OverUnity Research

A bench post about my replication of the high ohm series coil motor that is Lasersaber's EZ Spin.

After watching Lidmotor's marvellous homebuild version. I asked him what he thought of using the 500ohm coils from Dollar Tree solar 'dancing flowers'. Instead of persuing a hope that he might try the idea, I decided to have a go myself.
I've reused such coils in other motors, but only ever 1 at a time.
The Lasersaber method entails connecting multiple coils in series...the end of 1 to the start of the next, all the way around the rotor. More coils are needed than magnets.
It's presumably an old practice, but one which he highlighted with the EZ Spin and was previously unknown to myself.
All of this starts to to gain a 'Muller' feel with the coil amounts and magnet ratios and, indeed my own project may be heading in a similar direction.

This non 3D printed version has 4 coils and 2 magnets...had no more dancer coils than the 4.
Coil ohm'age is approx 1800ohms, 1 coil is of a slightly different design for some reason and 1 had its wire snap the other day during a video'd run.
If you've ever seen these little wonders of a coil design, you'll know the wire is something like 48 AWG !!!
The first version of it used Lidmotor's complimentary pair transistor circuit. Worked great, but unfortunately not as well as a reed switch subsequently proved.
The rotor is 'semi-levitated', in that a needle point sits on the bathroom type tile and its top end sits under a couple of magnets. The same type of arrangement worked great for 2 years daily with a 1 coil motor sat under my desk light to solar power it, so wear on the tile is of no importance.

Here's the YouTube upload that shows the build:


Lidmotor then showed a video, where a 5F supercap had powered his motor for over 10 hours. I decided to try with a 1F supercap, just to see how long the thing would go. Previous personal record was 19 minutes for a 1F cap on any motor.
Result was 11hours !
The only reason it stopped, was because 1 of the thin coil wires had broken, from moving the wiring each time a voltage reading was made with the multimeter.

Here's the 11hr run video:




Where to next ?
Well, since that video the other day, there is a biasing magnet that's being tested for combating reed switch metals pull (i've already snipped the original connections and attached copper wires instead).
A washer has been glued to the middle of the rotor, for stability and low speed running improvements. it will now run down to about 0.165V.
The next build will be an 8 coil setup, then 16. Gen coils and output usefulness tests will include adding a blocking oscillator circuit and/or returning the output to the input via various disconnection methods.

Oddities !
The motor will back charge very very readily. No diode or loop around is needed. Spinning the rotor by hand will charge a cap up, to the tune of 10mV every few seconds on a 0.1F cap. The output with no cap is somewhere around a constant 60mV.
That explains the run times :)
Connecting one of my multimeters to the motor, with no capacitor fitted and with the rotor moving slowly from a hand spin, I can set the meter to ohms and the rotor will then spin up to a high speed !
Wonder what causes that ?
Also, a stepper motor makes a great 'power source'. Taking any 2 wires from the stepper and adding a diode to one of them, the stepper will charge a 0.1F cap to 0.8V with approx 20 turns of the shaft forward and backward. It can be left 'in circuit'.

I'm following your awesome progress on this motor-generator, Mark.  Awesome is the word!  O0

Hi Mark,

Nice and interesting build, thanks for showing it. Would like to notice a few things if you do not mind.

During the run down tests I assume you left the probe clips of the voltmeter connected to the 1 F capacitor and you occasionally switched the meter on when wanted to observe the voltage level, right?
If yes, then the presumably 10 MegOhm inner resistance of the meter remained connected in parallel with the 1 F capacitor all the time (regardless of your switching the meter off) and this consumption surely stole charge away from the capacitor. Its loading was surely much less than the total motor consumption but during the 11 hours it may have had a presence and your run time may have been longer if you had fully disconnected the probe clips.
To get some numbers, and assuming 10 MegOhm meter inner resistance, at the starting voltage of 0.877 V the load current from the meter was 0.877 V/10 MOhm = 0.0877 uA (microAmper) and after the 10th hour the voltage was 0.454 V so the load current by the meter was 0.0454 uA. Maybe these tiny currents sound indeed negligible, I think they do count on the long run and may have reduced the run time by half to 1 hour as a rough guess.

To explain one of the oddities you mentioned: "Connecting one of my multimeters to the motor, with no capacitor fitted and with the rotor moving slowly from a hand spin, I can set the meter to ohms and the rotor will then spin up to a high speed ! Wonder what causes that ?"

Multimeters use part of their inner battery voltage as a test voltage when measuring resistors, this means that they put DC voltage across a resistor when you connect one to the probe clips. So this voltage can easily spin up such motor setups. Just check your multimeter set in Ohms (and then in kOhms) with another multimeter set in DC Volts and see the output voltage across the Ohm meter probe clips.

Another notice if you do not mind, you mention that you are going to use some generator coils to utilize the torque of this motor. May I suggest not to use similar high Ohm'age coils for the generator coils because the huge DC resistance would dissipate precious output power. It would be much better to have coils with under 10 Ohm or so DC resistances, though I know this involve more cost. If you could use ferrite cores for the gen coils, cleverly positioned and choosing odd number of them with respect to the magnets number to minimize cogging, that would be a good step I think. 

Finally, you mention parallel connected coils for Lasersaber's 3D motor setup but he connected his coils in series, no? That is why he has less than 2 uA current consumption. I may have missed his earlier setups where he used paralleled coils though.

Thanks,  Gyula

Thanks for the responses guys.

Steve - thought i'd best write it down, as per your suggestion earlier...it's become a research path, unexpectedly and well worth documenting.

Gyula - thank you for the detailed response.
First things first, OOPS, yes, series not parallel. Will change the OP. Luckily that was not part of a series of errors :)
Yes, the thought was that the meter has to use some form of voltage reference for the tested resistance to be measured, but the actualities of that process are appreciated. Another meter doesn't seem to show the effect, but lights LED's nicely on the diode test setting. Neither is a Fluke in any case, so are likely to do things the simplest and cheapest ways.
I did indeed leave the meter connected, mainly for the video, to show readings for the interested and the doubtful.
With the motor now being able to turn at ~0.185V i'll have another go at a 1F run.
Half to 1 hour difference is reason enough to try again too.

Indeed, hand wound lower ohm coils are to be used for the gen part. There is a charge back to the cap with the 500ohm coils, yet for generator coils I was thinking that the resistance may well impede when attempting maximum recovery. In other motor experiments, i've found that mounting a pick up coil on top of and at 90 degrees to a powering coil, has enabled great results and not formed a drag on the motor. Long time ago though, a few years now. Hopefully, something similar will enable 1 coil to run a blocking oscillator and give a light output at least.
Also, a single diode can work just as well in some situations as a FWBR and smoothing cap. Will try that.
Am tempted to go 'full Muller' with a second build. Spaced magnets such that only 1 at a time is in alignment with the powering coils and tiny neodymiums to offset flux (for the reed metals and if a core material is used in the generator coils).
As far as core material, powdered ferrite is one thought.  
Got opinions on this idea ?
I have bunches of ferrite beads and toroids around here, of different unknown specs. A hammer has proven itself worthy with coffee beans in a bag, so that's the idea for the ferrite.
The powder will then go in pieces of drinking straw, sealed at the ends with disks of thin cardboard, which will form the former for a coil.  The coil form will then look like a sewing bobbin type.

An alternative, is a 24/7 runner via solar.
A small solar panel picks up ambient room light, charging say a 2F cap. That should allow enough run time between usable daylight hours even in winter. Especially if I sit the motor on a windowsill. that is readily possible right now  :)
If it went that route, i'd use old 1970's EPROM's for the solar power, just for the quirky factor. Got 2 pairs of them on the kitchen windowsill and they power blocking oscillator LED flashers very well.

Another question.
What's the best and most practical simple regulator ? I'd like the voltage to be no higher than 1V at any time.
One thought is an LED connected at the power input and when it starts to conduct another circuit runs (such as blocking oscillator). If voltage climbs, the LED begins to glow, brighter if full sun were to send the cap far higher than the required run charge. Have used such a technique before, when running a digital watch circuit with wireless electricity. The brighter the LED, the more it's burning off to keep the voltage down. A diode could be added after the LED, to bring down say a 1.7V conducting point.
Perhaps if the 'Muller' idea worked, such a method would save having to buy a Radioshack multi-switch transformer and the need to solder a red wire to run through the table top. Or maybe they're the requirements for obfuscation  8)

Some good ideas to test out, Mark  O0

"Indeed, hand wound lower ohm coils are to be used for the gen part. There is a charge back to the cap with the 500ohm coils, yet for generator coils I was thinking that the resistance may well impede when attempting maximum recovery. In other motor experiments, i've found that mounting a pick up coil on top of and at 90 degrees to a powering coil, has enabled great results and not formed a drag on the motor."

I'm wondering if one can buy lower ohm coils, ready made?  would also make replication easier for those following your path later.

In any case, I would encourage you to keep trying for longer runs ...  shooting for the elusive "self-runner".  I still have a 1-oz gold eagle I want to give - to the first person to reach that lofty goal (and let me test it out, to make sure he's got it!)

And the price of gold is going up lately, too....  Good time to "go for the gold!"
Cheers,
Steve

Now there's a thought...and a golden opportunity.
It's a good thing that gold isn't linked to gas prices :)

Prewound gen coils would be highly useful, being as these powering coils are prewound. Personally, i've no idea about stockists or availablilty, that would indeed be handy to know.
Am thinking 30AWG and of similar diameter to the dancing flower coils, 3 or 4 times thicker. Then the crushed ferrite in a straw technique and see how that fairs.

Tests today were with balancing the rotor at all speeds and all power inputs. That involved shifting magnets around and changing them out on the top piece above the rotor needle.
I've raised the max input to approx 1.3V, but there is some wobble still in mid speed ranges. A failing of this semi-levitated rotor I think, a small wobble sets in on the unsupported top of the needle and then dampens out again. Might well build another while waiting for parts to arrive for a larger build.
Ideally, the rotor would run at a governable speed (a variable pot or such) for max generating speed, linked to max run time. 
Tomorrow will see a pot tested, in series with the coils, just to check it all out for possible usefulness.

Hi Mark,

On ferrite hammering: it would be good to smash ferrite material with known permeability (in the upper hundred or lower thousand permeability range) because there are lot of beads and toroids with rather low permeabilities, these may give less inductance increase for an air coil. Powdered iron usually has lower values.  Otherwise I agree with your "core and coil manufacturing" like that.   8)

Otherwise, there are the audio crossover inductors but these are normally expensive and have relatively high mechanical sizes with respect to your present setup (unless you redesign it).  Here is a cheap ebay offer for any two air core coils between 0.1 mH to 1 mH coils: http://www.ebay.co.uk/itm/Visaton-air-core-coil-inductor-for-hi-fi-crossover-0-1mH-1-0mH-2pcs-/271742498050?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item3f451d7502 

Any such crossover coil elsewhere may cost higher than that, see one at random:  http://www.hificollective.co.uk/components/inductors.html

I do think that radial ferrite choke coils in the few milliHenry range would also serve nicely, with cleverly choosing their position to minimize mechanical cogging (odd-even as I wrote yesterday) around the rotor magnets. A cheap source for such coils is here:
http://uk.mouser.com/ProductDetail/Bourns/RLB0912-102KL/?qs=%2fha2pyFadugxwaL23JRBqzBtAbNivwLR1mUQF%2fEHyxd6BBX8JdU%252bxQ%3d%3d

I will return to your simple regulator question. I assume you would have an input voltage range from say 2 V to 5 or 6 V or so? Or maybe different?

Gyula

Dear Mark.

