T nuts. I'm still tweaking though. My layer adhesion sux atm. Still working out the software. I'm telling you.. you conceive of something, design it and print it. Custom brackets, bearing holders... but I do also have a kiln here so I will be able to sinter metal prints :)
well that did not go well - looks like Y axis issues.
Quote from: JimBoot on 2019.01.19, 04:14:33
well that did not go well - looks like Y axis issues.
Yep,id be saying Y (why) as well lol.
Now Jim
I have been trying to print your coil bobbin,and cannot get it to print flat side down.
It either tries to print on it side,or if i rotate it in Cura so as it's the right way up,it wants to print upside down ???
If i turn it upside down in Cura,it still wants to print upside down C.C
Brad
Quote from: TinMan on 2019.01.07, 05:16:52
In regards to my last post,i have decided to build another coil the same as the one with the magnet core,but with a steel core,and do a side by side comparison.
Brad
I was thinking something similar, but rather than a steel core for comparison, use an air core.
I'm not sure if you've reported your results yet, what did you find with the comparison, and could you try it with an air core as well?
Brad, in at least some of your videos, you present the following measurements on your circuit:
~ 500mW(avg) Pin... O0
~ 2W (rms) "...by the cap to the system (https://youtu.be/LyNCcnXbwdY?t=1393)..."... >:(
Question: Why are rms computations being used for v(t) x i(t) power measurements? :o
Try again, and this time use avg[v(t) x i(t)]
Jim - Yep Cura 3.5
3.6 is out but am fine with the older version. They seem to bring a new one out every few weeks, no idea why.
Brad - Congrats. You'll have Zelda rings for the wife and all sorts of stuff printed out, all over the house. Guys who are handy with their hands always take well to these printers. I just considered it a decent learning for how they work, by having to build and tweak.
Your prints look like 0.1mm quality, very nice.
For Jim's spool, I set Cura to print a raft, then add 'Support', 'Everywhere'. Seems to always be WYSIWYG for me. Am not sure why it would flip it.
Quote from: TinMan on 2019.01.19, 13:29:38
Yep,id be saying Y (why) as well lol.
Now Jim
I have been trying to print your coil bobbin,and cannot get it to print flat side down.
It either tries to print on it side,or if i rotate it in Cura so as it's the right way up,it wants to print upside down ???
If i turn it upside down in Cura,it still wants to print upside down C.C
Brad
LOL - no sorry that sounds like user error. I print it on it's side as it needs support material to print. If I print upright I have to clean all the threads.
Quote from: poynt99 on 2019.01.19, 17:53:41
Brad, in at least some of your videos, you present the following measurements on your circuit:
~ 500mW(avg) Pin... O0
~ 2W (rms) "...by the cap to the system (https://youtu.be/LyNCcnXbwdY?t=1393)..."... >:(
Question: Why are rms computations being used for v(t) x i(t) power measurements? :o
Try again, and this time use avg[v(t) x i(t)]
Maybe you missed these ones at 24v
P/in is AVG CVR1 current x 24.1v
P/out from cap is AVG CVR2 current x 24.1v
V is across the cap in both cases.
Now,if CVR1 and C1 were part of the power supply,then there would be no questions asked(see schematic below),as P/in would be 81.2mA @ 24.1v + the small amount of P/in from the FG.
This would be measured at the power supply output.
But because CVR1 and C1 are considered to be part of the circuit,questions are being raised as to how or why power from the cap seems to be greater than power being sent to the cap.
Brad
Quote from: TinMan on 2019.01.20, 02:59:21
Now,if CVR1 and C1 were part of the power supply,then there would be no questions asked(see schematic below),as P/in would be 81.2mA @ 24.1v + the small amount of P/in from the FG.
This would be measured at the power supply output.
I'm not following you here. Can you elaborate?
Quote from: poynt99 on 2019.01.20, 16:27:40
I'm not following you here. Can you elaborate?
Well,power to the coil would be calculated by using the voltage value of the cap x the current value flowing through CVR2. But because CVR1 and the cap are considered part of the circuit,then we are calculating P/in by voltage across the power supply x current through CVR1
Quick to setup... and 60 hours to calibrate... egggh. Should have bought a kit.
Getting closer. Will do a full run tomorrow, then I have to print a spool winding jig.
Quote from: TinMan on 2019.01.20, 23:01:51
Well,power to the coil would be calculated by using the voltage value of the cap x the current value flowing through CVR2. But because CVR1 and the cap are considered part of the circuit,then we are calculating P/in by voltage across the power supply x current through CVR1
Wouldnt we want to use the voltage of the coil and the current of cvr2 to measure the power to the coil?
Mags
Quote from: Magluvin on 2019.01.21, 23:49:56
Wouldnt we want to use the voltage of the coil and the current of cvr2 to measure the power to the coil?
Mags
Yes we would
Quote from: JimBoot on 2019.01.21, 12:12:22
Quick to setup... and 60 hours to calibrate... egggh. Should have bought a kit.
Getting closer. Will do a full run tomorrow, then I have to print a spool winding jig.
Well,I've never used a printer before,and i am now printing near perfect prints.
This el cheapo printer dose an excellent job when all the parameters are set correctly in cura O0
Brad
Quote from: JimBoot on 2019.01.21, 12:12:22
Quick to setup... and 60 hours to calibrate... egggh. Should have bought a kit.
Getting closer. Will do a full run tomorrow, then I have to print a spool winding jig.
Hey Jim
If i give you a couple of pictures and dimentions,can you draw up and make a file to print out the parts?,as im yet to learn that stuff.
Quote from: TinMan on 2019.01.22, 04:45:01
Hey Jim
If i give you a couple of pictures and dimentions,can you draw up and make a file to print out the parts?,as im yet to learn that stuff.
of course. When you setup a tinkercad account, I can share and you can edit your own version. I had Y axis creep, Z motors out of sync, didn't realise that after I'd leveled the bed 10,000 times., and a stuffed nozzle I had to replace. Cable management was also an issue so I'll printing some brackets for that.
Quote from: JimBoot on 2019.01.22, 05:45:35
of course. When you setup a tinkercad account, I can share and you can edit your own version. I had Y axis creep, Z motors out of sync, didn't realise that after I'd leveled the bed 10,000 times., and a stuffed nozzle I had to replace. Cable management was also an issue so I'll printing some brackets for that.
