Levitating Motor

[Ken I]

I bet I could cut them on the waterjet here. I'd cut them for the heck of it, but shipping to Cape Town might be prohibitive.

Thanks for the offer and a great big slap upside my head - I know plenty of customers with waterjet who might oblige - why didn't I think of that.

Ken

 

[rhitee93]

Cutting glass is a bit tricky on one. The piercing tends to cause a lot of chipping on the surface so you have to use a long lead in before you get to the cut line so the chipping doesn't end up on your part.

Optionally, you can start the jet off the edge of the glass and move into the material with the abrasive flowing. That works well, but requires creative programming.

You can also by 30mm disks here http://www.edmundoptics.com/

 

[Herbiev]

Ken. One KP for the brilliant "scrounging" ideas. Being on a part pension here in Oz really restricts the money available to the hobby. Little freebie tips like these really do stretch the dollar. Love to hear any other tips from members :bow:

 

[Ken I]

Progress.....

I have extended the shafts and I'm working on the base, in the meantime I have a few observations - maybe a member can chime in with some enlightenment.

The motor seems to run better in one direction - if you turn the rotor around it still goes the same way but one way arround is beter than the other. Doesn't make any sense.

I reversed the field magnet - which reverses the motor - still the same - here there is also a difference which I suspect is constructive / destructive interference with the levitating poles.
That does make sense. But the difference on reversing the rotor remains.

The output from the photocells seems to be temperature dependent, they seem to give more output warm.

Windings - I was interrupted and lost count during winding - as a consequence my 2 windings are not the same (37 vs 45 Ohms) - this gives the motor the collywobbles - unrelated to ballance - due to uneven thrusting by the windings - Lesson - make sure you count the number of turns accurately - I'm going to rig up a counter if and when I redo it. I don't see how the uneven winding has anything to do with the difference in performance when the rotor is turned around.

I have run it up to 200 rpm under a 50W downlighter the collywobbles settle down - I guess in direct sunlight its going to get to about 1000 rpm.

Contamination - Iron filings (our shops are full of them) getting into the laquer, pressed into or even in the MDF also cause ballance-like issues.

The finished rotor weighs 241g and from a ballance point of view is sensetive to 0.01g, - I can measure to only 0.1g and using modeling clay a fraction of this induces it to turn - statically.

I'm still fiddling with the ballance issues - again some curious observations - I'm getting different results with the field magnet in or out (if you ballance with the magnet in - it wants to rotate under the bench lights - a winding disconnect point might have been a good idea) - here I suspect contamination but I'll be damned if I can find it.


Ken

 

[steamer]

You never cease to amaze me Ken! That's cool!

Dave

 

[Ken I]

Well here it is almost finished (apart from a bit of polishing) I managed the glass disks by sandwiching the glass see..

http://www.homemodelenginemachinist.com/index.php?topic=18793.msg194667#msg194667

Here it is running on the bar at 20 rpm under a 20W downlighter

the blur is because of the camera / low light level.

Here it is stopped.

I used stops at both ends to prevent das dumpfkoppenrubbernekkenundsightseeren from knocking it out of the cradle - it runs fine with only one.

And a video of it running.

http://www.youtube.com/my_videos_edit?ns=1&feature=vm&video_id=fplVwOuAv38

The speed it was turning (when I grabbed it) was the stable speed it was running at the time.


The speed has been going down as ambient light levels dropped (I hope that's why its been slowing down) and its now dark outside and ticking over at a turgid 15 rpm which is nice as a conversation piece for the bar.

At this speed its not self starting which I suspect is a ballance issue - the torque is so very low - with a brighter light it self starts.

I haven't tried it in direct sunlight yet it will probably go ape.

Give me a couple of days to tidy up the drawings & build notes and I will post it under the downloads section.

Ken

Edit - P.S. Apart from this project I have been desinging and building some powerful magnetic grippers for robotic use.

These powerful magnets are a curse - everything that comes in contact becomes magnetic (yeah sure - I'll demagnetize them later), they can jump together violently enough to shatter against each other - and really big magnets can do some serious damage to your digits if they get in between a pair.