It seems PM motors are " flavour of the month " !!   :)

I have, on many occasions pressed into service the humble Automotive relay. Particularly the small Black cased 4 pin version. They have a coil resistance of C 113 Ohm, an easily removed soft Iron armature, and a sturdy plastic bobbin, ideal for filling with your chosen powder!!  :)

BTW, I found on eBay the other day Silicon Steel powder readily available, this might prove more suitable for this lower frequency application??

Just some thought.

Cheers Grum.

Hi Grum,

On Silicon Steel Powder, the question is whether the powder conducts electrically when you stuff it into a container volume? Also a question what permeability it may have. If such material can conduct in itself, then means should be found to isolate the particles from each other, unfortunately, and this may be cumbersome and if the resulting permeability turns out to be low, then it is better to hammer a high permeability ferrite piece where isolation is already granted.

On the 113 Ohm coil resistance: this may sound reasonable for such low power motor like Mark has showed, provided the shape, length, etc is also good for the particular task.
Generally you should drive for as low generator coil resistance as possible.

Gyula

Dear Gyula.

Thanks for your prompt response.

Perhaps I should have been a little more explicit. Yes I agree the powder on its own would be highly conductive but combined with a quality Epoxy resin should reduce conductivity greatly.

I have to fully agree with your sentiments as regards coil resistance, I am learning FAST !!   :)

Cheers Grum.

Many thanks for your input fella's, it's fantastic to receive such encouragement  O0

Firstly, I have another reason now to 'go for Gold'.
News is, my eldest daughter will be getting married in September !
If I were to be awarded the coin, it would be gifted to her and her new husband.

The motor is on a run at the moment. Starting voltage of just over 1V and I have several thoughts.
@gyula - that Ebay link to the audio coils is superb, thanks for that ! They look to be exactly the sort of thing wondered about as ready made. Ferrite can go within the bobbin hole.
Would you happen to know whether the ferrite from CRT TV deflection yokes is suitable ?
The size of them seems highly suited for a good quantity of exactly matched ferrite powder.

@Grumage - I concur. The 'White Crow' of a few years back used a 12V relay, was Hall sensor based and charged a 6V lead/acid gel fully from flat when started via a 1F supercap. There is a thought that it's the chemistry of lead/acid that plays a part, to which Berdini has seemed to allude very often, at least with regular car batteries. That battery came from a kiddie toy electric car and was completely flat, having been out of service for probably a couple of years. The motor had only single diodes on its AV plug rectifier (not doubled) and no rear of coil magnet to help with rotor pull toward the relay core, which was still in place. Will attach the schematic, just for reference.
That thing was what gave me an insight into OU. The 'trick' seemed to be the offset position of the relay away from the magnets and the micrometer by micrometer fine positioning over weeks of the Hall sensor to get the exact position. It was only a decent motor until 1 extremely tiny sweet spot was found. Run time on the cap went up to 19 minutes on its own, then I added the 6V battery on the output.
All is sort of off-topic to discuss the White Crow, but is an insight maybe into what has gone before on my own workbench.
Here's a video, where after hours and hours of it running, I started running a PC fan with that formerly flat battery.



I believe that if you witness true OU, or, in fact fully believe you have witnessed it, then replication and building another becomes more possible. A barrier within the mind is broken down and you just want to 'do it again'..
Lots of Dremmel'ing is needed for such relays, some deft cuts with a sharp knife on the plastics and housings, but as run coils relays are superb.

Hi Mark,

Yes the cores used for TV deflection coils should have a relatively high permeability (due to the need for the 50 Hz operation for the vertical coils) so you can start hammering.   :)

Any comment on the voltage regulator question I asked?

Gyula

Good ideas, Mark and Gyula!

Note that Mark lives in Oklahoma...  so shipping to US is important.
http://www.ebay.com/itm/Visaton-air-core-coil-inductor-for-hi-fi-crossover-0-1mH-1-0mH-2pcs-/271742498050?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item3f451d7502
is another source, but from Lithuania...

Mark - just let me know what coils you need, or wire..   ;)

Hi Steve,

Thanks, I somehow thought Mark is in the UK...  :D   I do not know why, perhaps due to the more English than American-like accent...

I think you managed to find the same ebay seller I did in my post #6  ;)  the shiipping cost seems to be same worldwide from Lithuania and amounts to the price for the two coils.  

By the way there are (for crossover purpose) plastic bobbins for cheaper price than those two air core coils. The sellers are mainly from China though, see this at random:
http://www.ebay.com/itm/10PCS-28-14mm-Plastic-Bobbin-Wire-Coil-Former-FR-DIY-Speaker-Crossover-Inductor-/221440869971?pt=LH_DefaultDomain_0&hash=item338ee78653

Gyula

I thought south Cheshire !!  :)

Cheers Grum.

It's all about internet anonymity Grumage, you'd never guess i'm Mexican LOL
Nah, very very close guess there of South Cheshire. Born in Nantwich, Cheshire, Grew up near Northwich, Cheshire. 10 years in Leigh, Lancs then back to Northwich. 2005 met my (2nd) wife and moved to Minnesota where we married. 2008, moved to near Tulsa, Oklahoma :)
The point of that history is that we seem to be moving south and my wife has said that by retirement age we ought to finish in Florida, with millions of other retirees !

Will email you Steve. Am thinking that home wound and made would be the thriftiest route, yet am indeed mindful of the replication avenue, should there be a success. It's obvious there is a belief in this project and i'm humbled by the interest in this thread too.

Gyula. Oh sorry, oops  :-\
The partial reasoning being the improvements of the rotor of today and yesterday. It's far more balanced...and still running without much of any wobble, from a 1.028V start, to 0.435V some 9hrs later. Not sure of any unwanted effects, now that there's a small washer added to the rotor, but that's the point of the run. If it's lost some run time, it's made up for that with the stability in a way.
So as to a regulator, I would think 1V, but it has run fine at 1.3V. That's a test for tomorrow, if this run gets close to the 11hr personal record. It used to have trouble with magnetic flux anywhere past a volt, but the washer weight and flywheel effect has worked well. The cap itself is rated at 5.5V and i'm sure all that would happen is the rotor would fly off. But many optimum parameters are unknown.
Ideally, 1.5V seems logical. Start it up with an AA and then remove for continued running. Everyone has AA's, could just borrow one from a TV remote and off it would go.

An intrigue is the self charging ability...how there can be 60mV or so directly measurable from a spin of the rotor by hand, with no diodes. I mean, feasibly, a completely flat capacitor could be charged by the very same device that would then go on to run it  ???
Another tidbit discovered, is that the LED is still in place across the reed and does light when the multi-meter is used to spin the rotor up (set on ohms).

Still going !  :D
14+ hours and now using 20mV per hour. It's at 0.320V on the 1F supercap and i'm tiring at near 2am.
No doubt will keep getting up to check on it. At least it won't need a bottle warming or diaper/nappie changing.

Have wound a homemade bobbin (to the procedure described) with 30AWG. 280 turns, 3 ohms resistance.
The form approximates the diameter of the 'dancing' coils, similar to the diameter of an AA battery, but is about 4 times as deep as the coils. I took the cue from those forms shown on the kindly researched links above.
More turns could be fitted on the form, it was just that I ran out of the salvaged previous project wire that it was taken from.
Will try it out as aircore tomorrow, then with crushed ferrite in the drinking straw.
To note, the drinking straw is one of those thin types...will there be enough ferrite material in such a size ? should it be epoxy mixed or do I just push the stuff in and glue over the open end of the bobbin ?

Amigo Mark.  :)

Nantwich is 22 miles NE of me and very familiar. 'Twas an old friend, sadly departed, who christened me my pen name " Grum " lived there!

I spent many happy hours at Stapely water gardens, run by a family friend, learned such a lot there.

You have not lost your accent, it came over loud and clear to me !  :)

Best wishes, Grum.

QuoteStill going !  Cheesy
14+ hours and now using 20mV per hour. It's at 0.320V on the 1F supercap and i'm tiring at near 2am.
No doubt will keep getting up to check on it. At least it won't need a bottle warming or diaper/nappie changing.

O0

That 20mV per hour has got to be close to the "leakage" loss from the cap...

Keep that baby going! (and get some sleep man!)

@Grum - Was born at the Barony Hospital...which they knocked down soon afterward  :o

@Steve - Will get some rest when the thing stops lol. Still going, over 26 hours now.
I've uploaded the video already though, of the full day run.
It also shows the first coil form and the now wound 30AWG to test with.

Of course, I hope it keeps going, but you need some rest too!

Does look like slowing down
               ... also the motor LOL

A minor thought... is it possible (hope not) that somehow the LED produces a voltage due to exposure to light, that feeds the motor??  (we've gotta think of such nuances...)
Fairly easy to test -- see if it runs as long in the dark...

Its my experience with fusion research --- scientists look for any "weird" source of the phenomenon, and variations of the experiment, to be sure they've "got something new!"

Ah, good idea..hadn't thought of that.
Just took a reading without the cap and it's -1.8mV
Even though tiny, it's a negative number and would seem to be a negative toward running ?
I put the whole thing right up close to the 18W CFL bulb and it went up to -100mV.
Anything above ~200mV will run the motor, but also unfortunately LED's struggle for any current indoors. 


The motor stopped just short of 27hrs. I then put the biasing magnet in place and it ran for 2 more hours, down to 0.187V.
The biasing magnet seems most effective when positioned as shown in the LED reading pic below.

27 hrs, then another 2hrs -- just WOW!

Quote from: Slider2732 on 2015.01.20, 23:42:55
Ah, good idea..hadn't thought of that.
Just took a reading without the cap and it's -1.8mV
Even though tiny, it's a negative number and would seem to be a negative toward running ?
I put the whole thing right up close to the 18W CFL bulb and it went up to -100mV.
Anything above ~200mV will run the motor, but also unfortunately LED's struggle for any current indoors. 

so it seems the LED might have a negative effect on the operation?

QuoteThe motor stopped just short of 27hrs. I then put the biasing magnet in place and it ran for 2 more hours, down to 0.187V.
The biasing magnet seems most effective when positioned as shown in the LED reading pic below.

would the whole thing run longer -- with the biasing magnet;  and in the dark? 

Good work, Mark.  Hope you can get some rest now, as its getting dark outside, at our longitude. 

I think the likelihood is that the LED doesn't factor. it would though if the motor was near a South facing window.
I'd take the LED off, but do want to be safe with the reed plates not sticking. The best solution all around, I think, is to paint it black :)

The biasing magnet still has a lot of unknowns. It definitely helps when the reed metals finally begin to have an impact at ultra slow speeds. Yet, higher speeds I don't know. Tomorrow i'll have a look see, using a 0.1F capacitor.

Tried the self made coil earlier..it was rubbish lol, I didn't get anything from it, with or without diodes, silicon or germanium.
The dusted ferrite isn't in there yet though. Am actually wondering if I can use 'dancing' coils, with the ferrite stuck to a sheet behind. I know that the pickup coils are supposed to be low ohm, but it's worth a try.

This evening saw a spin up of the rotor for some more voltage and current output tests of the existing coils.
The intention is to keep sampling after build changes, with furthered builds or new builds etc, for comparisons.
A fast hand spin was measured again and up to 250mV is generated. At speed, approx 70uA is generated.
A linear fall off happens with both voltage and current. There seems to be some spiking of both V and uA lower down in the speed ranges (roughly the 5hr part in the last video for example of speed), but i'm only using one meter. Will try with others + with 1 set on volts, another set on uA.