Is tinkercad the program you use to draw up your prints ?
Brad
Quote from: TinMan on 2019.01.22, 06:53:37
Is tinkercad the program you use to draw up your prints ?
Brad
Tinkercad is for design with it's own editing environment. You export an stl file of your model and get it ready for your printer in Cura.
Quote from: JimBoot on 2019.01.22, 09:28:19
Tinkercad is for design with it's own editing environment. You export an stl file of your model and get it ready for your printer in Cura.
So,design on tinkercad. Then save as stl file.Then import to cura,make changes to settings as desired ,then save to card and print ?.
Quote from: JimBoot on 2019.01.22, 05:45:35
of course. When you setup a tinkercad account, I can share and you can edit your own version. I had Y axis creep, Z motors out of sync, didn't realise that after I'd leveled the bed 10,000 times., and a stuffed nozzle I had to replace. Cable management was also an issue so I'll printing some brackets for that.
OK,below is a couple of pictures of the bits i have to print,along with a drawing showing the sizes.
Is this enough info ?
Brad.
Yep I reckon I'm capable of that.
Maybe one more pic so I can see the depth of the 5mm lug.
Quote from: JimBoot on 2019.01.22, 10:35:09
Yep I reckon I'm capable of that.
O0 O0 O0
Well this will be my first paying print job ;)
These bits would cost the boss $23.00 each from the manufacture,so i said i would do them for $10.00 each.
Need four of them,so thats 2 more rolls of PLA ;D
Quote from: JimBoot on 2019.01.22, 10:35:09
Yep I reckon I'm capable of that.
Maybe one more pic so I can see the depth of the 5mm lug.
I dont have anymore pics
Bits are round,so widest part of the lug is 5mm
It would look like a woodruff key from top looking down.
Top looking down pic of lug
The one in the actual pic is slightly damaged,and so leans out a bit.\But should be vertical
Here you go.https://www.tinkercad.com/things/lWWJOSfbFlb-tinmans-thingy
Quote from: JimBoot on 2019.01.22, 11:46:53
Here you go.https://www.tinkercad.com/things/lWWJOSfbFlb-tinmans-thingy
Bloody hell Jim,that was fast O0
Thank you very much.
I just had a quick go on tinkercad,and have a cylinder printing ATM--but couldnt figure out how to put the angle and lug on it.
Cheers
Brad
Quote from: TinMan on 2019.01.22, 11:51:05
Bloody hell Jim,that was fast O0
Thank you very much.
I just had a quick go on tinkercad,and have a cylinder printing ATM--but couldnt figure out how to put the angle and lug on it.
Cheers
Brad
I surprised myself actually. I used a cylinder, the rooftop triangle in "hole" mode then chopped off the waste with a cube hole, then grouped it all together. You can ungroup that and see the pieces. edit. and another cylinder in the the base as a hole.
First one done Jim.
Worked out a treat O0
3 more on the go now.
Oh,and an RC prop i scaled up and printed last night.
This one is for the 2 meter solar boat O0
Brad
Ha sweet. I'm getting stringy prints again witht the filament I'm using and need to do some more work on the z motors as they wander 4mm out over a few hours of use. I don't like the threads peaking through on the rotor so I'll see if I can tidy the model up a bit.
Quote from: JimBoot on 2019.01.22, 20:09:50
Ha sweet. I'm getting stringy prints again witht the filament I'm using and need to do some more work on the z motors as they wander 4mm out over a few hours of use. I don't like the threads peaking through on the rotor so I'll see if I can tidy the model up a bit.
What printer did you get Jim?.
Im just using the cheap $18.00 filament from ebay,and seems to work just fine-so far.
Perhaps we should open a 3D printer thread , so we have somewhere to share all this 3D printing stuff?.
Brad
Quote from: JimBoot on 2019.01.22, 20:09:50
Ha sweet. I'm getting stringy prints again witht the filament I'm using and need to do some more work on the z motors as they wander 4mm out over a few hours of use. I don't like the threads peaking through on the rotor so I'll see if I can tidy the model up a bit.
Hey Jim
Have you tried using retraction?
Mags
Quote from: Magluvin on 2019.01.23, 00:36:43
Hey Jim
Have you tried using retraction?
Mags
Yeah. I think my problem is the right z axis going out of sync, so bed leveling goes to shit and the filament doesn't lay down properly. Lot's of extruder clicking as well. I got the Anet E12 Brad. Build volume of 300x300x 400. I think a thread would be good. I'm a big fan of this iron PLA. I'm wondering if we could produce a structure that would minimise eddy currents with max permeability. Seems to behave like ferrite but obviously not in the heat dept. I will be running some side by side tests of ferrite vs iron PLA. Have not got a protocol for them yet though.
Quote from: JimBoot on 2019.01.23, 00:55:50
Yeah. I think my problem is the right z axis going out of sync, so bed leveling goes to shit and the filament doesn't lay down properly. Lot's of extruder clicking as well. I got the Anet E12 Brad. Build volume of 300x300x 400. I think a thread would be good. I'm a big fan of this iron PLA. I'm wondering if we could produce a structure that would minimise eddy currents with max permeability. Seems to behave like ferrite but obviously not in the heat dept. I will be running some side by side tests of ferrite vs iron PLA. Have not got a protocol for them yet though.
Hey Jim
I have started a 3d printing thread.
Perhaps repost your Z axis problem there,as i have a fix for that.
Brad
Quote from: TinMan on 2018.12.30, 23:58:27
Ok,so i was thinking about this some more,and i believe there would be a benefit in putting a slot along the length of the steel tube,so as it is not seen as a shorted turn.
At these low RPM,it may not have much effect,but every bit helps.
So if you have not built your drive coil yet,it may be something to consider ?
Brad
Just thinking, if it were not a slit tube, it could act as a shorted turn to the magnets on the rotor when they go by?
Mags
Quote from: Magluvin on 2019.01.25, 02:33:52
Just thinking, if it were not a slit tube, it could act as a shorted turn to the magnets on the rotor when they go by?
Mags
That is what i was refering to when i said-at these low RPM. O0
What if it was just a solid iron core? Would it also be seen as a shorted turn by the rotor magnets?
