DIY Brushless Motor?

I was out in the workshop last night, and came across an old motorcycle alternator in one of my junk boxes. I've no idea what it's off, some small Japanese bike I think.



Anyway, looking at it I was struck by the resemblance to these RC motors we're playing with. The magnets in the rotor are probably rubbish, but they are easy to replace with multiple neodymium magnets from eBay. Likewise, the stator windings are way too small for serious power, but there's plenty of room to rewind the stator with something better. The stator has twelve poles, so, providing that I can find some suitable magnets to fit the rotor I should be able to make a decent 12 pole, 14 magnet motor, just like a big RC one.

This motor has a 3 1/2" diameter stator, so is around twice the diameter of the stator in the 2.8kW Towerpro motors I've been playing with. As torque is directly proportional to stator diameter, a motor built from this core would have around double the torque I'd guess. I think it should be possible to wind a low Kv, high Kt, motor using these parts, that will be optimised for high voltage operation, so I might well end up getting even more torque from it.

Someone else must have spotted the potential to build motors from these things, as they are so obviously made for the job. I've googled about a bit and not come up with anything much, which is a bit surprising, as both the mechanical and electrical work needed for a conversion seems pretty simple. As these parts cost me nothing, I reckon I may be able to build a decent 5kW or so motor just for the cost of some wire, magnets and a couple of bearings.

Jeremy

Looks like a good candidate for experimentation, Jeremy.

Did you see this thread? http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=6584

Make sure the laminations are thin. A lot of the RC stuff is 0.2-0.5mm per lam. Thicker ones saturate sooner from my limited understanding. Looks like a ton of room for fill in those things though! Should be able to make a wicked re-wind.

Interesting.... I reckon they all have facility for sprockets, like the harley I found on ebay for $25.

Yes, as you point out the magnets are not very good. The lamination thickness is probably a bit marginal too. These things are not designed for efficiency. Replacing the magnets might turn it into something useful if you can find magnets that fit.

I probably need to go and do a bit more research before buying stuff to experiment with this, but it looks interesting, all the same.

A quick look on eBay has found some magnets that look as if they'll fit OK, at a very reasonable price, too. I need to find out more about the iron loss problem though, as this thing seems to have fairly thick (around 0.5mm, 0.020in) laminations.

There is masses of room for windings though, which means a nice low Kv wind should be possible.

It's got me thinking about looking around for a nice large stator, though. My guess is that there are some other potential donor parts around. For the application I have in mind (3000rpm max, around 10kW) something around the size of a big vehicle alternator might make a suitable donor.

Jeremy

There's a bunch diy wind mill folks out there converting alternators for low speed...

A few folks have redesigned the rotors (inrunner) with new magnets for less cogging...etc..

Ring magnets can be bought from KJ magnetics with alternating poles.

anyway....I got some good leads on this topic going down the windmill path..

Sounds like fun!!

I used to want to play with this stuff called metglas...It was supposed to revolutionized the motor world...
but it's probably too exotic. But the mu values are amazingly high...

Then it can go back into the crank casing like in this link http://www.planetetrial.com/articles.php?lng=en&pg=10583 instead of a monster RC motor 8) There is a small youtube vid of this bike doing a section in a trial, looking good its just a shame that the audio of the vid you can here the gas bikes

edit:
Just re-read the web page and it looks like its not a vid of the e-tbike, its a vid of the guy who is helping develop it,

i have a delphi alternator converted as the load of a dyno tester.

this picture is the dyno i am building and now it's working.

http://www.rcgroups.com/forums/attachment.php?attachmentid=2466051

what i was going to say is that....this alternator has stator punch thickness of ~1mm (yes, that's one mm) and the core loss is huge (i put 3A on the field winding and almost could not rotate the rotor with bare hand; the stator winding is left open). i guess i could drive this alternator with a RC motor controller but never bothered.