Will be an early'ish night tonight, that's for sure :)

Must just mention a brilliant idea that my wife has had.
Was saying to her about the rotor and the improvements...she said "have you added graphite to where the needle rests ?"
Now THAT is a great idea !
The rotor sits in a little piece that came from a DC motor, it's what the stator wires would attach to. Anyway, it stops any lateral movement of the needle end, but underneath is just the bathroom tile. Adding graphite may make that area even slicker for spinning on.

   Graphite is a good idea.

"Tomorrow i'll have a look see, using a 0.1F capacitor. " 
Right - I was going to mention that, to permit shorter-time runs for comparative tests as you make small changes and test again.
 

  I see that Lidmotor has built the 12-coil version of Lasersaber's EZ spin.  Nice, but as he says in the vid, very tedious to wind 12 coils with thousands of turns of very fine wire!

https://www.youtube.com/watch?v=TZLEQfUixZg&feature=em-uploademail  = Lidmotor build

What you have done, Mark, is to replace this tedious winding process with coils ready-made and cheap from dancing-flowers, and shown that they work very well indeed!

As we continue to share ideas in the community, fast progress should ensue...

@ Slider-or anyone that knows.
How is a DC cap being charged with an AC current?-i believe there are no diodes in the circuits?
The only way i see this happening is if the reed switch is closed on the forward current side of the AC sign wave,and open on the reverse (lower)side of the AC sine wave-would this be correct?.

Think i may join in on the fun here,and see how efficient i can get it to run.
Are we useing the 1 farad cap as a test point?,if so,what voltage are we charging it to at the start of the run/.

Great job Mark O0

In reply to both chaps.
How this started with the 1F, was because that had been the efficiency testing sizing of capacitor for previous builds. Regular 2200uF etc didn't give enough power to spin up and run, sometimes at all and 1F was the only size of supercap owned for years (recovered 2 of them from a junk circuitboard). I note that Jimboot also tested with 1F on his 'Ossis'. Up until LS's motor design they made sense.
This motor would lock up with magnetic flux at about 0.8V at first. The rotor was too light and there was hardly a flywheel effect. Hence the low charge voltage to start a run. 1V seems right for rotor run speed, being easier to calculate from too.
Also, Lidmotor usually shows a 5F and I was intrigued with how the 1F would compare...anything past 1/5th of his builds 10+ hour run was the intention when using a reed. It would allow seeing if I had got the build on the right track.

There seems merit in piggy backing the coils. 1 for power, 1 for pickup ?
As I see it, there is EMF flying out from the front and from the back with each powering pulse.
Why not collect the otherwise wasted rearside energies ?

I'm delighted that you are here too Brad. The Cool Joule got me really interested in these coils, with the idea to use them on that circuit. When first looking for 'dancing' coils to use, the CJ that had run continuously for a few months last year was deliberated on for its coils to salvage. For the heck of it I put a clip lead end across the transistor, the LED fired up again and it's garden light AAA's have been running it for the past few days...nope, aint touching that thing !
In fact, the two systems are not very far apart. Simple, effective, mind blowingly efficient.

It has to be something with the reed closing for the motor charge, I agree. Doesn't that seem very handy for any motor build ! avoids the diode losses and smoothing aspects. Needs a scope study to inspect what is happening.
At only 1 fine placement, that i've found anyway, the motor picks up and runs differently to any other placements. Even a millimeter difference drops that sweet spot and bolsters the peak of wave thoughts. Though, is this not simply pulse wave, on/off ? It also relates to the Hall positioning of the White Crow (mentioned on the previous page).
Intriguingly, LS and Lidmotor have the reed between coils, whereas that spot seems to be actually on a coil on this build.

@ Mark
well i do have hundreds of those 240 to 120 volt stepdown transformers that have the wrong plug for us here in OZ-the coils on them are like the ones i used on the cool joule.
Or,i do have this other coil,but we'll leave that one as a supprize O0

Hi Mark,

Regarding the voltage regulator, first I thought of suggesting a Joule thief (or Amstrong oscillator) with a rectifier added to use as a step-up DC-DC converter, then use a micropower voltage regulator IC  but this may not be the best solution because the oscillator itself may consume more power for its own operation than your motor needs.

Nevertheless, see these interesting posts and tests on these here (I mean the four posts, not the total thread):
http://overunity.com/13175/25mv-joule-thief-powered-by-peltier-merely-using-our-body-heat-free-energy-247/msg348682/#msg348682  
http://overunity.com/13175/25mv-joule-thief-powered-by-peltier-merely-using-our-body-heat-free-energy-247/msg348682/#msg348682
http://overunity.com/13175/25mv-joule-thief-powered-by-peltier-merely-using-our-body-heat-free-energy-247/msg350384/#msg350384
http://overunity.com/13175/25mv-joule-thief-powered-by-peltier-merely-using-our-body-heat-free-energy-247/msg351060/#msg351060


So I turned to boost converters designed for energy harvesting applications and Texas Instruments has the BQ25504 integrated circuit which can start up with an input voltage as low as 330 mVDC and then can continue working down to an input voltage of 80 mV while producing output between 2.5 to 5.25 V, able to charge Li-ion and thin-film rechargable batteries, supercaps or normal capacitors, with MPPT (maximum power point tracking).  Data sheet is here: http://www.ti.com/lit/ds/symlink/bq25504.pdf

Another interesting IC is a buck converter (from also TI), the TPS6273X family, see data sheet here:
http://www.ti.com/lit/ds/symlink/tps62736.pdf  though this has an input voltage range between 2 V to 5.5 V and a resistor programmable output voltage range between 1.3 V and 5 V.

Latest ultra-low power DC/DC converter family is here: http://www.ti.com/lit/sg/slyt593/slyt593.pdf

Yet another good possibility is to study charge pump circuits, the TPS6030X family, they give 3 to 3.3 V output voltage from 0.9 to 1.8 V input, though they have 35 uA (microAmper) self consumption (while the above ICs all have some hundred nanoAmper consumption only). With these switched capacitor charge pump circuits the advantage is that the ICs need five 1uF capacitors only, and no coil. See data sheet: http://www.ti.com/lit/ds/symlink/tps60300.pdf

You may not like tinkering with tiny IC packages like most of the above chips have and I can understand it...  :D

Would like to return to the generator coil question: it is natural that coils with thicker wire and with some hundred turns only, the induced output voltage turns out to be way less than with coils with very fine wire and with some thousand or so number of turns. To compensate this to a certain degree, perhaps stronger magnets could be fixed to the rotor for generating,  and also the use of good ferromagnetic cores for the gen coils can also enhance induced voltage.
The piggy backing sounds ok as long as you start loading the power coil and see what output remains from the 500 Ohm coils, to get a reference and then check a cored home made coil and compare to the (same comparably chosen) load you attach (note also the drag with an Ammeter).  If you have an L meter, you can check the 500 Ohm coil inductance and then approach that value with a cored home made or cored off the shelf coil wound from thicker wire. If you manage to approach it say 35-40 % in value i.e if the 500 Ohm coil has say 1 Henry, then approach it to say with a 350-400 milliHenry cored coil of thick wire,  and there would be a certain value load current where the loss in the 500 Ohm coil will render that coil  a poorer performer than the cored coil. If not, then for lower than that certain value load current just use the 500 Ohm coil.  Maybe I have expressed this understandably I hope...  :)

While it is okay that when the load current for a generator coil is relatively low (and your present motor setup consumes in the microAmper range) then the 500 Ohm DC resistance of a gen coil (like you use for driving the rotor of the motor now) may not introduce a huge loss, remember that in this scenario the best output would be most likely less than the input... I think. one reason for instance, with piggy backed coils the outer coil is further away from the rotor magnets hence induction is different with respect to the coil closer to the magnets.

Sorry for the long post and the many links, I know it takes time to digest...  C.C  And I am not an agent for TI... ;D

Gyula

QuoteI am not an agent for TI

Pffft..yeah sure buddy  8)

Only joking :)

Thank you so much for all those details !!!
You sent me on a wild foray with that OU thread, have read most of it. Spent a couple of hours replicating the FET circuit, though using an MPSA18 transistor. I know i've seen K170's here somewhere on some old junk, just have to find them.
Anyway, with the MPSA18 I can light a white LED to partial brightness, using a 12706 peltier. Should say, it flashes and needs the Emitter/Collector start up with a piece of metal. Incredibly intriguing though to need just 50mV or so.
Also apologies for the delayed response..a large A/C unit fell off the house ! It's one of those that goes where a window should be and has been in situ fine for the 7 years we've been here. An old arcade game back door is now filling the hole in the house C.C

Those boost converters are great. I think an ideal would be a self limiter, where it always outputs the 1V and anything else is handled up or down. The BQ25504 looks to be along the thinking lines. Do bear in mind i've no actual clue, just a vision of a system....and it won't be the likes of Tesla's visions, that can be assured :)
Many thanks for those detailed looks into what is available.
Am fine with SMD by the way, but not the micrometer pin distances of some chips. At 45, more than a magnifying glass would be needed.

Not got an inductance meter, but the logic was easy enough to follow. I think the idea would be to recover the back of the coil energy with a piggyback coil, that formed no load on the front of the coil. Am much more likely to fit the coils between powering coils however and merely try out the piggyback version.

@Tinman - looking forward to the coil surprise !

-----------------------------------------

An idea, remote rotors.
Some time ago, I showed multiple rotors being driven, magnetically, from 1 powered rotor.
The thing being, that such rotors can spin at an exact double multiple of the powering rotor.
If the powered rotor has 2 magnets and the remote rotor has 4, it stands to reason that twice the energy can be extracted.
The powered rotor sees a pull from the remote one, which, in this case is beneficial...because these motors use less power the slower the rotor is moving.
To explain and demonstrate the initial concept, here is a short 'Unlisted' video made this evening.

Hi Mark,

I think it is good idea to use 'remote' rotors, coupled to the powered rotor magnetically. Although I assume if you try to 'load' the remote rotor very gently by your finger, you would see a proportional small increase in the input current draw of the powered motor, the double speed may reward you with some juice from the generator coils applied at the remote rotor. So it is worth trying to explore I think. The diameter of the remote rotor compared to the powered rotor may deserve further thinking as do the number of poles I think.  

What I tried to tell with the example of the 500 Ohm coil vs a much smaller DC resistance coil is that the obviously less induction into the smaller coil, at a given load, may easily become comparable to the loaded output of the 500 Ohm generator coil just because the loss in the latter 'consumes' the higher part of the induced voltage and the outer useful load cannot receive it. But this may be so beyond a certain load current value and not likely at the microamper load currents your present stator cois in the motor may consume.  (Have you measured the uA, by the way? just curious...)

With the piggyback coil you propose to fit it between two powered coils, it is surely worth testing and make conclusions.

MY intention with the first 4 links to overunity.com was not to distract you for a couple of hours from this EZ spin motor...  :D  but to illustrate how a low level voltage source could be enhanced to a higher level. BUT those circuits do consume precious milliAmpers instead of microAmpers for the conversion and unless you have the means to supply the charge consumed from your supercap, the supercap will discharge more rapidly than with the dedicated converters of microwatt self consumption.