Tools leap off the bench to attack you or whatever it is you are working on - my normally docile scriber leapt off the bench and impaled itself in my finger. In a jealous rage my pliers took a snap at my project.

My advice - clear the workspace and keep it as clear as possible for the duration (do as I say - not as I do !).

 

[doubletop]

Ken

This an absolutely wonderful insight into a subject a lot of us probably had no idea about, and like me are now on the lookout for the various components. I'd found the source of neo magnets in NZ only the other week for another application. No doubt there will be a few more of these motors presented here over time.

I now realise that the levitation is bought about by the magnets in the end plates and at the rotor tips and the coils only provide the rotational drive with commutation provide by the ambient light on each of the solar cells. I had thought that the magnets at the rotor tips were bearings and the armature would rise off the stand as the speed increased

I like your use of acrylic to add to the effect of "look no hands, hidden wires or magic". Maybe an acrilic rotor is on the cards? First thought on seeing it was it needs some leds embedded and would some magnets on the rotor with an induction coil in the base. But maybe not.

Dare I say the second end stop does need to be removed it detracts from the effect. However I can see it's there to deal with the effects of Kinderfingerpoken

Nice job

Pete

 

[George_Race]

Hi Ken, very nice project description and build. Your final design really does run great!
I have been a fan of these kind of motors for some time, and started putting one together just this week.

Gosh, do you think we can convert this group of engine builders to Mendocino builders?

These little devices are sure fascinating to watch run!

 

[Ken I]

Pete,
I've come to the same conclusion - having two end stops detracts from the effect - but I can flip one down by hand easilly enough - I will post another clip of it running with one.

Yes a totally acryllic rotor is possible and would look nice.

My caution to anyone building this is to make the rotor complete and then do a lash up to determine the centre height - it will vary depending on the mass of the rotor and the strength of the magnets etc. etc.

I could have targeted a faster running motor using more turns of thicker wire - but I was deliberately aiming for "slow".

George, thanks I will be following your thread.

Ken

 

[Ken I]

Here it is running at 50 rpm under ambient lighting (no illumination lights turned on).

The Windhoek Draught is to amuse Arnoldb - and I was thirsty.

This is to show it running with only one support end (as mentioned in prior posts two spoils the effect - but are there because I have grandchildren). The screws are loose enough to manually fold them up or down.

I took it outside in direct sunlight (low angle winter sun) and it spun up to about 1200 rpm before the collywobbles caused the rotor to strike the field magnets (I've only got ±1mm clearance) and it would have run over the fields and far away has I not had my hands in place for such an eventuality.

I think a couple more milimeters clearance would not hurt.

Ah well - it ticks over nice and slowly on my bar - day or night and that was the intention so I'm happy - Calling this project finished (for now).

(Sorry the vid's at the bottom but for some reason if I place it further up the post then everything below it dissapers ?)

Ken

 

[Ramon]

Ken,

I have just caught up with this thread and would like to say what a super piece of 'engineering' you have created.

I confess that to date this sort of project is not exactly my cup of tea prefering I guess those of a much more conventional nature. However, this has simply 'bowled me over' and though I'm not sure I will ever make one your description of the build and the way in which it works certainly makes me 'want to' I await the drawings etc then with great anticipation - who know's? I/C could go on hold for a bit

I thought the way you overcome the glass cutting problem was truly brilliant. :bow:

Thanks for sharing it with us - good on you Thm: Thm:

Regards - Ramon

 

[doubletop]

Ken

Moving the second end stop is just what it needed.

Another question. Do the support magnets adjacent to the end stop need to be slightly lower than the opposite set so the rotor tends to slide down hill towards the stop or is it the effect of the armature and field magnet that pushes it towards the end stop? (just had a flashback, Lenz's law?)

Great job

Pete

 

[Ken I]

Pete,
The end stop magnets were a bad idea - they're gone.

The idea was by attracting the ball in the end it would stabilise any ballance issues - it didn't work for that purpose.

With them repulsive from the rotor magnet, (it still attracted the ball but very lightly) when you folded one stop down it pushed the rotor off its magnetic perch.

