Author Topic: New project  (Read 5324 times)

0 Members and 1 Guest are viewing this topic.

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
New project
« on: April 25, 2015, 02:14:44 AM »
Synchronous conversion of a 5 hp 3 phase 220/440 1.25sf 4 pole motor.
Rotor diameter is 120mm, lamination stack is 5.25 inches deep/long. (works out to a nominal 4.75" stator internal diameter, with a .35mm air gap, or .013 inches

"back iron" behind the coils is only 15mm thick, thickness of the stator core is 35mm, so the slots are 20mm deep.

I have no intention of destroying the rotor, so my tentative plan is to make a new one.
Given the 15mm thick back iron, the rotor will require twice that for the salient pole core.
Conveniently this is the same as the 30 and close enough to the 28mm wide core of larger than average microwave oven transformers. Presuming I can take the E cores, and cut a dovetail in the center leg of the "E", press the core onto a shaft, then cut the E core down and end up with a "T" shaped core with a 60mm radius cut into the top of the core, then i should be able to get a rather legit core for very little effort. (on the other hand i threw out 30 pounds of microwave oven tx cores a month or three ago, so i don't have enough on hand to do this tomorrow.. so.. Alternatively I could attempt to make a stator from solid steel.. but i'm not sure i will have the measurement capabilities to extract the rotor core losses from such a system.)

the motor is pressed into an extruded aluminium case, so i expect the saturation losses to be a bit of a sharper brick wall than i have noticed before.. 
My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: New project
« Reply #1 on: April 25, 2015, 03:53:02 AM »
You don't say what you are intending to use this for. You seem to be aiming to saturate the whole core, which is not the normal way of using these things. Normally there is little point in using more magnet than is needed to bring the teeth near saturation but of course you can squeeze out more output beyond this point.

I can't think of an application where you would saturate the back iron and be worried about induced losses in the aluminium case.

Similarly using solid steel below the magnets would normally not significantly add to any iron loss, with balanced 3 phase the mmf in this region is virtually constant but with a tiny 3rd harmonic ripple. As the neo is saturated this ripple is induced through a very large air gap so in normal conditions it will add no significant loss. If you can make it laminated then fine but otherwise I can't see it being a big deal.

Flux

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: New project
« Reply #2 on: April 25, 2015, 05:00:59 AM »
I see that this is in the hydro section so I assume that is the intended use. If so then I would not use more magnet than to bring the teeth near saturation as the gain becomes one of very diminishing returns, the leakage reactance is lowered with more flux but with proper loading that should not be a big issue.

Flux

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
Re: New project
« Reply #3 on: April 26, 2015, 08:50:15 PM »
should probably go in user diaries, i forgot we had that subforum.

I don't know whether its more tooth saturation or core saturation (i know its some combination of both) on small induction motors but I've had a very difficult time getting nominal line voltage out of a synchronous conversion without heavy core losses.
So i'm going to try a larger motor this time around, nema nominal efficiency is 90%

My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: New project
« Reply #4 on: April 27, 2015, 03:20:45 AM »
Yes I think induction motors are pushed very hard on core saturation, that is how they have reduced the size over the years. The same thing has happened with transformers over the years as higher temperature wire and insulation has become available.

The best full load efficiency happens where the core and copper loss are about equal. For synchronous alternators for most applications a lower core flux is likely to give better overall characteristics. Not so bad for hydro but for wind the high core losses at start up are a big deal. Even if cogging is entirely reduced, the start up of iron cored PMAs can be very poor with high speed ( lower solidity props ).

I suspect you will find things a little better with a bigger motor but to get to nominal volts I think the teeth will be saturated. When running as a motor I think the field in the tooth region will be close to 1.7T so the air will be carrying a fair bit of it.

Flux

joestue

  • Hero Member
  • *****
  • Posts: 1759
  • Country: 00
Re: New project
« Reply #5 on: June 12, 2015, 05:06:22 AM »
close to starting on machining a rotor for this project.. i've come up with a more topologically friendly method of building up a salient pole rotor..

For what its worth, rewiring this 5 hp machine for 138 volts double delta, and starting it and running it from 120vac single phase, while it draws 12.5 amps from the line.. it only consumes 200 watts.
leaving a 55 uf cap connected between the line and the third phase drops the watts down to 170, and the amps down to 10.5.

I suppose i expected significantly more no load losses.. however, 120/138 is 86% of line voltage.. which is close to the "knee" of saturation as I've discovered before.

i've performed one test so far of the effectiveness of using this 5 hp as a three phase rotary converter.. and so far... while the motor does generate the third leg and provide the rotating field needed to start a three phase motor..
the test motor i used today was a 1/2 hp 460v bench grinder (no grinding wheels on it) and i had already rewired it for 138v double delta.
so while it starts up real nice and smooth.. disconnecting the third phase (the generated leg from the 5 hp motor) actually dropped the total power consumption by 1%.

so from what i can tell.. the books are right. unless your generated leg is within 5% for voltage and phase.. sure it works. but a vfd would be more efficient.
My wife says I'm not just a different colored rubik's cube, i am a rubik's knot in a cage.