Quote from: poynt99 on 2019.01.25, 02:47:14
What if it was just a solid iron core? Would it also be seen as a shorted turn by the rotor magnets?
We can print any shape. If solid, same diameter all the way? Are there better structures that open up with 3D Printing iron PLA.Seems ferrite (non-conductive) like chacteristics. soft iron particles suspended in PLA.
Quote from: poynt99 on 2019.01.25, 02:47:14
What if it was just a solid iron core? Would it also be seen as a shorted turn by the rotor magnets?
It would be seen as just a short,which would result in eddy current heating of the solid core.
The joule energy value of heat produced will be equal to the required energy to move the magnets past the core.
Brad
Quote from: TinMan on 2019.01.25, 04:35:50
It would be seen as just a short,which would result in eddy current heating of the solid core.
The joule energy value of heat produced will be equal to the required energy to move the magnets past the core.
Brad
Speaking of shorted coils, just realised that the bobbin I wound as the drive coil uses Copper PLA as the keepers which I thought was copper in colour only.... doh!
Quote from: JimBoot on 2019.01.25, 22:25:33
Speaking of shorted coils, just realised that the bobbin I wound as the drive coil uses Copper PLA as the keepers which I thought was copper in colour only.... doh!
Hey Jim
Well Id try to put a meter to it. I dont suspect that the copper particles in the pla are all in physical contact to cause a problem. I believe there is more plastic involved than the metals it contains. could probably judge that by weighing the impregnated pla in comparison with the real deal, both having the same volume for the comparison, and even make the comparison with just raw pla to see the difference between it and the metalic pla. It would give you a good idea of the differences. Had messed with iron powders and resins. Its never as good magnetically as say a ferrite or solid magnetic metals. even tried super thin CA glue, hoping it would just soak in without increasing the distance between particles. The CA wouldnt soak in, like water on oil. Just seems better to use the real thing. The best I could do was thin resin and mixing till it was a stiff mud. But still doesnt compare to solid iron or ferrites, of which are a compressed and baked if I remember correctly.
Mags
Thanks Mags, interesting. I've been reading up on powdered cores and their limitations in tranformers due to their round loop soft saturation characteristics. I'll b e testing different structures though, as the research I've read involves solid powdered cores. I'm thinking different infill structrures may create more complex paths for the flux resulting hopefully sharper saturation. I'm probably completely wrong but hey, won't stop me trying :) I would have thought I would see eddy currents though in a copper powered core like my bobin keeper. I thing my iron pla is up around 80% not sure about the copper. There's another which is up around 95% so I'll be getting some of that soon.
Scope dramas still with CVR2 C.C
https://www.youtube.com/watch?v=7FIty_SgbZI
Quote from: TinMan on 2019.01.27, 06:01:34
Scope dramas still with CVR2 C.C
https://www.youtube.com/watch?v=7FIty_SgbZI
Damn I need a new scope?
Thanks for the video Brad! (and thanks for not rubbing in your DMM results).
I am surprised indeed that the AVG measurements in the CVR's aren't the same. >:( I don't have an explanation as to why.
Regarding the scope/meter discrepancy for a moment, what is the AVG CSR2 voltage by manual calculation?
If we assume that you are seeing about 750mVp across CSR2 and using a 5% duty, that would calculate to 37.5mV AVG. This is lower than your meters were indicating, and right around what the scope is saying. :o
In either case, you are measuring more AVG current in CSR2 than in CSR1, so something is different there.
If you have indeed managed to extract some energy from the magnet, perhaps you might consider pausing at this time and investigating further to see if a SS device is possible.
Perhaps doing a differential measurement with two probes across the CSR2 could shed some more lights on the voltage amplitude (just to widen the measuring possibilities).
Gyula
Quote from: poynt99 on 2019.01.27, 13:51:34
Thanks for the video Brad! (and thanks for not rubbing in your DMM results).
I am surprised indeed that the AVG measurements in the CVR's aren't the same. >:( I don't have an explanation as to why.
Regarding the scope/meter discrepancy for a moment, what is the AVG CSR2 voltage by manual calculation?
If we assume that you are seeing about 750mVp across CSR2 and using a 5% duty, that would calculate to 37.5mV AVG. This is lower than your meters were indicating, and right around what the scope is saying. :o
In either case, you are measuring more AVG current in CSR2 than in CSR1, so something is different there.
If you have indeed managed to extract some energy from the magnet, perhaps you might consider pausing at this time and investigating further to see if a SS device is possible.
ATM,i am still not 100% sure about the DMM reading--we have to remain open minded.
The scope is what is bugging me. Why dose ch2(CSR1) always give the correct reading,and ch1 on CSR2 always change with the number of cycles displayed on the screen ?.
If the scope is having trouble reading CSR2,then the DMM maybe as well.
Brad
The two wave forms are very different.
The peak-to-average ratio in CSR1 is very low compared to that of CSR2. CSR1 has a high DC initial average already, and the peak amplitudes are on the order of the avg level. CSR2 on the other hand has a 0V DC initial average, with a relatively high/short duration peak, so it is going to be more challenging for any measuring device than CSR1, even though it should be no problem for either the meter or the scope.
What I might suggest is to perform some tests on a few different meters, AND the scope to see when/if they ever agree. Just pipe the FG signal straight in to all at the same time. Keep about 10 to 20 pulses on the scope display, and vary the PW from 1% up to 50% duty. Write down the levels of each at a number of PW's. So that you have minimal variability from the scope's trigger, use the FG output to trigger the scope. That will keep it constant so it can be factored out of the test.
Quote from: gyula on 2019.01.27, 14:48:53
Perhaps doing a differential measurement with two probes across the CSR2 could shed some more lights on the voltage amplitude (just to widen the measuring possibilities).
Gyula
OK,so how would i go about that?,as in ,my scope and FG share a common ground,and so one probe would then be connected directly to ground,as one side of CSR2 is the ground connection.
Brad
For differential voltage measurement, you do not need to connect any of gnd clips of the probes to the circuit, they are left floating. Just attach the two input pins of the two probes across CSR2. I meant to do separately this measurement for both CSRs
See this video on differential measurement but I think you know it https://www.youtube.com/watch?v=VWodjUSkYVE (https://www.youtube.com/watch?v=VWodjUSkYVE)
Gyula
author=poynt99 link=topic=3684.msg71663#msg71663 date=1548597094]
QuoteThanks for the video Brad! (and thanks for not rubbing in your DMM results).