Gyula

  Mark,
   I very much agree that experimenting on the use of coupled magnetic rotors - and drawing power from the driven rotor -- is a very good path to explore.  We have seen some claims of anomalous effects before, when magnets are coupled at a distance -- I've seen some anomalies myself...  It's as if the wheelwork of nature "helps" magnets to remain coupled at a distance... ?
Fun stuff!
Steve

  If you have two low-leakage caps, one to run the motor, the other to collect energy from the coupled-rotor (for instance, using a diode), then it is a simple matter to calculate the input energy used and the output energy collected, using E = 1/2 C(Vf^2 - Vi^2), where Vf = final voltage on the cap and Vi = initial voltage on the cap. 
  That is, it is not necessary to re-cycle the power (that may involve too many losses to be practical at this stage) -  one can just measure Eout/Ein ...!  (or if easier, Pout/Pin. )

Will try that certainly.
While the exact mechanism for how the charge enters the run cap isn't known, it may be an idea to set up a similar motor, with no drive cap. Such that, the reed switches by the rotation, but that rotation is derived from being a remote rotor.
As it so happens, I built something very similar a few weeks ago, using 1 'dancing' coil. It was to have been fully dismembered to reuse the coil, but luckily was kept mostly together (pic attached).
If I can get the 2 close enough it might.just.work :)

The 24hr leakage test of the 1F cap produced a start of 1.023V and 29hr end of 0.710V.


Something odd happened last night - the rotor pulled itself to a halt in the usual clockwise direction and started up in reverse !
Then the LED started flashing !
That was with the 0.1F cap. It seems related to the so called 'bias' magnet positioning.
But the reed is set up for clockwise motion. Also, the realisation that the LED is in backwards (doh!).
Try as I might to pick off the returning energy through the LED, it keeps flashing and that doesn't seem right. Normally, putting a diode each way to a collection cap will draw the spike energy from an LED.

Here's a quick video demo of what happened.
Another funny thing, was that I discovered the CFL emissions watch circuit that I timed with, kept running when the light was switched off for up to 5 minutes. By using it for timing and it getting its power from the lightbulb, a panic set in of losing the time...but it came back up fine when the light was switched back on.

  Anomalies are fun.  I think that's how I learn best - facing an anomaly, and one day, that anomaly will turn into a discovery of some significance (IMO).

QuoteWhile the exact mechanism for how the charge enters the run cap isn't known, it may be an idea to set up a similar motor, with no drive cap. Such that, the reed switches by the rotation, but that rotation is derived from being a remote rotor.
As it so happens, I built something very similar a few weeks ago, using 1 'dancing' coil. It was to have been fully dismembered to reuse the coil, but luckily was kept mostly together (pic attached).
If I can get the 2 close enough it might.just.work

Glad you kept it mostly together!  I realize you have a lot of dancing toys coming and I was hoping they would come sooner; but if you are in a rush, I have four or five here I could send immediately.  Dancing flower, bear, elf, etc.

Because of you, Mark, when I go to the Dollar Tree, I sometimes pick up one or two of these.  O0

Also, I wonder if having the LED the "wrong way" around actually helped you get a longer run?

Haha, Greenbrier Imports may well be a link in OU success  O0
The store throws up some useful gear.
The coils should be here in the next few days I would expect - but - on making inquiries with staff, they did say 'the truck' will be bringing Valentines products this Saturday. There is a Valentine dancer, the devil. So if no word by Saturday afternoon then any needed for now will be picked up right off the shop floor.
Thanks for the kind thought though of course !

Maybe the error with the LED wasn't, in retrospect !
However, the 0.1F did run it backwards, with flashing for an hour. The flash was seen down to around 0.7V, then faded to nothing lower down the voltages. That's not bad, would equate to 10hrs on 1F, though that isn't a scientific appraisal.


It turns out that the remote rotor can't get close enough. So, i've taken the coil off and put it on the small one (seen in the above vid). A very rough arrangement with a germanium diode see's output of approx 40mV from the 1 coil, which doesn't seem half bad.
Its glue is setting at the moment, then a real test will be done, with sending that output to the drive cap too.
Thinking onward. Do I connect individually with individual outputs, or, connect up in the same fashion as the driven rotor in series. If in series then only 1 diode would be needed.

Hi Mark,

Just briefly,  the remote rotor may benefit from using strong Neos there, thus giving more juice.  You may have to place it further away from the powered motor though.
IF the output waveform from the generator coils has a zero crossing and negative amplitude, then a single diode is not enough.  And to avoid using the 4 diode bridge with its double forward voltage losses, there is the full wave voltage doubler circuit with two diodes only, you surely know it:
  http://www.daenotes.com/images/full-wave-voltage-doubler.png   from this site:
http://www.daenotes.com/electronics/devices-circuits/voltage-multipler

A question on your LED, as is shown in the video the LED is in series with the reed with a forward direction, i.e. the LED conducts and lights up (or flashes) whenever the reed is ON, right?  At least this is how you used it in the last couple of days?.

Gyula

Ah yes, I know that circuit quite well. It's similar to the 'inventor3' circuit on YouTube, a variant of which powers my CFL powered clocks, as seen in the timing runs so far.
One thing i've never tried, is a rotor larger than the powering rotor. All of mine have been made from water bottle tops.

The LED is in parallel with the reed. It works similarly to an induction coil with transistor switching, where a diode will be across the coil to prevent reverse energy damage.
As far as my knowledge goes, the LED flashes at the instant the reed opens again.

-----------------------------------

Here's some run data, from the first remote rotor loop return test.
This is where those with a lot more mathematics skills than myself would be appreciated.
Mainly, how many coils of this output would be needed to equal energy used by the driven circuit ?
The best harvesting would appear to be at the start, in the 0.8V to 1V area. Unfortunately I didn't keep an eye on the output from the remote coil, was seeing if the remote rotor dragged the whole thing down. Which, it doesn't, the 0.1F cap put in the same 1hr run as with no remote rotor.
Edit: -or- did the returning energy allow the 1hr run ?!  ???
As usual, 'never thought of that'.

0.1F cap, charged to 0.827V
At the beginning of the run, the output of 1 'dancing' coil and 1 germanium diode was:
40mV @ 250uA

Voltage Time
0.827   1.50pm
0.535   2.10pm
0.475   2.15pm
0.431   2.20pm - Remote rotor lost sync
0.363   2.30pm
0.332   2.35pm
0.305   2.40pm
0.280   2.45pm
0.259   2.50pm
Stopped a couple of minutes later.

Hi Mark,

Thanks for the details on the LED connection and yes it should flash when the reed just opens, the flash due to the voltage spike created by the collapsing field of the coils.

On your remote rotor: maybe it is not the best to use an even-odd (i.e. 2:3) combination for the two rotors, I think either use 2 or 4 magnets on the remote rotor. Or you meant using the 3 larger magnets on the "input" rotor fed by the supercap? 

If you have used a single diode, this means half of the induced power from the gen coil remained unused, it did not get into the puffer capacitor.

Your wrote "At the beginning of the run, the output of 1 'dancing' coil and 1 germanium diode was: 40mV @ 250uA"
How do you mean: the diode drives a separate puffer cap or it drives the 0.1 F supercap or? And how the 250 uA was measured? by an ammeter that shorted the output? Sorry but I do not wish to guess.

I think first the generator side ought to be made to produce at least as high output voltage in a similar puffer cap than the run cap which has the voltage level around 0.8 to 1 V or so AND then load the gen coils puffer cap by a resistor to draw as much current as the input motor draws from input cap. When these two the voltage and current are more or less in the same range, you could try the looping. Of course this is my take on this...

Thanks Gyula

And I appreciate your input  O0

The larger magnets line was deleted from the post soon after posting (thanks for inquiring though, you were quick!). Because of something stupid, by me. I'd put them on another rotor and was aligning them, the rotor has 2 tiny neos embedded in the plastic and then larger or more magnets simply snap on to those embedded ones. Wellllll, the 2 large magnets are roughly the size of an LR44 button cell each and somehow snapped together and sprang from my hand  C.C they landed 'somewhere' in either of 2 filled up parts boxes behind me !
Needless to say, the idea is on pause.
That idea was to put the 3 on a larger rotor, with 6 coils around them.

The output of the remote rotor coil + diode was measured directly via a DMM, then with an analog meter (Sperry SP6-A).
In circuit, the output can't be measured as yet, because of feeding to the 0.1F cap.
Next up will be to fill a capacitor, to look at resultant top voltage and fill rate. Will use a 1000uF and then a 0.1F.
Then to build the circuit you linked to for full rectification.
Yes I did jump the gun a bit there, but was intrigued to see if the depletion rate of the capacitor changed.
ETA: In wireless electricity experiments, my usual field,  i've found that 1 diode will run motors, fill caps etc. Rather than being rectified to Positive and Negative, it seems more a case of Positive and positive + negative, equaling zero. That may not be right, but it works and where a FWBR can give inferior results.  

Mark, I found a site where there is an online calculator for supercapacitors discharge time with known load current, capacitor start and end voltages, capacitor size and its ESR. They consider two methods: the constant current discharge and the resistor load discharge and they calculate an R resistor value which would discharge the given cap with the above given parameters during a measured time in seconds.

Using the link http://www.circuits.dk/calculator_capacitor_discharge.htm  the leakage resistance value for you 29 hour long time data (it is 104400 second) gives R=2858532.01 Ohm i.e. 2.858532 MegOhm. So your 0.1 Farad capacitor has a 2.85 MegOHm parallel resistor built-in forever. 

I got this R value by entering (in this order) the 1.023V, the 0.71V, the 0.1 F, 20 Ohm for ESR and played with the Imax values which approached best the 104400 second for the resistor load discharge line, it was 0.357876 uA, this gave 104400.105 seconds, close enough I think...

Notice that I assumed at least a 20 Ohm ESR for your 0.1 F supercap because for the 5.5 V rated caps this may be true. If not, then it causes a small mistake in this low uA discharge current range but in the several mA or higher current ranges you should know the ESR from data sheet for that type of supercap. I think the supercaps with 2.6 to 2.7V voltage ratings have an ESR of less than 0.2 Ohm or so, the better qualities have 0.05 Ohm or so, this is frequency dependent too.

Quote from: Slider2732 on 2015.01.22, 22:52:05

Next up will be to fill a capacitor, to look at resultant top voltage and fill rate. Will use a 1000uF and then a 0.1F.
Then to build the circuit you linked to for full rectification. 

I would agree with that 'agenda'  8)      Regarding the remaining two larger magnets, if you agree, you could just use them for the remote rotor, so the running motor need not be changed.  So you would have a 2:2 magnet combination for the two rotors, this would not result in a double speed for the remote rotor but maybe you could 'compensate this virtual loss' by using at least twice as big diameter rotor than the running rotor has?  What do you think?

Quote

ETA: In wireless electricity experiments, my usual field,  i've found that 1 diode will run motors, fill caps etc. Rather than being rectified to Positive and Negative, it seems more a case of Positive and positive + negative, equaling zero. That may not be right, but it works and where a FWBR can give inferior results.  

IT would be good to see on a scope what the induced waveform looks like, if you do not have a scope than there remains a comparison test: build the full wave voltage doubler and check the DC voltage levels in the two series capacitors. If they are rather different from each other then the induced voltage has a differing positive and negative half wave amplitudes but still. when you sum them in the two series caps, the result will surely be higher than either one of the halfwaves.  A FWBR "eats up" twice the forward voltage drop from each half wave.  I cannot comment more precisely what you described because of lack of other data but this is not the issue now.

Gyula

Nice link there Gyula and thanks for the calculations.
I wish that Ebay would keep older purchases details (do they ?) because then i'd know some better specifics. Just thought it might be an idea to avail you of my supercaps, in case something jumps out for runs later down the line. I expect to see the ability soon to run well from 1000uF and other electrolytics, but am wondering about the discharge resistance and if it could work in favour of such a motor. I don't want to put a speed brake on the motor with a resistor, but would like to start at higher than ~1.3V with maximum return charging possibilities from pickup coils. In other words, run longer at good generating speeds, with the aim of the charge amount overtaking the run cap storage.
6x 0.1F 5.5V (Maruti ?), it's a letter M in a squared box.
Half dozen bag of mixed from 1.0F to 10F nanoforce by DRL 2.7V.  
2x circa year 2000, 1F 5.5V much larger ones by Maxcap, found on salvaged commercial equip circuit boards.
Finally a 1F 5.5V the size of a small rhinocerous by NEC. That last one must be 20+ years old by the look of it, it's got rust on it too.
If the age brings internal resistance, but all caps actually charge to the correct voltage and Farad rating, does that help a motor at all that wishes to sip power ?

