Boeing 737 APU generator as an EV motor

[filip]

This is my first post on this forum, so hello everybody!

I live in Poland and I came across a Boeing 373 APU generator (this is what my research indicates) on a Polish Ebay
equivalent.

I suspect that it could work really well as an EV motor, and I am planning to convert a car (possibly a Smart Fortwo)
to electric power. My plans include a DIY motor controller.

Basics specs:
- vendor: Westinghouse
- model: 976J497-1
- power: 30 kVA
- current: 83,3 A
- voltage: 120/208 V
- frequency: 380/420 Hz
- power factor: 0.75
- phases: 3
- speed: 5700/6300 rpm
- length: 38cm
- diameter: 28cm
- weight: 32kg
- nameplate: http://www.tdm-electronics.com/katalog3/d/108200-3/P7190006.JPG
- the auction with pictures: http://allegro.pl/pradnica-30-kva-83-a-400-hz-westinghouse-i2996648198.html

Terminals: http://www.tdm-electronics.com/katalog3/d/108206-3/P7190017.JPG
My guess is that the top 3 ones are phase coils terminals.
The bottom ones - excitation coils? speed control?

Advantages:
- better power/weight and power/volume ratios because it's 400Hz and it's for an aircraft - close to 1 kW/kg
- lower operating voltage meaning less battery cells, a simpler BMS and possibly MOSFETs instead of IGBTs in the controller
- quite cheap if I manage to negotiate a lower price (I can't imagine a high demand for it)
- it has a nice flange to connect to an adaptor plate

Disadvantages:
- possible issues with servicing - sourcing replacement bearings
- not sure about its past - might have been abused
- low power factor could indicate low efficiency
- air cooling looks like passing air through the interior, which might not be a good idea in a car
- controller might be tricky if it's not an ACIM
- achieving 400Hz with a controller might be a problem.

Dillemmas:
- is it an ACIM? I think it was hard to get power from an ACIM on an aircraft back when the 737 was designed - no inverters
- if it's not an ACIM, how is it excited?
- are such machines symmetrical in performance as generators and motors?

An alternative would be ordering a custom wound (for low voltage) 112 frame 4 pole industrial ACIM like 5.5kW
and then driving it up to 200Hz to get more power out of a small motor
(like it is being reserarched here: http://forums.aeva.asn.au/changing-an-induction-motor-voltage_topic1237.html),
but this might be problematic due to hysteresis iron losses, cooling issues, bearing longevity and rotor balancing.

So, what do you think about the idea?

Thanks.

 

[Farfle]

Wow, that looks like a winner. Looks like all six winding ends are brought out, so changing it delta/wye gives you some options on what voltage/amperage you can run it at. The dual 120/208 indicates delta/wye. A scott-drive should run that pretty well in wye if you can run a high-v low Ah battery pack (300v IIRC) Or if you want to run a low voltage high ah pack the curtis 132v 650A controller is a great choice.

 

[filip]

Wow, that looks like a winner. Looks like all six winding ends are brought out, so changing it delta/wye gives you some options on what voltage/amperage you can run it at. The dual 120/208 indicates delta/wye. A scott-drive should run that pretty well in wye if you can run a high-v low Ah battery pack (300v IIRC) Or if you want to run a low voltage high ah pack the curtis 132v 650A controller is a great choice.

Hmm, but is it an ACIM? I can't imagine getting a three phase power out of an asynchronous machine without another power source feeding the phases...

 

[Farfle]

Its just like an ACIM controller does regenerative braking. Theres some unit onboard the plane that's making it happen, might be the same unit that converts it to the stabilized AC used by the plane.

 

[filip]

Its just like an ACIM controller does regenerative braking. Theres some unit onboard the plane that's making it happen, might be the same unit that converts it to the stabilized AC used by the plane.

As far as I know the ACIM controllers were not available back when 737s were designed, and some kind of a clutch (fluid based?) was used to get the exact needed rotational speed and AC frequency.

Here is what I found in http://www.smartcockpit.com/download.php?path=docs/&file=B737-Electrical_Systems_Summary.pdf :

Primary power is obtained from two engine Integrated Drive Generators (IDGs), one mounted on
each engine. The IDG is a common housing containing a generator and drive which allows the
generator to maintain a constant generator speed and thus a constant frequency throughout the
normal operating range of the engine. The IDG is coupled directly to the engine and operates
whenever the engine is running.

So if there were no ACIM controllers, then what kind of a machine is it?

 

[Farfle]

That's a very good question. Connect the full color leads ( the solid red-yellow-blue leads) together, and then apply current to a pair of the half color leads (a 9v battery can typically do this, if it doesnt, then use a cordless power tool battery) . If it orients to a position, its a PMAC, and the Scott drive, or a big Kelly controller will work, if it does nothing, then its an ACIM, and the Scott drive or a Curtis controller will be good options.

 

[Farfle]

Sorry, thought you had the motor in hand, hard to tell what the motor type is. Looking at google, the apu was used both to start the apu turbine from ac, or a battery, and to generate power for the aircraft. So they had to have some way to convert battery power to the high frequency AC needed to spin the generator and start the apu turbine.

 

[filip]

So if there were no ACIM controllers, then what kind of a machine is it?