Regarding the scope/meter discrepancy for a moment, what is the AVG CSR2 voltage by manual calculation?
QuoteIf we assume that you are seeing about 750mVp across CSR2 and using a 5% duty, that would calculate to 37.5mV AVG. This is lower than your meters were indicating, and right around what the scope is saying. :o
In the video,i was using a 7% duty cycle,which gives an average of 52.8mV.
The meter was showing at 46mV.
QuoteI am surprised indeed that the AVG measurements in the CVR's aren't the same. >:( I don't have an explanation as to why.
Well,if you are interested in my thought's,here they are.
With a standard coil,having a steel or ferrite core,when we send a current through the coil the magnetic field is built from a 0 value up. So we start with no field,and then that field begins to build and cut through the windings as the current begins to flow. When we disconnect the current,the field collapses back through the winding's and the current continues to flow in the !same! direction<--this is important to remember.
Now,when we have a PM as the core we already have a magnetic field that exists through the winding's,but where that field is opposite to that of which the coil will produce once energised.
So now when we energise the coil the existing field must first collap's through the windings before the opposite field can be built by the induced current through the coil.
Now,the question is-->when the existing PM field is collapsing,which way will the induced current flow?
Well we know that when the current to the coil is cut off,the inductive kickback current keeps flowing in the same direction. This can only mean that the current produced by the collapsing PMs field through the coil must produce a current flow in the opposite direction to that of the induced current.
Why can we see this reversed current?
Well we can,we just dont know we are looking at it via the extra current at CVR2.
One side of CVR2 is X amount above 0,and the other side of CVR2 is X amount below 0,instead of being a 0 value.
It would seem to me that we have two opposite flows of current meeting at CVR2,or the potential across the coil becomes opposite to that of the supply for a brief period of time.
If we look at the scope shot below,how is the inductive kickback period nearly twice as long as the current input time?.
Anyway,they are my thoughts for what there worth.
QuoteIn either case, you are measuring more AVG current in CSR2 than in CSR1, so something is different there.
If you have indeed managed to extract some energy from the magnet, perhaps you might consider pausing at this time and investigating further to see if a SS device is possible.
I did post a design for a SS version some pages back on this thread.
I am not sure anyone took up the !theoretical! project?.
Perhaps i will take the time today to give it a shot,while waiting for the rotor to be perfected by Jim.
Brad
Quote from: gyula on 2019.01.28, 00:18:43
For differential voltage measurement, you do not need to connect any of gnd clips of the probes to the circuit, they are left floating. Just attach the two input pins of the two probes across CSR2. I meant to do separately this measurement for both CSRs
See this video on differential measurement but I think you know it https://www.youtube.com/watch?v=VWodjUSkYVE (https://www.youtube.com/watch?v=VWodjUSkYVE)
Gyula
I understand that Gyula,but my FGs ground is between the two CVRs,and shares a common ground with the scope.
This means that leaving the scope grounds floating makes no difference,as the SGs ground will be attached to the circuit between the two CVRs,and thus the scopes ground will also be attached to the circuit between the two CVRs.
This also means that one of the scopes probes will be attached directly to ground,regardless of whether or not the scopes ground leads are left floating.
Brad
Quote from: TinMan on 2019.01.28, 00:42:57
In the video,i was using a 7% duty cycle,which gives an average of 52.8mV.
The meter was showing at 46mV.
OK, that makes sense.
Quote
So now when we energise the coil the existing field must first collap's through the windings before the opposite field can be built by the induced current through the coil.
Now,the question is-->when the existing PM field is collapsing,which way will the induced current flow?
Well we know that when the current to the coil is cut off,the inductive kickback current keeps flowing in the same direction. This can only mean that the current produced by the collapsing PMs field through the coil must produce a current flow in the opposite direction to that of the induced current.
In terms of the collapsing field and direction of current, I don't think anything changes with the magnet scenario, other than the core is pre-biased with flux in the opposite direction the coil is going to take it when it is energized. Since the PM field is static, as far as the coil is concerned it is starting off at "0" so to speak. So the prebias allows for more flux swing, in theory twice the amount. There may also be a commensurate increase in the average current in CSR2.
Quote
If we look at the scope shot below,how is the inductive kickback period nearly twice as long as the current input time?.
The length of IK time is not the relevant issue; it can be shorter or longer than the ON time of the switch. The important question is: is the energy in the ON phase and IK phase roughly equal? I think you can do a rough calculation on the IK phase to see if the area under each curve is equal (of course there will be some losses).
Quote
I did post a design for a SS version some pages back on this thread.
I am not sure anyone took up the !theoretical! project?.
Perhaps i will take the time today to give it a shot,while waiting for the rotor to be perfected by Jim.
Brad
Yes I saw that SS idea way back too, but I guess no one ran with it yet. Looking forward to your next video on it.
author=poynt99 link=topic=3684.msg71686#msg71686 date=1548646601]
QuoteIn terms of the collapsing field and direction of current, I don't think anything changes with the magnet scenario, other than the core is pre-biased with flux in the opposite direction the coil is going to take it when it is energized. Since the PM field is static, as far as the coil is concerned it is starting off at "0" so to speak. So the prebias allows for more flux swing, in theory twice the amount. There may also be a commensurate increase in the average current in CSR2.
Well lets have a good think about this.
We agree that there is a static field from the PM core cutting through the coil windings-yes?
Now,the coil produces a magnetic field that is opposite that of the PM field.
In order to induce a current through the coil,the magnetic field must change with time.
Would you not agree that when a small amount of current flows through the coil,the PMs field through the windings will change with time. So the PMs field must be removed from the coil windings before the opposite induced magnetic field from the coil can cut through the windings.
Please know that this is only my understanding,and could quite easily be incorrect.
Brad
Quote from: TinMan on 2019.01.28, 00:42:57
author=poynt99 link=topic=3684.msg71663#msg71663 date=1548597094]
I did post a design for a SS version some pages back on this thread.
I am not sure anyone took up the !theoretical! project?.
Perhaps i will take the time today to give it a shot,while waiting for the rotor to be perfected by Jim.