Oh, have only got one of those magnets left unfortunately. 2 flew away and there were only 3.
However, the 'dancing' toys on order should have the same neos on them as you see on the main motor. They can be doubled or trebled on a rotor. Those other two were about to be tested on the remote rotor.
Indeed i'd like to try a larger main rotor and see if smaller rotors will spin fine with it.
Important to all of this, may be that i've had 4 remote rotors all going before now. They will chain too, gets tricky to match speeds but is doable.

Got 3 scopes and all 3 need a scope to sort them out !
1 is an EICO TR-410, older than Noah with a round green screen. Actually that one might still work, you've got me thinking :)
A dual input 20MHz Protek 6502, traces only mains hum now for no reason I can fathom.
A DS-203 pocket scope with firmware issues. Function gen works, but scope inputs don't. We do now have a registered version of Windows 7 since a couple of days ago and i'll reflash it. It didn't want to flash with Peppermint Linux.

1000uF electrolytic charge test, from single 'dancing' coil + germanium diode on remote rotor.
Run supercap was a 0.1F at 0.750V
Test cap repeatedly discharged for 10 secs, for dialectric absorption rest of a repeat tested ~0.036V
Both rotors were running, in sync.
10 secs to charge to 0.136V, 5 times.
Top out ~0.180V after a further 2 minutes.

Starting at 0.900V, the top out is 0.253V.
Starting at 1.000V, the top out is 0.328V.

Thanks for taking these measurements, Mark.
I know you're doing a lot, and it is much appreciated.

Hi MArk,

Well,  although you could do some discharge tests with your supercaps one by one when you have the time, with known (measured) value resistors and with precise time measurements (say with a stop watch) but in order to use the online calculator link you should also know the actual capacitor value too. So there are two unknowns actually, maybe you have access to C meter able to measure in the Farad range, that would be good. Also, ESR meters are also available for capacitors, they look like a smaller DMM, but this depends on the make of course.
By the way, here is a link to an Application Note to measure supercaps parameters, though it is a specific product with known parameters, it may help to explore your caps too. https://www.tecategroup.com/app_notes/CAP-XX_AN1005%20Simple%20Supercapacitor%20Measurement%202-1.pdf  Of course, there are other methods, other members here may be able to suggest further methods.

For your 1 Farad 2.7V Nanoforce capacitors here is link where a value on short circuit current is given, 3.17 Amper and assuming the cap is charged up to the max 2.7V the ESR may be 2.7V/3.17A = 0.85 Ohm. Although the short circuit current is given. other manufacturers do give the direct ESR value for their super or ultracap products, Nanoforce seems not to do so for this 1F type: http://www.alibaba.com/product-detail/super-cap-samll-capacity-2-7v_1644343150.html 
With similar search you may explore some other types of your caps, I used these keywords in google: " nanoforce dlr 1F 2.7V supercap ". 

On your question: "If the age brings internal resistance, but all caps actually charge to the correct voltage and Farad rating, does that help a motor at all that wishes to sip power ?"

I would say yes it does. Aging has larger effect on supercaps when they are not used for years but if you charge them up to just under their rated voltage (say to 2.5-2.6V for a 2.7V cap) and keep the 2.5V voltage on the cap for some hours then disharge it (not with a srewdriver bang but with a resistor) and then repeat the charge up again, your cap will be revived in most cases and would serve like a new one. 

You wrote: "Indeed i'd like to try a larger main rotor and see if smaller rotors will spin fine with it."

When the number of magnets are two and two on the driving and the remote rotors, respectively, then syncronism may be more difficult to maintain because the flux connection between the rotors changes (fluctuates) in a greater degree versus the case when the number of magnets are higher than two on each rotor. So it would be advisable to use say 4 on the driving and 4 on the remote rotor, this latter having twice (or maybe 3 times) as high diameter than the driving rotor has. Of course you decide.

Thanks for the measurements, the "top out" means that the run cap has discharged (in one example) from 0.9V to 0.253V while the test cap was charged up to just 0.253V during the same time, correct? If not, please explain.  Also, try to check with the full wave voltage doubler whether you get similar output voltage on the second 1000uF?

Gyula

@All -
Today will see tests with the doubler circuit and, variants with different components.
The best results managed with 1 'dancing' coil as energy pickup will allow furthering, or, dropping of those as the pickup coils.
It would be preferable, for any future replications at least, to use the 1 coil type throughout.

@Gyula - If this were an indie films forum, you'd be technical director..thanks so much again :)
I have 2 capacitance meters, or should say, DMM's with capacitor measuring function.
Craftsman 82139 - bought new from Sears, simply quit with an unknown issue (not internal fuse) shortly afterward. Sears has shut down !
UNI-T UT136B - a few months old, limited to 10nF to 100uF testing.

For a couple of years, there  was a solar powered rotor spinning under my workbench light. Was a converted 'dancing flower' in essence. Same look to it as this new motor, but just the 1 coil, it used the circuit from the toy and a garden light solar panel.
That was where the remote rotors were mainly researched, for how many could be run and how good the lock to each other could be. The link to your comment, was that the rotor only had 2 magnets, with transistors for balancing on the other 2 points. Looked funky spinning around :)

Want a quick mind messer upper ? lol. Here is a demo of remote rotors turning in the same direction ! no need for clockwise/anti-clockwise/clockwise.




Back to now, I was delighted to see a 1hr run on the 0.1F with or without the remote rotor.
What that means (to me), is that the concept is more valid, for adding remote rotors until energy generated is more than that used. But would have to be aircore. Plus, of course, spinning up and matching 4 or 5 remote rotors would become a P.I.T.A. very quickly !

Yep. the 'top out' was an interest for what the present idea can achieve as a maximum figure (with meter drain) of a 1000uF cap. Just initial findings and exploring. The 'top out' is achieved and maintained well, dropping as rotor speed diminished. So one thought is what speed the rotor needs to spin at to achieve a set criteria. Still thinking 1V is optimum.
100mV is generated very quickly in comparison to the final 'top out' and that is where I think things are headed.
Am relatively new to BEAM engines, but have had great results with especially the Easter and now with a 2 transistor variant. That's my working idea for cap dumping.
The TI range of chips haven't been forgotten about and may well feature, for improvements if something looks promising.

The summary -
2 remote rotors, with 4 'dancing' coils around each in series
Voltage doubler circuits on both
Out to a BEAM harvester triggering at ~1.6V, or lower would be better
Into the run cap, which is likely to be 1000uF or similar, depending on needed throughput and charge. Supercap may be better.
If the voltage on the run cap climbs, an LED bleed off of the extra will form a crude regulator, but allow the motor to have a function, however slight that may be.
(better solutions, such as the TI chips will then move it forward)

Mark: "Back to now, I was delighted to see a 1hr run on the 0.1F with or without the remote rotor."

Most of the energy is evidently consumed in the coils; not in turning the rotors.

It would be good to get our hands on some very low-leakage caps in the 0.01-0.1F range (or so) - does anyone have suggestions? 

Quote from: PhysicsProf on 2015.01.23, 18:07:42
Mark: "Back to now, I was delighted to see a 1hr run on the 0.1F with or without the remote rotor."

Most of the energy is evidently consumed in the coils; not in turning the rotors.

It would be good to get our hands on some very low-leakage caps in the 0.01-0.1F range (or so) - does anyone have suggestions? 

Tantalum caps come to mind ;)

@ Mark

The two spinning the same direction is very strange, the magnets are all south or all north, are they not?

regards

Mike 8)

Lasersaber made mention of using tantalums on his EZ Spin...aha, makes sense with the low leakage !

Yep Centraflow, the 'trick' as it were is mentioned in the description box over at YouTube.
The 2nd is an unpowered rotor - all magnets are North outward
The 3rd is an unpowered rotor - all magnets are South outward
It was surprising that it worked though, thinking it all had to be the same as regular mechanical cogs.
:)

Hi Steve,

Unfortunately, leakage current for supercaps are usually not readily given in data sheets, exception are perhaps the biggest manufacturers.  This is my conclusion after some search, and I have also been aware of tantalum caps being among the low leakage types.  Here is some info on capacitor properties, found useful in the search:
http://www.delta.dk/imported/images/DELTA_Web/documents/Innovation/capacitors-for-energy-harvesting.pdf 

Yet another useful collection on the self discharge property of different capacitor types is here:
http://www.robotroom.com/Capacitor-Self-Discharge-2.html

Sorry that I cannot really offer a particular type, with links, so far I found for instance AVX capacitors with leakage current between 5 to 10 uA (page 6 in http://www.avx.com/docs/catalogs/bestcap.pdf )   This 5-10 uA value sounds still high though.
High quality electrolytic capacitors for demanding audio or measuring equipment surely exist with low leakage current, though their capacitance range may not go up to 0.1 F and this can be true for tantalum types too.

Mark, thanks for the answers. A curiousity question: I wonder whether there were four - four magnets on the two unpowered rotors or maybe more?

Gyula

Yep, 4 and 4.
The powered one had 2 magnets, with 2 transistors glued for balance.

Say, I can't get the voltage doubler to work. No idea what may be wrong, but did get sidetracked with the LED output on the driven motor. The 1000uF cap will go to 0.672V on a drive voltage to the motor of 0.9V. So instead of flashing the LED, it seems a good idea to use that output anyway.
I decided to try a blocking oscillator circuit from it and, although an LED can flash readily, it doesn't seem to be doing much when returned to the driving capacitor. Fun though and the oscillator runs to a lower drive voltage than the flashes do from the original LED.

Hi Mark,

On the voltage doubler: a simple test would be as follows.
In the same setup where you find to get 0.672V in the 1000uF cap, just reverse the same diode and the cap's polarity and measure the induced voltage in the cap then. It is is possible that due to the all N or all S inducing magnets the induced voltage's AC polarity is uniform and does not change, so the doubler has nothing to rectify on the other half wave, this is where a scope would serve you fine.

Gyula

I think you're right.
For reference, the doubler was built with 1uF electrolytics and 1N4148 diodes.
Will take a good look at the flashable scope today.

Great news, the 'dancing' toys have arrived, so full remote rotor building will be underway later today.
The magnets from the toys will go on rotors as single and doubled up, 2 and 4 and probably a larger rotor with more coils as pickup coils.

Great news Part Two. Was talking with my wife about the original Lasersaber EZ Spin and how it uses sapphire bearings.
She said "i've got a bag of uncut sapphires"
My answer "who da what now ?" and my face ->  :o
They were a bag lot on Ebay some years ago, tiny rough pieces and her idea was to try and cut them into something resembling jewellery pieces. They'll be experimented with, that's for certain.  

Mark,
  Glad the dancing toys arrived!  all those pre-wound coils, a good deal...

  Had an idea about the generator - instead of trying to build a gen from scratch, how about buying a small, hopefully high-efficiency DC generator?  pre-built...  A DC motor might do
the trick also, driven to produce a voltage.
  The tiny DC generator could be driven perhaps by a coupled rotor, or the "main" rotor;
using the bearings in the DC gen ...

