DIY Electric Car Forums banner
1 - 10 of 10 Posts

·
Registered
Joined
·
24 Posts
Discussion Starter · #1 ·
I think it should be obvious, but I'd benefit from confirmation of it. Even though I could confirm it by experiment, I'm not set up to make such experiments right now, and have limited time between regular work to fool with such things. I need to spend my limited time on more important aspects.

My ? is; will a normal 3 ph motor run on 3 ph pulsed DC?

The USelectricar I repaired awhile back and sold had a 300V variable frequency 3 ph inverter running the drive motor, I'm almost 100% certain it had to be running on 3 ph pulsed DC because there were no transformers between the IGBT devices and motor. Because it seems to me in order to get 3 ph AC without a transformer it would need 6 devices with 3 running upside down (electrically) in relation to the other 3. Seems like the motor itself would serve as its own transformer with back emf and/or induction from the other phases supplying the other half of the waveform.

I'm working out the design on a controller that may be very cheap and easy to build, but it may have 3 ph pulsed DC output without using expensive and/or unmanageably bulky devices.

Now I'm not a degreed design engineer, but I do have some interesting men of inventive qualities in my direct ancestry and the good Lord seems to have given me a large share of their mind. My ideas are always running far ahead of my ability to prove them by experiment or make use of them.
 

·
Registered
Joined
·
24 Posts
Discussion Starter · #3 ·
Thanks, I thought it had to work just fine. When I get a controller workable I'll put my scope on it and see what it's doing for myself, with dual trace I can see 2 phases at a time.
 

·
Registered
Joined
·
74 Posts
You'll have to bring all 6 taps out of the motor or at the very least the usual 3 plus the star point for a Y wired motor. You'll also have to drive each phase much harder to get the same power out (vs a normal ac trapezoidal drive waveform) leading to increased I^2R losses, voltage overshoot problems, high ripple current and high torque ripple (vibration, noise). 3PH inverters use 6 power switches for good reason. A lot of the bulk in a medium-high power inverter is dedicated to handling the ripple current (caps), snubbing the voltage over-shoots (caps/snubbers) and cooling. All of those parts will need to be as big if not bigger in your design. You do save the cost of 3 power drivers but you'll be plowing it back in elsewhere to compensate for the design inadequacies.

Good luck, I don't mean to be negative. There's no reason for a workable, affordable home-brewed inverter to be beyond the reach of an enthusiastic designer/experimenter but I'd suggest significantly more reading is required before you spend any money or let the smoke out of something expensive.
 

·
Registered
Joined
·
24 Posts
Discussion Starter · #5 ·
Ah; something just triggered another idea. I just might be able to get AC out of the thing with an extra 3 $7.00 devices. I think it would be more square wave than sine though, but I could separate the + to - swing of the wave form as much as I want to by creating some 0V flat time between the + and - waves. Oh well. things to play with as I can find time. The most expensive part is the capacitors but I got some that are hard to kill.
 

·
Registered
Joined
·
514 Posts
Once you pass the station of understanding how to make AC currents out of DC. you should come to the conclusion that the real issue with three phase speed controllers is how to feed the motor in a way it produces actual the torque you request.
It is possible to make a very crude controller which takes encoder speed and translated it to a output frequency and voltage limited by current. Most modern (1980+) VFD's/three phase inverters allow the user to give accurate control of the torque, which is highly desirable.

There are some very nice papers available regarding Field Oriented Control which describe a large portion how to make (controlled) torque in induction (and also synchronous 3ph) motors.

Im current at trying to understand how to feed the clark / park inverse-/transforms with the proper angular information. :eek:
 

·
Registered
Joined
·
74 Posts
Ah; something just triggered another idea. I just might be able to get AC out of the thing with an extra 3 $7.00 devices.
Switching power levels of use to an EV or are we talking about model cars :confused:

For the core of a small homebrew EV inverter you're looking at something like this http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=0194912 Something like that would allow you to build an inverter loosely in the 20-30kW output range. If you've figured out how to do it with much less then I eagerly await the unveiling and testing of your prototype.

jk
 

·
Registered
Joined
·
24 Posts
Discussion Starter · #9 ·
the $7 devices will handle nearly 20 KW apiece (surplus market, $10 new) If it will work the whole thing may cost 250-$300 in parts and be capable of up to 150KW. but I'll have to thoroughly test it on some motors I picked up from the local scrap yard. 25 hp is the biggest motor I have. Nice IGBT units there, but I hope I can avoid using them.
 
1 - 10 of 10 Posts
Top