Brad
Sure, been tinkering with some ideas. Still waiting to turn out a decent print on my machine. Just watching they way it prints though some of the threads may need a bees dick more centring. Wishing there was a Permalloy filament now too :) EDIT : Successful print yay
Quote from: TinMan on 2019.01.28, 07:35:53
author=poynt99 link=topic=3684.msg71686#msg71686 date=1548646601]
Well lets have a good think about this.
We agree that there is a static field from the PM core cutting through the coil windings-yes?
Now,the coil produces a magnetic field that is opposite that of the PM field.
In order to induce a current through the coil,the magnetic field must change with time.
Would you not agree that when a small amount of current flows through the coil,the PMs field through the windings will change with time. So the PMs field must be removed from the coil windings before the opposite induced magnetic field from the coil can cut through the windings.
Please know that this is only my understanding,and could quite easily be incorrect.
Brad
The way I see it is:
- The polarization of the field from the core/coil/magnet will decrease then change polarity through the process.
- Pre-bias allows one to utilize the full B-H curve of the core, rather than only half.
- All else remains the same.
I might try to play with a saturable core in spice and see where that goes.
Quote from: poynt99 on 2019.01.28, 14:24:02
The way I see it is:
- The polarization of the field from the core/coil/magnet will decrease then change polarity through the process.
During this decrease,will a current be induced through the windings that adds to the current flowing through the windings that causes the decrease in the first place?.
I
Quotemight try to play with a saturable core in spice and see where that goes.
I would be interested in seeing your results from that O0
I made a similar magnet cored coil as Brad uses, and redid my tests.
I have 10 5mm / 12mm ceramic magnets stacked together isolated by PVC tubing and a steel tube with a slot over it.
A coil wiht about 200 turns in 3 layers (0.7mm) was wound around this, measuring 1.3mH and 0.8 Ohm.
Same circuit as before, C1 10000uF, C2 470uF, UF4007 diode, power led, IRF530 MOSFET, 100 Ohm gate resistor,
CSR1 and CSR2 1 Ohm 1% inductionfree resistors.
Battery operated FG running at 86Hz square wave (DC) at 7% duty cycle.
Result is still that both CSR's have the same average current going through them measured by DMM's and the scope.
Not sure if Brad still uses a transistor or MOSFET.
Video here: https://www.youtube.com/watch?v=s1FlYIm-4FE
Itsu
i used my magneto sensor (MLX90393) to sense/show the magnetic field of the BMCC (Brad Magnet Cored Coil).
The FG frequency was lowered to 1Hz @ 6% duty cycle.
It seems to work nicely allthough somehow the flipovers stall for some time, then continues again.
Guess the caps need to be filled up faster.
The magneto sensor and application was discussed here:
http://www.overunityresearch.com/index.php?topic=3588.msg70836#msg70836
Video here: https://www.youtube.com/watch?v=kGftPqiOTTI
Itsu
Quote from: Itsu on 2019.01.28, 16:40:11
I made a similar magnet cored coil as Brad uses, and redid my tests.
I have 10 5mm / 12mm ceramic magnets stacked together isolated by PVC tubing and a steel tube with a slot over it.
A coil wiht about 200 turns in 3 layers (0.7mm) was wound around this, measuring 1.3mH and 0.8 Ohm.
Same circuit as before, C1 10000uF, C2 470uF, UF4007 diode, power led, IRF530 MOSFET, 100 Ohm gate resistor,
CSR1 and CSR2 1 Ohm 1% inductionfree resistors.
Battery operated FG running at 86Hz square wave (DC) at 7% duty cycle.
Result is still that both CSR's have the same average current going through them measured by DMM's and the scope.
Not sure if Brad still uses a transistor or MOSFET.
Video here: https://www.youtube.com/watch?v=s1FlYIm-4FE
Itsu
Im still using the transistor.
As my scope is having issues reading CSR2 correctly,and your measurements are normally spot on Itsu,i can only put it down to my DMMs also reading CSR2 incorrectly.
Brad
Hey Brad. Have you tried to increase the freq, but keep the pulse in the same time width where as soon as all events are done, the pulse and the field collapse currents subside, and then the new pulse happens. Just eliminating all the dead time, then the scope and meters shouldt have to work around that dead time. Not that it cant, but maybe it would be better to having it look like an ac or continuous waveform. Just thinking
Mags
Brad,
I doubt very much that both your meters are off, and they do jive with what you had on the scope display when calculated manually. So either you are somehow getting the AVG current to increase with your circuit, or something else is amiss.
You double-checked the CSR2 value? Checked its inductance? Swapped with CSR1?
Thanks for the latest vids Itsu. :)
Quote from: poynt99 on 2019.01.29, 02:45:27
Brad,
I doubt very much that both your meters are off, and they do jive with what you had on the scope display when calculated manually. So either you are somehow getting the AVG current to increase with your circuit, or something else is amiss.
You double-checked the CSR2 value? Checked its inductance? Swapped with CSR1?
Yes,i bought all new resistors,transistor and breadboard.
I also bought a new bnc cable to hook the FG to the scope,but that made no difference to the outcome of the scopes readings for CSR2.
As always,i swapped the CSRs around,and the result was the same.
They are just 1 watt carbon resistors,but at only 50Hz i really cant see any inductance from the resistors playing a part in this.
I have used this effect in another device-one im sure you remember,and the results were much the same,and efficiency values are also very close.
I just cant think of any other way to show that CSR2 has more average current flowing through it,other than all the methods i have used so far.
But the question remains--why cant Itsu get the same results?.
I have just about finished the solid state version,so lets see what happens there,as it is extreemly easy to loop to see if it self runs.
Brad
Jim
If you have time,and as you seem to be able to make printer files fast,i have the design for the solid state version we can print up.
Let me know,and I'll post the drawing here.
Brad
Quote from: TinMan on 2019.01.29, 05:04:09
Jim
If you have time,and as you seem to be able to make printer files fast,i have the design for the solid state version we can print up.
Let me know,and I'll post the drawing here.
Brad
Post away buddy and let's get cracking
Quote from: JimBoot on 2019.01.29, 05:31:08
Post away buddy and let's get cracking
Ok,here is our first attempt at the SS version.
Hope the diagram makes sense Jim.
Hollow through former to be 10.2mm,as 10mm is a bit tight when we get some stringing or overlap.