Now there's a thought, high output current from something like a 6mm coreless motor.
It reminds of wind generator experiments, where small motor size needn't result in similarly small output current for that size.
A lot being about the internal windings.
These EZ Spin type motors do have some torque, so, along with the fact that they use less power when running slower, the friction side to it may be more than made up for with the current generated.
Superb, will try that !
I'd mount the drive rotor on the 6mm motor shaft. The remote ones need to have a fine magnetic lock. It's not too difficult to achieve that lock, but torque would be next to zero.
Have got a few 4mm and 6mm, on R/C aircraft. but am sure 1 or 2 are around from a few years ago when I was building from scratch.

Hello, my pretties !

Dear Mark.

I spent the last couple of years of my career with the supply industry, remember MANWEB ??  In the " Test laboratory ", below us was the Meter test dep't. All the old disc type electricity meters had very nice, large spring loaded Jewelled bearings.

I am suggesting this as an alternative to making your own, as the grinding and polishing will be quite tedious. I shall have to look in the back room of the workshop, I know I rescued ( Liberated ) a few !!   ;)   If I come across them I will let you know and send a couple.
The weather here at the moment is COLD I am in no mood for a foray, but when things improve I will have a good look for you.

Cheers Grum.

That would be truly appreciated !
Those meters spun freely for many years, what a great bearing set up.
Manchester West Electricity Board ?
We had one on our house until last year...told the electric company we didn't want a Smart meter and even explained in writing that I have electrosensitivity issues (many poo poo it, but i feel queasy around strong electromagnetics, neodymiums, wi-fi, get static shocks off car doors very readily and somehow sense lightning before nearby strikes lol). Also, this meter is banned in California, because of the 3W output, the limit being 1W. There were concerns and at least questions.
Anyway, my wife or I are in the house 24/7 nearly all the time. We both went out one day in the car and came back to find a new Smart meter on the wall !
Sleep was affected, is affected, it sits right outside the bedroom wall.



---------------------------------------------------

Dancing toy coil removal - a mini How-To

Many people have trouble with removing the coils on the Dollar Tree dancing items.
The problem, is that a decent glue holds the coil to the plastic base - see pic.
The wires snap readily with even careful removal. Some people therefore cut around the coils with a knife or scissors.
Today, I found that a coil broke with the usual careful removal of bending the plastics down, while lifting the coil upward. The glue is stronger !
Here's the fix.
Firstly, desolder the circuitboard and solar panel, leaving just the coil attached to the base.
The glue they are using is a version of hotglue, it melts again with a decent amount of applied heat.
If you have a hotplate or electric cooker, switch on a ring to the highest heat.
Bring the bottom of the base to around 1" (couple of centimetres) to the hotplate and wait until the plastics begin to deform.
Remove from the heat and gently bend the plastics away from the coil.
Repeat if necessary, until the glue can be seen to be melting away from the coil.
Remove the coil.

Quote from: Slider2732 on 2015.01.24, 21:11:34
That would be truly appreciated !
Those meters spun freely for many years, what a great bearing set up.
Manchester West Electricity Board ?
We had one on our house until last year...told the electric company we didn't want a Smart meter and even explained in writing that I have electrosensitivity issues (many poo poo it, but i feel queasy around strong electromagnetics, neodymiums, wi-fi, get static shocks off car doors very readily and somehow sense lightning before nearby strikes lol). Also, this meter is banned in California, because of the 3W output, the limit being 1W. There were concerns and at least questions.
Anyway, my wife or I are in the house 24/7 nearly all the time. We both went out one day in the car and came back to find a new Smart meter on the wall !
Sleep was affected, is affected, it sits right outside the bedroom wall.

Dear Mark.

Merseyside And North Wales Electricity Board. The amalgamation of the Mersey Power company with the North Wales Power company in 1948.

Our department head had been in the employ of both the Penrhyn Slate Quarry electricity generation company and the fledgling N W Power company, he had some stories to tell, that's for sure !!

OOOOH, Smart meters, I am dead against them. At the moment I have been changing to suppliers that do not offer them.  The thought that your usage can be monitored 24/7 is surely an invasion of privacy ?? Are you able to get your supplier to remove it ?

Cheers Grum.

Thanks for the company name. One of those things seen all over the place and never knew what it stood for.

The supplier won't remove it. It's all that mess of 'implied consent' because people don't explicitly say they don't want one.
Well, we did, by letter and then they must have stalked us to know when both would be out !
It's sad that a 'trick' to stopping them knowing habits is to use things at the wrong times of the day. Switch the cooker on at 3am. use solar on different items, switch everything off for an hour as though out  etc. Why bother ? well privacy yes, but I don't want habits known, exactly when people are out or away etc, for just such reasons as the meter swap episode.
Mind you, while slightly off topic anyway, here's how you mess up stores and their records of what you buy -
Works if you have neighbours that you get along with well. Both may spend $200 on shopping. Swap lists sometimes, shop for each other and swap goods on return. Their items go on your card, therefore database trackers are messed up. Same thing for browsing...i've no doubt certain people think I have a vast knowledge of curtains and are waiting to intercept pictures of the new couch that doesn't exist LOL
In my opinion, such things will become important when health prying and the Internet of Things really begins.
4th Amendment and i'm English !
That's why technology needs to be Open Source, anything online connected especially and communications allowed to be securely encrypted when persons are not suspected of terrorism (I noted that Cameron wants to ban encryption in the UK)..

Quote from: Slider2732 on 2015.01.24, 17:16:48
...
For reference, the doubler was built with 1uF electrolytics and 1N4148 diodes.
Will take a good look at the flashable scope today.
...

Hi Mark,

I guess you used the 1N4148 diodes because they were at hand or you may not have more Ge diodes?  Tha latter would give you a free 0.55V extra output voltage (0,65V-0.1V=0.55V,  the approximate difference between Si and Ge at the low current levels involved)  for each half AC half wave unless one of the half waves are missing indeed due to the nature of the single pole induction.
So in your test where you got 0.672V in the 1000uF cap, the DC level could have been approximately 0.672V + 0.55V =  1.222V for a half wave.

The ultimate solution to defeat the 0.1 - 0.15V Ge diode voltage drop would be to use a syncronous rectifier, this is a controlled MOSFET switch which is made to conduct whenever the desired polarity half wave comes from the coil.  When you waded through the overunity.com thread on the 25mV Joule thief topic, you may have seen a link to a paper entitled: "Design of a low-input voltage converter for thermoelectric generator"  see the 2nd link in that forum post:

http://overunity.com/13175/25mv-joule-thief-powered-by-peltier-merely-using-our-body-heat-free-energy-247/msg351059/#msg351059  

and study Fig. 4 in Page 4 of the PDF document.  MOSFET T3 shorts diode D2 whenever the AC wave becomes positive at its gate with respect to its source, this way the voltage drop of the diode D2 is killed by the low value drain-source channel resistance of the MOSFET (R_DSon). Of course this solution demands an extra coil to control the MOSFET but maybe it pays off when you need just a 200mV extra juice in a low power setup...  :)  (thinking here to shunt at least two Ge diodes).  There are MOSFETs with low (logic level) threshold V_GS voltage specifications, this may reduce the needed number of turns for the extra control coil.  

Gyula

Well now that it has cooled down a bit here(8.30pm) I'm hitting the work shop.
I'm going to try something completely new here with the coils,and i suspect about a 3 hour build time for each coil-and we will need 6 of them. I'm going to try a C coil design,and the round 1inch x 1/4 inch disk magnets will pass through the middle of the C coils. The coils will be designed so as there is a concentrated north or south field inside the C section of the coil,and the outer circumference will carry the other pole. I'm hopping that this concentrated field will give maximum force against the magnet as it leaves one coil,and approaches the next-->but time will tell.

C field sounds like Searle...
better have some ballast around just in case [anti Gravity]  :)

maybe place it on a scale...... :D

do you guys have dancing flowers down under? [I crack up every time Mark says "dancing flowers"]

Quote from: Chet K on 2015.01.25, 12:56:08
C field sounds like Searle...
better have some ballast around just in case [anti Gravity]  :)

maybe place it on a scale...... :D

do you guys have dancing flowers down under? [I crack up every time Mark says "dancing flowers"]

Searle-bla bla bla :P

We have all sorts of fancy toys that use those small coil's.
Infact,after all this time,i just thought of something-and i had to go check right at this time--> I bought my wife these two dolphin's that rock back and forth on a pivot over one of the little coil circuits.It takes 1 x 1.5 volt battery. I bought it for her 30th birthday,as she loves dolphin's. she is now 43(i didn't tell you that),and the dolphins still rock back and forth as i type this. That is 13 years running on one 1.5 volt battery-->holly sh-t :o They have only been stopped when we shift house,but other than a few days all up,these two dolphins have been rocking back and forth for 13 years on one battery ???-as far as i know. I'm going to have to ask my wife if she has replaced the battery a few times at least,as i see no way possable that this could be right.

Edit-I just had to check this thing out,and it has a 9 volt battery in it,not a 1.5 as i thought. The wife has never changed the battery,and it turns out i bought it for her 25th bday,not her 30th as i thought-the mind is growing weaker lol.
So that makes a 17 year run time on one 9 volt battery.Its been going that long that one of the balls the dolphin's nose hits has worn right off lol.
Video to come.

Dear TinMan.

I'm smiling as I post this response.... We have a similar thing, a kiddies toy. This toy has continued to operate for 13 years on the same battery!

Perhaps 2002 was a good year for Alkaline?

Cheers Grum.

Tinman
Gravity ??
doesn't hurt to pay attention to that one, after all ....nobody really has a clue about gravity ..
and I know you have a scale [your very intriguing recent weight change vid ].

17 years on 9volts and still running...we could have saved a zillion bucs on nuclear batteries and gone to
the five and dime for the batteries to run the  mars lander...  C.C  8)

Grum
its a good thing that you purchased those batteries prior to the "new planned obsolescence protocol" !:"}.

but as many have pointed out ,a well balanced system utilizing all input to its advantage ..can run a really really long time..
        like a watch.

Quote from: Grumage on 2015.01.25, 14:10:44
Dear TinMan.

I'm smiling as I post this response.... We have a similar thing, a kiddies toy. This toy has continued to operate for 13 years on the same battery!

Perhaps 2002 was a good year for Alkaline?

Cheers Grum.

Well at the moment,I'm thinking why even bother trying to make this easy spin motor as efficient as i can,when a toy maker was making these super efficient toys 17 years ago. The efficiency of this dolphin toy makes LS easy spin look under efficient. I just checked the voltage in the old 9 volt battery in the dolphin toy,and it still has 9.32 volts in it. Now I'm wondering how much voltage(if any) would a 9 volt battery still have in it if left sitting on a shelf for 17 year's?.

At this point in time,im afraid that the wife's dolphin toy is now up for grab's in the name of reserch-->this is ofcourse after she has left for work tomorrow,and i should have it all back together by the time she gets home. O0

Oh P.S
I might add that the motion of the dolphin's is the motion of a pendulum.

@Gyula...the K170 MOSFET's would suit ?
I've always wanted to find the perfect diode, zero loss. There's also so much out there that falls just short of transistor switch on of around 0.5V. Atmospheric collections, aged crystal cells and such.
Funny thing, at very low voltages and currents, I find that 1N4148's can drop something like 0.1V and they only do their silicon stumbling block of 0.7V with more realistic amounts, such as 2V at 10mA. But, yes, am nearly out of Ge, having about 3 left.


