EZ Spin motor - homebuild

[Slider2732]

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'.

[PhysicsProf]

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

[gyula]

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

[Slider2732]

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  

[PhysicsProf]

Some good ideas to test out, Mark 

"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

[Slider2732]

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.

[gyula]

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.   

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

[Grumage]

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.

[gyula]

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

[Grumage]

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.

[Slider2732]

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

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.

[gyula]

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

[PhysicsProf]

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..   

[gyula]

Hi Steve,

Thanks, I somehow thought Mark is in the UK...     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

[Grumage]

I thought south Cheshire !! 

Cheers Grum.

[Slider2732]

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).

[Slider2732]

Still going !  
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 ?

[Grumage]

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.

[PhysicsProf]

Still 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.

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!)

[Slider2732]

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

@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.

[PhysicsProf]

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!"

[Slider2732]

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.

[PhysicsProf]

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

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?

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.

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. 

[Slider2732]

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

[Slider2732]

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.