Former ends (4x) are 15mm outer edge,so if 1mm thick the two smaller coils will be 13mm wide,and main coil will be 30mm wide.
Any other problems-just ask.
Cheers.
Quote from: TinMan on 2019.01.29, 08:32:04
Ok,here is our first attempt at the SS version.
Hope the diagram makes sense Jim.
Hollow through former to be 10.2mm,as 10mm is a bit tight when we get some stringing or overlap.
Former ends (4x) are 15mm outer edge,so if 1mm thick the two smaller coils will be 13mm wide,and main coil will be 30mm wide.
Any other problems-just ask.
Cheers.
Where does the iron pipe site?
Quote from: JimBoot on 2019.01.29, 08:50:57
Where does the iron pipe site?
No iron pipe in this one
Quote from: TinMan on 2019.01.29, 08:57:00
No iron pipe in this one
Ok as a single piece you'll need to print it with some support. Leave it with me.
Quote from: JimBoot on 2019.01.29, 09:07:18
Ok as a single piece you'll need to print it with some support. Leave it with me.
Yep
Got the hang of this printing stuff now O0
https://www.tinkercad.com/things/iCNj2FgDnXF-non-sense-ss-former
Quote from: JimBoot on 2019.01.29, 09:35:31
https://www.tinkercad.com/things/iCNj2FgDnXF-non-sense-ss-former
I'm printing now. See how we go.
Quote from: JimBoot on 2019.01.29, 09:55:34
I'm printing now. See how we go.
Yep,printing now ;D
I'll race ya Jim lol.
Good job--fast O0
Brad
Quote from: TinMan on 2019.01.29, 10:20:15
Yep,printing now ;D
I'll race ya Jim lol.
Good job--fast O0
Brad
you won
Quote from: JimBoot on 2019.01.29, 10:39:23
you won
No,im only 51% through the print.
I think i screwd up the supports though,unless there suppose to look like a bowl of spagettie lol
Brad
Quote from: TinMan on 2019.01.29, 10:51:10
No,im only 51% through the print.
I think i screwd up the supports though,unless there suppose to look like a bowl of spagettie lol
Brad
started again after 1st layer
Quote from: JimBoot on 2019.01.29, 11:00:12
started again after 1st layer
Well i selected support everywhere,but it did not put in the support ???
Trying again lol
And i thought i had this printing all sorted lol.
Brad
Quote from: TinMan on 2019.01.29, 11:07:52
Well i selected support everywhere,but it did not put in the support ???
Trying again lol
And i thought i had this printing all sorted lol.
Brad
Yeah bleah. I'm printing on the side. The learning curve is steep but well worth it I think. Now I have to work out whethe to buy a coil winder or print one. Edit. Just manuallly cranked mine up to 80mm /sec mid print. This is a big test for it.
Quote from: JimBoot on 2019.01.29, 11:33:33
Yeah bleah. I'm printing on the side. The learning curve is steep but well worth it I think. Now I have to work out whethe to buy a coil winder or print one. Edit. Just manuallly cranked mine up to 80mm /sec mid print. This is a big test for it.
After 3 go's,i to have laid it on it's side,as it just will not put support between the two middle disk's.
i am printing at 100mm/sec,as i have been for the last few prints.
All turned out well.
Also what is the calculation you are using to determine coil size vs magnet? I have a stack of 20mm neos giving a combined length of 42.24mm . Thinking of building the ss with them.
Quote from: JimBoot on 2019.01.29, 11:52:54
Also what is the calculation you are using to determine coil size vs magnet? I have a stack of 20mm neos giving a combined length of 42.24mm . Thinking of building the ss with them.
20mm neo's :o
You will need a lot of current to neutralise that field.
So far-so good with the horizontal print O0
Quote from: TinMan on 2019.01.29, 12:06:42
20mm neo's :o
You will need a lot of current to neutralise that field.
Yeah I know but I have a fisher & Paykel stator staring at me begging to be stripped... 90% through the print and the keepers look a little brittle. May want to make them 2mm.
Quote from: JimBoot on 2019.01.29, 12:27:02
Yeah I know but I have a fisher & Paykel stator staring at me begging to be stripped... 90% through the print and the keepers look a little brittle. May want to make them 2mm.
Im 50% done on the 4th attempt lol,but looking good. O0
I made it all 1.5mm.
Not sure what size the 10mm hole will be now?
And yes,that is where i get all my wire,from .4mm for the 40 series,up to 1mm for the 100 series motors.
Always go out collecting them on throwout week lol.
Looks like the tube is a tad thicker than 1.5mm lol.
Yep too brittle I'll start one at 100% infill, cross my fingers and got to bed :)
Quote from: JimBoot on 2019.01.29, 12:40:53
Yep too brittle I'll start one at 100% infill, cross my fingers and got to bed :)
Well the center core was strong enough on mine,but the keepers just snapped off like a potato chip.
Also 10mm magnet dose not fit in the hole,so up'd the scale by .2%.
So will try again tomorrow.
Brad
Quote from: TinMan on 2019.01.29, 14:00:42
Well the center core was strong enough on mine,but the keepers just snapped off like a potato chip.
Also 10mm magnet dose not fit in the hole,so up'd the scale by .2%.
So will try again tomorrow.
Brad
Mmmm internal dia was definitely 10.2 - I Have made the keepers 1mm thicker extended to 64mm and centre dia is 10.3.
Quote from: JimBoot on 2019.01.29, 21:01:04
Mmmm internal dia was definitely 10.2 - I Have made the keepers 1mm thicker extended to 64mm and centre dia is 10.3.
Did it turn out ok Jim?
Quote from: TinMan on 2019.01.29, 22:22:43
Did it turn out ok Jim?
swore a lot last night then went to bed. Bout to print now. The edited version is at the same link.
Quote from: JimBoot on 2019.01.30, 04:29:01
swore a lot last night then went to bed. Bout to print now. The edited version is at the same link.
Well its only 1.15pm here,so by the time i get home from work,you should have it all sorted O0 lol.
Brad
Quote from: TinMan on 2019.01.30, 05:17:39
Well its only 1.15pm here,so by the time i get home from work,you should have it all sorted O0 lol.
Brad
I'll put a et of verniers over it for you :)
Quote from: JimBoot on 2019.01.30, 05:44:30
I'll put a et of verniers over it for you :)
Ok not good. Don't print it. one of the smaller coils is 1mm too big. Keepers need more infill to0. Adjusting now.
