switched / variable reluctance motor conversion tests

[DanG]

This motor is a good flux bottle, I was trying not to destroy the original configuration by grinding or replacing windings, just simply adding neo's.

What I've seen so far is a good generator that gets extremly hard to turn when even one phase is applied to load. Possibly could be used for Hydro applications with right impeller and a way to dissipate heat, or a temporary charging arrangement attached to small internal combustion engine.

Variable Reluctance motor conversion first diary entry here:

http://www.fieldlines.com/story/2005/7/14/42640/9511

Motor is rated 120V and 2.5amps, intermittant duty @ 10 minutes on - 20 minutes off, no forced air cooling provided as-is and no open grills on cast aluminum motor bells.

N38 magnets are 1" by 1/2" by 1/4" (17 pounds pull force) for eight poles (16 mags) on a 'gear toothed' laminated rotor.

Shown in the first diary was what every-other coil in series output was, but canceling and low/no output trying for two-phase. Really interesting, placing the phases in parallel and obtaining full ouput with no complete wired circut, just the flux flowing betwen coil sets!

Here is waveform from that configuration.



10:1 multiplier at 5V per division at 1200rpm.

I next rewired coils to quad-coil sets to match the 8-pole rotor, so I get three phases of four coils each.

First try was four coils in series which yielded 212 VAC and cogged very hard - no pictures, just a near squarewave output form. Personally I didn't want to get shocked by 212VAC RMS so stopped immediately.  Holding onto the motor housing laminates still gives off shocks as potentials float around - I can't measure any coil as 'shorted' or 'leaking' to ground even under high-power, there is just sharp static like shocks snapping about. A label on motor warns against touching while in operation and motor is never to be grounded.

So next I changed to four coils in parallel, 'start' leads together & 'finish' leads together. Output went through a 15 ohm power resistor - One phase gave 40VAC across resistor with 2.7 amps displayed on my Fluke DMM - 110 watts calculated using I2xR. Also saw 6+ amps at 12VAC through auto headlight capsule at about 120-150rpm before both filiments burned open.

Waveform shot here:

For the electronic gurus - the switched (variable) reluctance motor style can be turned into a generator w/o magnets by using position sensor and a "field" coil sets nested within each coil -  small coils w/ precise timed pulses induce a larger flux cascade, they will be used more often in transportation in the future, the military likes them because they can take alot more damage and continue working then other motor styles.

My next step if I continue w/ this anymore would be to discard the motor housings laminate/coil section and replace with 5" PVC pipe, duplicate the 12-coil arrangement in as a air core arrangement to reduce cogging - but there are too many needy projects to continue w/ this...

[MountainMan]

99y,

Seems like you know some stuff about motors.  I've been thinking about electric vehicles and pondering the motor questions that go with that.  I was wondering, if one were to apply an appropriate waveform to the leads on one of the "from scratch" permanent magnet alternators that are seen around this forum, would it act as a reasonably efficient motor?  Is the architecture of a good motor horrendously different from that of a good generator, given similar speed and voltage requirements?  

Obviously this would require an unusual input voltage waveform rather than simple AC or DC, but assuming one had the electronics cajones to pull that off, is there some reason that a similar architecture would not work well for a low speed motor?  Could one do a good job of controlling motor speed simply by changing the frequency of the drive waveform?  My gut tells me that such a motor would have a fairly level torque curve as the speed varies, since you would be keeping the amplitude of the drive voltage the same and just varying the frequency of the waveform.

Also, do  you know of any forums where people invent "from scratch" motors instead of alternators?

thanks,

jp

[DanG]

Not that smart MM - just took a few months to read up on reluctance motors.

Opinion and conjecture follows, hope my answer is as good as the question.

Easier to design an airport tram or similar using the basic axial design but thats also w/ unlimited grid power to spin it up in the first place...

Wind means rpms loading via constant ramp against the mechanical supports of a resin-cast stator. Do much work and push a load it'd most likely overheat, flex, buckle etc - its made to cope w/ internal forces-losses only. Imagine windstorm and gen flying loaded into short circut, then adding 10, 20, 40% current, nearly the same scenario as dumping a mechanical loads into an electric motor. Whole different world in specifying design efficiencies.

Yes, if a single revolution of an axial flux generator was mapped out in steps via optic sensor, the position of rotor could be synched & made into a motor, but not very efficient since its optimized to be a generator. And few dual rotor machines stators are built for the abuse randomn loads will inflict.

RC hobbyists are about the only people to give away data on brushless motors - many people trying to oversize them, asking there would be good - alot of europeans are into that, make sure you read the german and scandanavia rc hobby pages...

[MountainMan]

Ich spreche nicht Deutsches, aber Dank.

jp