I've found a gallery with a similar generator (the white one):
http://s117.beta.photobucket.com/user/jaynethecat/library/Range%20Extender%20Stuff

This photo: http://s117.beta.photobucket.com/user/jaynethecat/media/Range%20Extender%20Stuff/IMG_0646.jpg.html?sort=6&o=2
shows a stationary coil connected to the smaller terminals.

This is a schematic of a voltage regulator for this generator: http://s117.beta.photobucket.com/user/jaynethecat/media/Range%20Extender%20Stuff/IMG_0513.jpg.html?sort=6&o=25
I think that the generator is the thing on the lower left enclosed in a dashed rectangle.
A close-up: http://s117.beta.photobucket.com/user/jaynethecat/media/Range%20Extender%20Stuff/IMG_0513-1-1.jpg.html?sort=6&o=5
A lot of coils (and a rectifier) seem to be there...

Here is a plausible explanation: http://www.diyelectriccar.com/forums/showpost.php?p=140148&postcount=20 :

F2-F1 is a field winding. G1 three phase winding, the rectifier and the armature coil are all spinning on the rotor. T1, T2, T3 are the stator windings

So the stationary coil excites a three phase winding on a rotor which after recitification powers the rotor flux coil...

A lot of coils...

I wonder how good is the efficiency of this complex layout compared to a classic ACIM.
Maybe "overclocking" a 50Hz ACIM is more efficient than dealing with all this rotor current conversions...

 

[liveforphysics]

I believe some of those are partially PM (weak-ish magnets) sized just to be able to supply the bare operating loads of the plane.
Then when they go under heavy loads they start to get slippy and work like an induction motor because the magnets are sunken into the induction motors rotor face.

 

[filip]

I believe some of those are partially PM (weak-ish magnets) sized just to be able to supply the bare operating loads of the plane.
Then when they go under heavy loads they start to get slippy and work like an induction motor because the magnets are sunken into the induction motors rotor face.

It doesn't seem to be the case with this one, as it contains a separate excitation system.

Having small PMs in a rotor and relying on an asynchronous principles under a heavy load would not need a separate excitation, right?

 

[filip]

There is a potential issue arising from the fact that it's a generator, described here:
http://www.diyelectriccar.com/forums/showpost.php?p=140278&postcount=23

At a rest/stall condition, G1 isn't spinning and therefore can't pick up the induced current from F1-F2. The motor would never start because there is no field set up on the rotor. The solution would be to use a 400Hz AC current into F1-F2 that reduces in frequency inversely to the motor increasing in speed. That would guarantee full torque from 0RPM. The AC waveform would need to be kept at the correct phase angle to the rotation of G1. You'll need a shaft encoder and a specially designed/programmed controller just to take care of the rotor field.

This is all getting quite complex and expensive unless you're looking to take on this task yourself - even I would be hesitant about attempting this.

On the other hand if I wanted to do field weakening for higher speed (does this kind of a motor need field weakening?), then I would need a PWM circuit for the excitation coil anyway.
During a stall condition I could just make the excitation current changing with the PWM circuit and get it through the F1-F2 => G1 coupling.

I wonder whether a single sided transistor would be enough or would I need a full H bridge to provide enough current?

 

[drewjet]

The APU and engine generators on a 737 are the same. On the engine they use a Constant Speed Drive (CSD). It is basically a hydraulic pump and motor in one unit. It takes the varrying input speed from the engine and outputs a constant speed. On the APU it has a seperate DC motor for starting. The APU runs at a constant speed through the use of a govoner built into the Fuel Control Unit (FCU). As far as I know these do not have any magnets in them. The is an external Generator Control Unit (GCU) that provides the excitation to the generator. As I recall the winding are fixed external to the rotor like an inrunner. I think that someone with good machine skills could remake the rotor to be magnets and could turn this into a 3 phase Brushless motor, like an oversized Astro.

 

[filip]

I think that someone with good machine skills could remake the rotor to be magnets and could turn this into a 3 phase Brushless motor, like an oversized Astro.

What do you think about using it without mechanical modifications - feed the rotor coil with a square wave from a H bridge (to pass the coil-rotor gap at 0RPM) and feed the phase coils
as if they were in a PMAC motor?

 

[drewjet]

I am going off my memory, and that is never to be trusted, and we currently have no spares for me to look at, but there are no brushes feeding power to the rotor. I think, and I am by no means an expert on the different types of motors, that it works on inductance like an AC motor. I could take a look at the wiring diagran and see if that may give a clue.

 

[drewjet]

Try this
View attachment ATA 24, ELECTRICAL POWER___R093 22.pdf
View attachment ATA 24, ELECTRICAL POWER___R093 23.pdf

 

[filip]

I am going off my memory, and that is never to be trusted, and we currently have no spares for me to look at, but there are no brushes feeding power to the rotor. I think, and I am by no means an expert on the different types of motors, that it works on inductance like an AC motor. I could take a look at the wiring diagran and see if that may give a clue.

Yeah, that is exactly reason why I thought about feeding it alternating current. Commutatior would pass DC nicely, but inductance passthrough needs changing flux.
Frequency of this supply shouldn't matter that much since it's rectified on the rotor anyway, and rectified from 3 phases - not a lot of ripple. So something like 100Hz could work.
Above something like 1000RPM I could switch to DC to reduce noise.

 

[filip]

Try this

Wow, thanks!

I don't understand much of the schematics (they seem more like block diagrams), but I'll try to dig something out of them .