Quote from: JimBoot on 2019.01.30, 06:55:07
Ok not good. Don't print it. one of the smaller coils is 1mm too big. Keepers need more infill to0. Adjusting now.
The width of each coil is not critical-- + or - 1 or 2mm wont matter much.
They're exactly 17.2 - 4mm for the keepers. 13.2 . Printing the solid version now. Tinkercad updated.
Ok,success O0
Printed upright.
I increased the size of your original by .07% Jim,printed at 2mm wall thickness,selected support everywhere at 5%,support angle at 0 degree's.
Still have some support material to remove,but the magnet fits spot on,and it feels strong O0
So now just a cleanup--maybe try the heat gun to remove the cobwebs ?.
Winding tomorrow night O0
Brad
To clean up your prints, just barely kiss it with a propane torch. Faster than a heat gun, works well for me. Just don't burn the house down ;)
You'll still want some sandpaper and an exacto-knife to clean up the smaller features. No matter how good your print, there's always a little bit of cleanup..
If you get excessive threading, you can try turning your nozzle temp down a few degrees; it usually helps.
Quote from: TinMan on 2019.01.30, 12:40:02
Ok,success O0
Printed upright.
I increased the size of your original by .07% Jim,printed at 2mm wall thickness,selected support everywhere at 5%,support angle at 0 degree's.
Still have some support material to remove,but the magnet fits spot on,and it feels strong O0
So now just a cleanup--maybe try the heat gun to remove the cobwebs ?.
Winding tomorrow night O0
Brad
Good to hear. What Reiyuki said. A little butane torch is great. I picked up an elcheapo from bunnings that uses a lighter as the fuel tank. $10 I think. My 10mm neos arrived today so it looks like I'll be joinig the SS party as well :)
Quote from: JimBoot on 2019.01.31, 06:31:35
Good to hear. What Reiyuki said. A little butane torch is great. I picked up an elcheapo from bunnings that uses a lighter as the fuel tank. $10 I think. My 10mm neos arrived today so it looks like I'll be joinig the SS party as well :)
Well,i have my tricoil coil wound,and after a couple of hours testing,i see nothing out of the ordinary going on here.
In fact,if i replace the magnets with an iron pin the same size,the efficiency of the coil/transformer rises. C.C
Also,using my old circuit to pulse it,i now have the same average voltage across both CSR1 and CSR2,as i do with any ordinary off the shelf inductor.
It would seem that the steel tube is needed to obtain the effect we are looking for.
Brad
Quote from: TinMan on 2019.02.01, 04:51:42
snip]]
Also,using my old circuit to pulse it,i now have the same average voltage across both CSR1 and CSR2,as i do with any ordinary off the shelf inductor.
It would seem that the steel tube is needed to obtain the effect we are looking for.
Brad
This is important! thank you for these tests.
Now, to get back to the effect by re-inserting the "critical element"! Good luck, friend.
Quote from: TinMan on 2019.02.01, 04:51:42
Well,i have my tricoil coil wound,and after a couple of hours testing,i see nothing out of the ordinary going on here.
In fact,if i replace the magnets with an iron pin the same size,the efficiency of the coil/transformer rises. C.C
Also,using my old circuit to pulse it,i now have the same average voltage across both CSR1 and CSR2,as i do with any ordinary off the shelf inductor.
It would seem that the steel tube is needed to obtain the effect we are looking for.
Brad
That's what I thought. Isn't the polarity of the steel pipe opposite that of the mag in the coil's unpowered state? Aren't we flipping the domains of the pipe and simply masking/supressing the neo with the coil? Meaning the mag field does not collapse on the neo only on the pipe.
Quote from: JimBoot on 2019.02.01, 06:32:55
That's what I thought. Isn't the polarity of the steel pipe opposite that of the mag in the coil's unpowered state? Aren't we flipping the domains of the pipe and simply masking/supressing the neo with the coil? Meaning the mag field does not collapse on the neo only on the pipe.
btw this was why I thought a tube mag ith a pipe on outside and iron core. My thinking is we get two fields of same polarity flipping, so no phase cancellation. Anyway I'm probably way off. I just can't see us flipping the neo dipole but we can the iron mags. Specifically I'm thinking about kunel and Sweet devices
Quote from: JimBoot on 2019.02.01, 07:07:42
btw this was why I thought a tube mag ith a pipe on outside and iron core. My thinking is we get two fields of same polarity flipping, so no phase cancellation. Anyway I'm probably way off. I just can't see us flipping the neo dipole but we can the iron mags. Specifically I'm thinking about kunel and Sweet devices
Well the tests seem to state that we do indeed need a magnetic material between the mags and coil.
Im wondering how a former made from that ferrite or iron filament you have would go?.
BTW,where do you get that filament from,as i should get myself a roll.
Brad
I'm using protopasta Iron PLA. https://www.proto-pasta.com/products/magnetic-iron-pla I'd also like to try Virtual foundary's that has a higher Fe content and you can sinter.
Sorry if its not to scale, but these are what Ive come up with to try the SS version. 1/2in nylon spacer and nylon washers. have to open the washer holes with a dremel to fit the spacer snug then sand the glue surfaces and superglue. Works pretty darn good for nylon. Done it before.
This is a very similar function to the SS orbo I have done back in 2010. This is just a different way of manipulating the magnets field path.
Mags
This version was what I imagined it should be if I could get the right size parts. Long ring magnet in the middle and thin ferrite tubes.
Mags
Quote from: Magluvin on 2019.02.04, 00:14:38
Sorry if its not to scale, but these are what Ive come up with to try the SS version. 1/2in nylon spacer and nylon washers. have to open the washer holes with a dremel to fit the spacer snug then sand the glue surfaces and superglue. Works pretty darn good for nylon. Done it before.
This is a very similar function to the SS orbo I have done back in 2010. This is just a different way of manipulating the magnets field path.
Mags
Have you tried one yet Mag's?.
Brad
Quote from: TinMan on 2019.02.04, 14:03:50
Have you tried one yet Mag's?.
Brad
Havnt tried yours yet. Have to finish that. The orbon I had fiddled with back then but didnt have proper equipment. Will be testing them side by side with yours here soon.