Hey Jeremy,

I am in the process of re-magnetizing a motor and found a source.
I actually talked to "George" today, (great guy) He can get anything
special you need. They carry flat as well as arced "neo" magnets specially
made for motors. All you need to do is specify an overall inner and outer
diameter for your application along with the desired number of magnet
segments to complete your circle and he will have them made. You need
to specify a desired strength as well. He claimed only about a 10% upcharge
from a comperable "stocked" item is needed for something custom
To be honest, I haven't done a lot of research for sources... but I'll be placing
an order tomorrow with them...

http://www.supermagnetman.net/index.php?cPath=37

Regards, and thanks for all of your pioneering efforts :wink:
Kevin

Thanks for that. One slight problem with trying to use a bike alternator is the depth of magnets that are needed. The scrap one I found needs magnets that are about 10mm thick, which makes for a heavy rotor. One way around this problem might be to hunt around some bike scrapyards and find some "mix and match" stators and rotors from different bikes that might allow a thinner magnet to be used.

I've found a supplier on eBay here in the UK that looks to have magnets that I think might fit, I just need to machine off the end of the rotor and remove the existing magnet assembly to check the size before going further.

Meantime I need to get back to fitting hall sensors to my newly modded Towerpro motor, now wired in wye, rather than delta, to see how it compares to standard.

Too many projects.............. So little time...............

Jeremy

Once I get the controller lag situtation solved, I'm going to be following your lead, and re-combining my windings into a Wye pattern. I'm glad I have adjustable sensor locations to compensate for the (30deg?) timing change in Wye.

Since that stator isn't any good with the windings it has on it now, how about cutting them off of a leg so you can check if the laminations are thin enough to be worth messing with.

georgeycc said:

i have a delphi alternator converted as the load of a dyno tester.

this picture is the dyno i am building and now it's working.

http://www.rcgroups.com/forums/attachment.php?attachmentid=2466051

what i was going to say is that....this alternator has stator punch thickness of ~1mm (yes, that's one mm) and the core loss is huge (i put 3A on the field winding and almost could not rotate the rotor with bare hand; the stator winding is left open). i guess i could drive this alternator with a RC motor controller but never bothered.

Click to expand...

Are you measuring mechanical torque on your dyno or just the electrical output? I setup the exact same dyno but when I tried to measure torque there was too much vibration for my scale to work properly.

I am building a push trailer with an alternator as a motor in this thread Yes, the laminations are quite thick, and the field winding will consume some power, but it was free as well as being an interesting challenge!

hi Paul,
i measure 4 numbers - battery V, battery I, output mechanical torque, and output mechanical speed.
i checked your post and i also ran into the same kind of problem with my earlier setup. the interaction between the motor and the load was messing up the torque readings. the secret of this 2nd setup is that it uses 2 couplings with one extension shaft to isolate the interacting force b/t the motor and the load (but not isolating the torque which you need). my current setup could measure torque reliably with ~1% fluctuating, that's pretty good in terms of SNR already.
btw, if you have a japanese car, check out the denso alternator, their stator lamination is a lot thinner at ~.3mm i eyeballed. i guess i should've got the denso alternator when i was looking for this delphi on ebay.

georgeycc said:

btw, if you have a japanese car, check out the denso alternator, their stator lamination is a lot thinner at ~.3mm i eyeballed. i guess i should've got the denso alternator when i was looking for this delphi on ebay.

Click to expand...

Delphi/Mopar/FoMoCo = The very lowest quality you can build, and still have something function.

Denso/Calsonic/NipponDenso = The highest quality parts available. Often the highest quality part available at any price.

Whoa! this was on my back-burner list of projects to do.

Jeremy Harris said:

Too many projects.............. So little time...............

Click to expand...

I have so many ideas rattling around in me brain-bucket but not enough time nor money. Most of the projects are on hold until I run into materials and solutions that I need, until then, I'm just doing what I can with what I have at the moment and playing catch-up with more pressing issues such as:

http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=3035&p=150760&sid=21ccaaf79be85039c8c87dda986f1593#p150760

Glad to see Jeremy has kicked off this motorcycle Alter-Motor project *tail wagging*

J