With it in attractive mode (strongly attractive to both ball and rotor magnet) it increaced friction spoiling the ultra slow low light performance and still did nothing for ballance issues at the free end.

It was just one of those experiments along the way that didn't pan out.

So ignore them. (Well spotted by the way.)

With the rotor magnets directly over the support magnets, the rotor is perfectly suspended - but unstable - it will slip from its perch to the left or the right - by placing the end stop about 1mm away the rotor is effectively falling down the magnetic hill towards the stop and is thus stable.

Literally a micron off is stable but a puff of wind can dislodge it, so more offset increaces the force against the stop - none of this is critical - if the distance between the rotor and support magnets is not the same (within reason) there is still a stable point and as long as your stop is just off from that it is fine.

I got the bright idea that I could also suspend this horizontal float with magnets - thus completely levitating the rotor - after a hell of a lot of experiments I dicovered its not possible. Turns out Earnshaw's theorem proved that Maxwell's equations make it imposible to achieve static magnetic levitation in all planes. Bugger.

It could be controlled electronically but that's not really in the spirit of this build.

Like I say build notes to follow - unfortunately my son has broken his collar bone in a mountain bike downhill crash at the weekend and I will have to fill in for him which is going to drastically reduce my hobby time for a few weeks.

Ken

 

[Ken I]

O.K. for those of you who may be interested I have finished all the documentation and drawings.

I've loaded it in the downloads section as a *.zip file containing the drawings as *.dwg, *.dxf & a single sheet *.pdf plus a *.doc file of the build notes.

Have fun but they'll cost you a KP point - and that's free.

Regards,
Ken

 

[golddustpeak]

Ken,
Interesting project and a very well done completion.

Just out of interest how did you end up choosing to use a 45 ohm (150 turn) winding?

Neil

 

[Ken I]

Just out of interest how did you end up choosing to use a 45 ohm (150 turn) winding?

By a process of calculation - which I got wrong, guesswork, intuition and dumb luck.

Max. output occurs at matching impeadance ie 3V @ 100mA = 30 Ohms (but I went and calculated 90) for two cells in paralell halved it to 45 - so I got that wrong from the getgo.

Next what gauge of wire - as that will influence the number of turns - clearly more turns improves youre Ampere Turns - but more turns increaces the back emf - so where is the optimum - pretty much anyone's guess given the imperfect field set up of this motor.

Why 36swg ? well it will handle 200mA and I had a roll and that equated to 45 Ohm at 150 turns.

I'm afraid it was a classic case of confirmation bias - I intuited about 150 turns and pretty much found what I was looking for by accepting answers that agreed with that - dumb - and a real problem for serious research - but in this case no harm done.

So ultimately I would have to write it off as a lucky guess.

Ken

 

[golddustpeak]

Max. output occurs at matching impeadance ie 3V @ 100mA = 30 Ohms (but I went and calculated 90) for two cells in paralell halved it to 45 - so I got that wrong from the getgo.

That's where my confusion came in.
I was figuring 2, 3v 100mA cells in parallel would yield 200mA @ 3v and require about 15 ohms, or even 2, 2.7v 68mA cells in parallel working with 19 ohms.

So with your 45 ohm loaded 3v cells you would be at about 67mA.

I'm going to be using 2, 4.7v 81mA cells in parallel and I am just struggling to decide on a number of turns to use as I've done some research and found no hard and fast rule.

Following your successful example (and other examples I've seen) I'm wondering about using say 75-80 ohms or even more heavily loaded at 58 ohms.

Looking at wire sizes and my choices in the shop I will be using #36 as well, as it seems to have a nice resistance per turn for a project like this.

Neil

 

[gld]

Tried to download the zip file..All I get is a 53 k file that Windows XP says is corrupted.. ???

Gary

 

[Ken I]

I dont know why that happens. :-X
try again tomorrow.
ken

 

[gld]

Ken

Retried on my laptop running Windows Vista and it worked perefect....

Very nice work and Thank You for sharing..it will be a fun project ..1 more Karma added.

Gary