Mags
Ive got a lil tidbit coming in with a sig gen I ordered from banggood.com. Its a small sq wave sig gen that will work perfectly with projects like this and can be built into projects like this for cheap. Up to 150khz with duty cycle. The input can be 3v to 30v and the output is what the input is. Will remember settings when powered of also, which is convenient. ordered 2 of them. I had made my own nitrous controller for my fiero. It pulses a relay around 20hz and I had set the duty cycle for 50%. The relay in turn opens and closes the solenoid to give me a first stage at 50% and then the second stage would be full on solenoid. But this lil do da would be a nice addition to that and could adjust it when desired vs the 555 and tiny pot resistors as I have it now.
https://www.banggood.com/1Hz-150Khz-3_3V-30V-Signal-Generator-PWM-Pulse-Frequency-Duty-Cycle-Adjustable-Module-LCD-Display-Bo-p-1270834.html?rmmds=search&cur_warehouse=CN
Hopefully I have all I need by this weekend. Im done doing jobs at my shop so it will be dedicated to my cars and these projects. So Ill be around more now.
Something i had thought of for this project including the orbons, was to try to integrate the Tesla igniter for gas engines patent circuit to drive the pulse coil. So it will be more of a cap discharge of higher voltages than the input working voltage. My tests on that with a standard ign coil(1ohm and about 6v across the input with 1ohm ballast resistor) were showing a standard points system with 1ohm ballast resistor basically pulls about 72w (resistor and coil consumption) while the points are closed and there is 6v across the coil primary and the igniter circuit reduced it down to 1w max draw. So if it works we could see some savings in the drive circuit here. With the standard ign points, the ring down of the primary oscillating with the condenser were around 300 to 400v, so that much into the coil primary when we discharge the cap via the igniter circuit, I was able to get as good or better spark on the cheaps at the input. And the ign coil stays cold as ice vs the standard points circuit, where even at 6v the thing would get hot.
Mags
please remove if too off topic [claimed energy from Magnets in 1990's
by a Swedish engineer in Aircraft /Aerospace manufacturing field for aircraft flight safety
Werjefelt Magnetic Battery".
topic here https://overunity.com/18144/critique-my-free-energy-idea-please/msg531368/#new
http://www.rexresearch.com/werjefelt/werjefelt.htm
https://www.google.com/search?q=bertil+werjefelt&rlz=1C1NHXL_enUS703US703&tbm=isch&source=iu&ictx=1&fir=_yxskLXTejFdXM%253A%252CJbQXEwIdc4pCMM%252C_&usg=AI4_-kSb1q0Ln98f8DAnJMtm4rL3gy-fjg&sa=X&ved=2ahUKEwj8g_a5xKLgAhVBKqwKHUfMDx8Q9QEwBHoECAYQBA#imgrc=_yxskLXTejFdXM:
Chet, I looked at the RexResearch page you cited; found this:
QuoteIt turned out that the Japanese were very interested in magnets and energy...
Well-funded Japanese research teams have engineered this discovery into reliable units for existing motors. Werjefelt spent two days with an official from Sumitomo and learned that the Japanese motors are running for hours, days, weeks. Japanese industrialists are switching over to the new units, which will use about half as much fossil fuel as existing motors. For example, the television program showed a refrigerator, a vacuum cleaner, and other common appliances with such motors.
Interesting, but I wonder if there is anything published/public about this use of Magnets and Motors running with higher efficiency? or, could we get one of these "common appliances with such motors"??
Quote from: PhysicsProf on 2019.02.10, 20:20:17
Chet, I looked at the RexResearch page you cited; found this:
Interesting, but I wonder if there is anything published/public about this use of Magnets and Motors running with higher efficiency? or, could we get one of these "common appliances with such motors"??
Thats a great question. O0 There was a Japanese company maybe 10yrs ago or so, Genepax that articles were coming out that they had a water energy device and put it in bikes and a car. It was said that as long as you have a bottle of water in the car, your good. Some of the stories disappeared after a month or so. But it never panned out like one might think
something like that would. This is what they are today. https://www.genepax.com/
Mags
Very interesting - thanks, Mag.
Now back to your FG: https://www.banggood.com/1Hz-150Khz-3_3V-30V-Signal-Generator-PWM-Pulse-Frequency-Duty-Cycle-Adjustable-Module-LCD-Display-Bo-p-1270834.html?rmmds=search&cur_warehouse=CN
Wow! Thanks for pointing this out! I like it especially since I can use DC in from a battery pack, or say a 9V battery, and do the experiments far from "60 Hz fog." That's important for some of the tests I want to do.
Quote from: PhysicsProf on 2019.02.11, 00:23:34
Very interesting - thanks, Mag.
Now back to your FG: https://www.banggood.com/1Hz-150Khz-3_3V-30V-Signal-Generator-PWM-Pulse-Frequency-Duty-Cycle-Adjustable-Module-LCD-Display-Bo-p-1270834.html?rmmds=search&cur_warehouse=CN
Wow! Thanks for pointing this out! I like it especially since I can use DC in from a battery pack, or say a 9V battery, and do the experiments far from "60 Hz fog." That's important for some of the tests I want to do.
Cool. Glad you like. Not good for everything but good for a lot of things. The output 3v to 30v(input voltage) is great. Should be able to drive mosfets well, for these sorts of projects. should have mine this week.
Mags
Thanks Mags,
I'll be getting mine from here. https://www.aliexpress.com/item/Signal-Generator-PWM-Pulse-Frequency-Duty-Cycle-Adjustable-Module-LCD-Display-1Hz-150Khz-3-3V-30V/32835541668.html $1.63
Quote from: JimBoot on 2019.02.11, 11:37:58
Thanks Mags,
I'll be getting mine from here. https://www.aliexpress.com/item/Signal-Generator-PWM-Pulse-Frequency-Duty-Cycle-Adjustable-Module-LCD-Display-1Hz-150Khz-3-3V-30V/32835541668.html $1.63
Nice. Cheaper too. Im very interested to try it when it gets here
Mags
Cool project tinman, any updates?
I've had a very similar idea to pulse coils, that induce other coils.
Quote from: unimmortal on 2023.11.18, 11:28:55
Cool project Tinman, any updates?
Here (https://www.youtube.com/@TinManPower/videos) is his YouTube channel.
He does things the right-way.