Dancing Dolphins FTW ! as they say.
Soooo, 2 small neo's could go on the dolphin, to equal the weight again.
A remote rotor could spin by the action of the dolphin at the front or/and one at the back.
Pickup coils and a synchronous rectifier (Gyula) MOSFET system could translate that to running the dolphin on a tantalum capacitor ?!?!
If anythings vintage is 2002 or older then it may not be a black blob chip inside.
The other thought is that these older designs may be 'wrong' and that's why they got discontinued  8)
A search on Ebay brought up nothing, but does it look somewhat like the one in the pic below ?

Quote from: TinMan on 2015.01.25, 16:02:08
At this point in time,im afraid that the wife's dolphin toy is now up for grab's in the name of reserch-->this is ofcourse after she has left for work tomorrow,and i should have it all back together by the time she gets home. O0

Oh P.S
I might add that the motion of the dolphin's is the motion of a pendulum.

Dear Tinman.

My wife said laughingly, it probably wont work after your investigation !!  " If it ain't broke, don't fix it "  certainly comes to mind !!   :)

Are you going to poke around with a scope ? Probably the safest bet !!

Our OU toy ??  Is a mat with footprints on it, these pairs of feet lead to a pool at the far end. The Grandkids have loads of fun with it as each footprint they stand on makes a different sound, culminating with a splash into the pool.
I can say, with hand on heart, this toy is running on the same Battery it came with, 13 years ago.

Cheers Grum.

 ;D
"no worries dear, it's all back together and working fine, only got 2 screws left over this time"
"what do you mean this time ?, what else have you taken apart of mine ?"

Best example of battery life in something commercial for me, would be the yellow Innova 3300 DMM seen in many uploads. Bought in 2010, it gets used every day and has even been left on by mistake overnight a few times. Still going on the factory fitted 9V. At least I think it's a 9V.
Oh, also my wifes digital temperature display, is a tiny thing with an LCD that comes on when pressing a button. Uses an LR44 and she got it way before her and I met in 2005.

Quote from: Slider2732 on 2015.01.25, 18:02:50
...the K170 MOSFET's would suit ?
I've always wanted to find the perfect diode, zero loss. There's also so much out there that falls just short of transistor switch on of around 0.5V. Atmospheric collections, aged crystal cells and such.
Funny thing, at very low voltages and currents, I find that 1N4148's can drop something like 0.1V and they only do their silicon stumbling block of 0.7V with more realistic amounts, such as 2V at 10mA. But, yes, am nearly out of Ge, having about 3 left.
....

Hi Mark,

No, if you mean the 2SK170, it is a junction FET and is normally on, here you need a normally off device (like many MOSFETs) which can be controlled with a single polarity voltage with respect to its source pin.
Because you have here a low level setup (voltage is much less than 20-30V, current is much less than a few hundred mA), low power MOSFETs could be used, they have low inter-electrode capacitances too. This means that the input power needed for the control of gate-source electrodes is in the nanoWatt range. Their channel resistance when they are on is less than few Ohms, some of them are under 1 Ohm.
All such MOSFETS have the body diode between their source and drain, for the N-channel types this diode anode is the source and the cathode is the drain, thus for N-channel MOSFETs this diode is always off whenever the drain is either more positive than the source or there is no any voltage between them AND the gate-source control voltage is less than the needed threshold voltage so that the conducting channel between the drain-source is open i.e has several MegaOhm impedance.
Of course the body diode is able to conduct like any other Si diode whenever the source voltage (the anode) is more positive than the drain (the cathode), in such situation you cannot control the MOSFET to work as a syncronous rectiifier.  

Remember that these MOSFETs i.e. controlled switches (I give some links below) need a V_GS control voltage higher than at least 0.5V, preferable around 1V, taken from the induced field of a setup by pick-up coils, so using them in atmospheric collections, aged crystal cells etc which are not able to provide the threshold control voltage is quasi impossible (unless you enhance a received and wanted signal by selective passive LC circuits at a certain single frequency).

Some low power MOSFET types are here, I chose Digikey because it is easy to browse through and select:

TN0604N               http://www.digikey.com/product-detail/en/TN0604N3-G/TN0604N3-G-ND/4902372 (http://www.digikey.com/product-detail/en/TN0604N3-G/TN0604N3-G-ND/4902372)    
TN0702                 http://www.digikey.com/product-search/en?x=20&y=14&lang=en&site=us&keywords=tn0702&stock=1 (http://www.digikey.com/product-search/en?x=20&y=14&lang=en&site=us&keywords=tn0702&stock=1)  
TN2106N               http://www.digikey.com/product-search/en?mpart=TN2106N3-G&vendor=150 (http://www.digikey.com/product-search/en?mpart=TN2106N3-G&vendor=150)  
ZXMN6A08E6        http://www.digikey.com/product-detail/en/ZXMN6A08E6QTA/ZXMN6A08E6QTADICT-ND/4747690 (http://www.digikey.com/product-detail/en/ZXMN6A08E6QTA/ZXMN6A08E6QTADICT-ND/4747690)
DMN3703u-7          http://www.digikey.com/product-detail/en/DMN3730U-7/DMN3730U-7DICT-ND/2792455 (http://www.digikey.com/product-detail/en/DMN3730U-7/DMN3730U-7DICT-ND/2792455)  
FDN337NCT           http://www.digikey.com/product-detail/en/FDN337N/FDN337NCT-ND/458950 (http://www.digikey.com/product-detail/en/FDN337N/FDN337NCT-ND/458950)    
VN0606L               http://www.digikey.com/product-detail/en/VN0606L-G/VN0606L-G-ND/4902394 (http://www.digikey.com/product-detail/en/VN0606L-G/VN0606L-G-ND/4902394)  
BSS806N               http://www.digikey.com/product-detail/en/BSS806N%20H6327/BSS806N%20H6327CT-ND/3196655 (http://www.digikey.com/product-detail/en/BSS806N%20H6327/BSS806N%20H6327CT-ND/3196655)  

There are more of them of course...  :)   In the above links the data sheet is included for each type.

Possibly you have some power MOSFETs scavenged from somewhere or bought, normally they have 2 to 4 nanofarad or so input capacitance between the gate source, and also some nanofarad capacitance between the drain source, so they need more input 'juice' to charge them up and it also takes some more time to discharge them, this may or may not be a drawback in a setup.

Gyula

"Our OU toy"??  Now that's funny...  ^-^

TinMan - and Grum -- any photos of these pushing-the-envelope toys?

Then we can shop around, see if we can get our ou toys on e-bay...   :D
That would be a hoot!

Quote from: PhysicsProf on 2015.01.25, 20:52:24
"Our OU toy"??  Now that's funny...  ^-^

TinMan - and Grum -- any photos of these pushing-the-envelope toys?

Then we can shop around, see if we can get our ou toys on e-bay...   :D
That would be a hoot!

Wife will be heading off to work shortly,and the video camera is alreading charging. ;D ;D ;D ;D

Oh man.....
a family heirloom taken to the slaughter.... :o

and you do remember rule number one in the OU builders hand book

rule number 1]  NEVER TAKE APART A WORKING OU TOY !!
rule number 2]  SEE RULE NUMBER ONE...  :P

@Gyula, thanks again...learning lots here.
I had always thought that MOSFET diodes were around 1.2-1.5V and often higher. I'd certainly never considered them for anything but MOSFET work. Very interesting indeed. Will certainly peruse those datasheets. I did look at one just now and saw it's small signal transistor type of packaging. Makes me wonder about some unknowns seen on scrap circuitboards (the source for most projects parts).

@TinMan, I find it hilarious how some fellas half a world away know what you're doing, but your own wife doesn't, AND that you are posting about it, with the camera charging up right when she's there ! LOL



Have built the 2nd motor.
8 coils, 4000 ohms seemingly on the dot. Coils have read between 465ohms and 544ohms, as taken from the circuits.
It's all completely rough, but that was the idea, to see the improvements based on the rough.
The rotor is, get this, from one of my wife's old tobacco chew tins ! She was never much of a fan of cigarettes, so opted for the chew. Stored a bunch of them in a drawer in case I needed them for anything, such as Altoid tin projects.
Hotglue has been used, it allows fine movements by remelting with the soldering iron (yep that old trick). Though, Fix-All and superglue don't seem to want to stick at the moment, warmer days needed.

It did an hour on a 0.1F cap charged to 0.9V, not bad for a first run. It's really slow though, because of the ceramic magnet weight. But, at least we know that ceramics will do an hour, if you see what I mean :)

Hey Chet, let's just hope the video doesn't contain this line:
"Oh cripes, she's gonna kill me"

Wow
talk about serendipitous moments...
I was just thinking about a Frantic Brad trying to get things Back where they belong
Ball bearings and springs flying out [the usual nightmares]
and then a plea to the populous for a quick replacement _until_ he gets this one back together  ...
so I clicked onto this thread to see where he's at and there you were
with the same thought....

funny stuff this OU Toy Biz...

But knowing Brad and the seriousness of this operation,I'm sure he suited up in the white suit with the pressure mask and is working in his 1.5 atmosphere, clean .dustless and hermetically sealed safety room.

this is a very touchy and sensitive area...taking apart the wifes historical [17 year] family heirloom for OU research
wars have been started over lesser offences.

Well yeah, how do you think the Irish got jailed and then carted off to Oz  ;D
Those wives wanted their fellas at more than arms length !

Only joking, of course.



Here is the 'Spider Rotor'.
The other is too heavy, i've decided, great low speed performance but that's not the idea.
So, its spindly legged self is on test at the moment.
Will probably have to grab some N52 5mmx3mm, similar to those on the first build. This was fun however, chopping up a hard drive magnet with pliers.




Do let me know if anyone following has bandwidth issues, if on cellphones and the like for videos and i'll post pics of progress.

Well just joining all the videos together,and i managed to put it back together with no left over screw's-->which would have been preaty silly as it only has two screws in it O0

Anyway-Slider,you will never guess as to what circuit this thing has in it lol.

Will upload video,and then post here as soon as it's done.

Quote Slider
@TinMan, I find it hilarious how some fellas half a world away know what you're doing, but your own wife doesn't, AND that you are posting about it, with the camera charging up right when she's there ! LOL

It is funny when you look at it like that,but my loving wife doesn't pay much attention to this crazy stuff i do-although she gave me a mouth full of disapproval when i electrocuted my self a couple of months back. :o

After proof watching the video,i just realised that i never did the capacitor run time test lol,trying to get everything done quickly so as i could get it all back together before !she who must be obeyed! returned home.
Ah-my wife isn't that bad,I'm sure she would understand O0. I remember some years back taking apart her electric cake mixer so as i could see how the speed controller worked. It all went back together,and worked just fine-except for that scraping noise when it was running. She of course took this opportunity to go purchase a new cake mixer,but on the up side of that,i got the old one O0

Quoteexcept for that scraping noise when it was running

You trapped the family cat in it ?

Looking forward to the vid !

As a guess, I'd say a high ohm coil, run cap, pair of transistors Collector to Base, throttled heavily with a large value resistor.
Or, something completely related to this thread project, like a high ohm coil and a reed switch.

Enjoy

https://www.youtube.com/watch?v=d9jj45CJsVk

Quote from: PhysicsProf on 2015.01.25, 20:52:24
"Our OU toy"??  Now that's funny...  ^-^

TinMan - and Grum -- any photos of these pushing-the-envelope toys?

Then we can shop around, see if we can get our ou toys on e-bay...   :D
That would be a hoot!

Dear Steve.

Photos attached. It seems ASDA, now part of Wall Mart, the term " Long Life " certainly lives' up to it's name !!  :)   Both cells read slightly > 1.5  VDC, 13 years on. This toy was bought for our first Grandson who is rapidly approaching his 16 th birthday !!

Cheers Graham.

Mark, I feel we are hijacking your thread, my apologies.

 Very interesting, Tinman and Grum   O0

Courageous of you to take the toy apart while your wife was out, Tinny -- made me laugh.
VEry simple device indeed...
  but how did the cap increase in voltage with the swinging of the pendulum?  that observation of yours, I don't understand yet...

Great video Tinman  O0
Was completely surprised to see that the transistor was a regular vanilla C945.
A replication will be made, using 2 x 'dancing' coils on top of each other (likely a new thread).

Oh no problems at all here Grum, all is related to efficiency.
I'm delighted about the learning from all the great posts. Fun too :)

Steve, I think this is the circuit operation (will draw it out at some point, have just worked it out in the mind):
ETA, am editing this as I take another look at TinMan's video, because it seems that 2 wires connect to the Emitter !
Negative from battery goes through a coil winding to the Emitter of the transistor.
Positive of the battery goes to the Collector.
A connection goes from the Emitter as the 2nd coil, around to the Base.
That's the full circuit.
When the dolphin rocks, the energy that would be picked up by the Base of the transistor is induced into the first described winding, sending it back to the battery. The EB diode of the transistor works somewhat like the reed switch of the EZ Spin, completing the return circuit.
So - when a pulse is fired through the transistor, it sends the dolphin rocking and when the dolphin comes back from the pulse and the transistor is off, the energy goes into the battery. On the next swing back down, the voltage is induced in the correct direction for the Base and the transistor fires again.

Dear All.

So he, who shall remain nameless, or at least expletive deletive, got copied by the Chinese, who created an almost self running house decoration !!  :) Irony ??

Brad, how do you think the decoration might run connected to a couple of 5 F 5V super caps connected in series ?

Cheers Grum.

QuoteA replication will be made, using 2 x 'dancing' coils on top of each other (likely a new thread).

Great! you're a mile ahead of me, Mark!

   We'll see what the circuit looks like as you get it (wish Tinman could have rotated that circuit a little more..) 
Note:  From what Mark describes, it differs significantly from the Ber-dini SSG (IMO).

Mark - keep in mind while doing the replication, if you can think how any energy floating around can be re-captured, that's a good goal.  IOW, the coil fly-back, is that being recaptured?  perhaps the return swing of the pendulum also...

   I like the simple pendulum action... there is already a type of resonance there, as the period P is well defined and depends on the length L of the pendulum (from pivot point = low friction!) to the center of mass of the mass M (in this case, magnet).

  Note that the period of the motion T is simply:
P = 2*pi*sqrt{L/g}
plugging in 9.81m/sec^2 for g, we find to a good approximation:

P = 2*sqrt(L) in SI units.  So if L=1m, then the period is
P = 2 seconds!   

The period is the time for round-trip, out and back, returning to the same spot.
The shorter the "string", the shorter the Period - but it does not depend on the mass of the pendulum at all!  just depends on the length L of the pendulum.

Quote from: PhysicsProf on 2015.01.26, 17:48:29
Very interesting, Tinman and Grum   O0

Courageous of you to take the toy apart while your wife was out, Tinny -- made me laugh.
VEry simple device indeed...
  but how did the cap increase in voltage with the swinging of the pendulum?  that observation of yours, I don't understand yet...

Im not sure Steve, but I think its due to capacitive or inductive coupling between the two coils being in the bifilar situation.

Quote from: Grumage on 2015.01.26, 19:44:42
Dear All.

So he, who shall remain nameless, or at least expletive deletive, got copied by the Chinese, who created an almost self running house decoration !!  :) Irony ??

Brad, how do you think the decoration might run connected to a couple of 5 F 5V super caps connected in series ?

Cheers Grum.

I did run it on the 2200uf cap with a starting voltage of 9.4 volts-same as the battery. But it only ran for about 4 minutes. I thought it would have done much better than this.
Im going to have to put my hand in the fire here , and ask my wife if I can carry out some test on it, as there is something I want to try with it. 2.5F @ 9 volts would give a good run time I would think Grum

To quote Steve:

Quoteit does not depend on the mass of the pendulum at all!  just depends on the length L of the pendulum.

I foresee an excellent use for the heavier ceramic magnets !
To explain, usually nowadays, the 'dancing' toys feature 5mm x 3mm neodymiums. Previously they had larger, heavier ceramics of rough equivalence magnetically. The store manager said on Saturday "last year we had 10 boxes of these devils and they all went in a day"...and there it is. Looking underneath, we see 2013 stamped and so they'll be to the ceramic spec.
For the dolphin, the flyback goes through the circuit backwards, enabled by the diode inside the transistor (do correct me if wrong anyone). The very movement of forward and back creates that effect by default, with no extra components required. It's very similar to the EZ Spin and what we know of its ability to return energy.

A mod to the dolphin, would be pickup coils up and down the travel distance, at least either side in the middle, where speed is greatest !


To quote Brad:

QuoteI did run it on the 2200uf cap with a starting voltage of 9.4 volts-same as the battery. But it only ran for about 4 minutes.

Aha, good info and thanks for doing that. The first EZ Spin that I made did 1 1/2 minutes on a 1000uF charged to 1V. it's somewhat similar and i'm really wanting to try the 'dancing' coils idea for the extra resistance.

Yeah, Brad, do you recall the actual circuit connections ?
I'll wing it, as I think I know how it's wired, but would like your input :)


---------------------------------------
Here we have the MicroEZ.
It's as small as an 8 coiler can be, using these coils.
It was tried as a build, to see if there needed to be a clear 1 coil space between coils, to allow for the phantom pole.
As it happens, nope, runs ok lol
Not very quick again and i'm really wondering now whether that first 4 coil build is the way forward.
Many coils = sipping of voltage/current but slow speed. 1 or 2 coils = high speed, fast power usage....are 4 coils and 2 magnets the Goldilocks zone ?





Edit: Fixed some spacings between coils.
There were panel gaps that British Leyland would have been proud of  C.C

@tinman and @grum. Our wives should never meet!

Quote from: Slider2732 on 2015.01.27, 03:13:19
To quote Steve:I foresee an excellent use for the heavier ceramic magnets !
To explain, usually nowadays, the 'dancing' toys feature 5mm x 3mm neodymiums. Previously they had larger, heavier ceramics of rough equivalence magnetically. The store manager said on Saturday "last year we had 10 boxes of these devils and they all went in a day"...and there it is. Looking underneath, we see 2013 stamped and so they'll be to the ceramic spec.
For the dolphin, the flyback goes through the circuit backwards, enabled by the diode inside the transistor (do correct me if wrong anyone). The very movement of forward and back creates that effect by default, with no extra components required. It's very similar to the EZ Spin and what we know of its ability to return energy.

A mod to the dolphin, would be pickup coils up and down the travel distance, at least either side in the middle, where speed is greatest !


To quote Brad:Aha, good info and thanks for doing that. The first EZ Spin that I made did 1 1/2 minutes on a 1000uF charged to 1V. it's somewhat similar and i'm really wanting to try the 'dancing' coils idea for the extra resistance.

Yeah, Brad, do you recall the actual circuit connections ?
I'll wing it, as I think I know how it's wired, but would like your input :)


---------------------------------------
Here we have the MicroEZ.
It's as small as an 8 coiler can be, using these coils.
It was tried as a build, to see if there needed to be a clear 1 coil space between coils, to allow for the phantom pole.
As it happens, nope, runs ok lol
Not very quick again and i'm really wondering now whether that first 4 coil build is the way forward.
Many coils = sipping of voltage/current but slow speed. 1 or 2 coils = high speed, fast power usage....are 4 coils and 2 magnets the Goldilocks zone ?

Edit: Fixed some spacings between coils.
There were panel gaps that British Leyland would have been proud of  C.C

It's just the cool joule circuit Mark, with both coils wound on the same former. What you didnt see was the small ferite rod that fited into the small hole in the coil former
So start of one coil to base, and end to emitter.start of the second coil to battery +, and end of second coil to collector. Negative of battery to emitter.

Thanks Brad...thought so, but also the Positive red wire seemed to have a thin wire connection within the solder and that confused. It is probably a loose strand from that battery connection
I have my best version of your Cool Joule still running. It's right there on the table.
Took some video a few days ago with various projects as an update, but didn't post it, the CJ was in that....kinda think I should have posted it, to link with the Dolphin.
Speaking of wives, the table you see in these vids was 'new' to us last year and Julie wanted it kept clean and nice, with a vase and flowers. Errrrrm *cough*

Question - after watching a video by Erfinder and considering the lack of speed etc of this latest MicroEZ build.
Can I short every other coil and gain anything ?
Edit - and the update. It's pretty funny, I linked every other one, therefore simply bypassing every other one. Spun it up by hand, with 0.855V on the cap and it slowed immediately to a crawl...and kept on going ! About 1 RPM.
The 4 will now be tried as paralleled gen coils instead.

Quote from: Slider2732 on 2015.01.27, 17:14:30
Question - after watching a video by Erfinder and considering the lack of speed etc of this latest MicroEZ build.
Can I short every other coil and gain anything ?

I can't comment on the shorted coil question, however, the arrangement of E's coils is important to taking advantage of Lenz' law. I can't see a reason why it can't be incorporated into this setup.
Bob

Thanks Bob.
Maybe there are clues in the latest set of experiments this morning, because it's odd to me and at least needs clarification from others.

All tests have been with approx 0.9V on the 0.1F capacitor, which has been connected throughout.
The MicroEZ had 4 magnets, 8 coils and ran slowly.
Shorting every other coil to basically remove every other resistance, resulted in 1 RPM running.
Well, there were now 4 magnets and 4 coils, fair enough.
2 magnets were taken off - same thing, 1 RPM, yet it was exactly the same configuration as the 29hr original build !
I decided that it must be some kind of friction issue, because without the reed it still slowed itself rapidly to a stop.
Adding graphite to the needle still resulted in a rapid slow and stop with no reed switch.
I disconnected 1 coil, so the path was now open circuit and the rotor spun around very freely again, to a very gradual free running stop.

In effect then, perhaps, these air cored coils are showing immense Lenz force, by simply being bypassed - the total opposite of what 'E' is showing ?
Another question, and perhaps a silly one. If I connected the 4x shorted coils backwards, would anything change ?

The set up of the MicroEZ, is now as follows.
4 compass point coils untouched, regular connections.
SE coil disconnected completely.
NE coil reverse connected.
NW coil shorted to either side coil
SW coil shorted to either side coil.

The rotor pulls to one side, but runs at twice the speed as it did when all 4 minor compass point coils were shorted.
Still very slow with rotation, but somehow has improved in this arrangement.
Is this related to the way that a magnet will drop slowly down a copper tube ?

Here's a brief video to hopefully explain.





Update: With all 4 of the minor compass point coils completely disconnected, the motor runs freely and at similar speeds to the original build.

After watching LS' s video a few times now on his latest EZ spin motor, dose it seem to anyone else here that the efficiency is way to high?. Im having trouble believing that that tiny cap could run that motor at that speed for that long :o
Dose anyone know the value of the cap he is using?

  IIRC, he's using a 5F super-cap in the latest EZ spin video.

Note that Mark used a 1F cap for his 25+hour run.

Quote from: PhysicsProf on 2015.01.28, 02:16:35
  IIRC, he's using a 5F super-cap in the latest EZ spin video.

Note that Mark used a 1F cap for his 25+hour run.

Yes, but what size cap did LS use-dose anyone know?
It looks very small for the run time he was getting throughout the video.

Seem to recall he used a pair of 100uF tantalums in one extended run.
It was within the comments to one of the videos.
I don't think tantalums are made at much larger capacities than that and are definitely where to head for low losses during run time.