Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[CollinK]

That curved bar idea is pretty slick! If I end up needing more capacitance I might have to go that route.

I'm going to try to do some more testing of my controller with the large film cap installed and see if the new (hopeful) lack of parasitic inductance on the input helps to keep the spikes and nastiness to a minimum.

Another thing I found is that I might have a bit too small of a gate resistor. I'm using 3 ohm resistors on 200A IGBTs. It seems like maybe that's a little on the small side. As such my di/dt is kind of fast and that doesn't help either. I'm thinking it might be safer to go with 4-6 ohm resistors. Any suggestions there?

[nlc]

Originally Posted by CollinK

Another thing I found is that I might have a bit too small of a gate resistor. I'm using 3 ohm resistors on 200A IGBTs. It seems like maybe that's a little on the small side. As such my di/dt is kind of fast and that doesn't help either. I'm thinking it might be safer to go with 4-6 ohm resistors. Any suggestions there?

The only way to answer your question is to look the gate signal with scope to see what happen exactly.
If your gate resistor is too small, igbt turn OFF can be too fast and provoke spike, but also igbt turn ON can be too fast and you will have a high current spike (and then voltage spike and ringing) because of the reverse recovery time of the opposed igbt's diode.
On the other hand if you put a higher gate resistor, you can easily reach autoconduction threshold when the opposite igbt turn ON, because of the high dv/dt on the turned OFF igbt.

The only way to adjust the gate resistor is to carefully look the gate signal, to see the turn ON time, and verify that the gate voltage don't reach the commutation threshold when the opposite igbt turn ON. When igbt are not controlled with negative gate voltage the 4-5V threshold can be reached easily with high voltage bus (and thus hight dv/dt), even with pretty low gate resistor.
With high voltage Vbus it's often mandatory to have different value of turn on and turn off gate resistor with the help of a schottky diode if you don't use negative voltage for igbt turn off.

[CollinK]

Thanks, that sounds like a plan. I'll stay with the 3 ohm resistors for now and scope the controller while carefully running a small motor.

[nlc]

Habitually it's also very useful to gradually increase bus voltage instead to directly test with full bus voltage. For example for the 72V/100A controller I work on (the one of my previous videos), I began with 24V, then 48V, then 60V, then 72V.
And at 48V I seen that my turn OFF rgate value was too high and I was at the limit of autoconduction. If I tried directly from 72V, I don't want to imagine the result that I had )))))

[CollinK]

Yes, that's a good idea. I have sort of been doing that. I tested first with about 80v worth of cells (shooting for 330v at the end) and then added a few batteries and got up to around 100V. Once that testing went ok I went right up to rectified 230VAC (the caps charged up to about 339V). Maybe that was too big of a jump. Also, trying to run on rectified AC causes issues with the sagging voltage from the mains so I wasn't helping myself there either. So, maybe I should stick with batteries and just gather up more of them and add them a few at a time until I get up to a reasonable voltage.

[mrbigh]

Originally Posted by adamj12b

Looks great MrBigh!

How did you end up bending those curves on the bus bars?

-Adam

I used an acetylene torch to soften the copper bar and then did the bends on a jig.
Then, I welded the two different planes with the same oxi-acetylene.

[mrbigh]

Originally Posted by CollinK

Thanks, that sounds like a plan. I'll stay with the 3 ohm resistors for now and scope the controller while carefully running a small motor.

I'm using 4.7 Ohms 2Watts........driving 600Amps x 144VDC rail, but still in "testing mode" with out any issues with a work bench transmition/motor assy.

[mrbigh]

Originally Posted by etischer

Make sure your bus bars don't rub through the vinyl insulation on your capacitors! You might also verify the vinyl insulation is rated for your bus voltage.

Tanks for the advice; actually I have some rubber insulation in the buss bars facing the elcos to avoid vibration "peeling" of the caps insulation.

[CollinK]

I have been working on the controller but recently most of my time has been spent elsewhere. Though, if you check the source repo you'll see that I have been making semi-regular updates to that.

Today I started compiling wiki pages which detail the trials and tribulations of making your own controller. The goal is to gather all the relevant info a person might need when designing or modifying a controller. You know, all of the stuff I wish someone had said to me before I blew up some IGBTs and ruined drivers.

It would be cool if any knowledgeable people could either sign up to be a member of the project and add to the wiki themselves or post what info they think should be included. As I can I'll edit the wiki to include more and more formulas and information. If, Heaven forbid, I got some detail wrong, then please tell me that too!

revolt-ac-controller - Open ReVolt AC Motor Controller Firmware - Google Project Hosting

As a side note, I do plan to actually do more testing soon and scope the controller and do more videos and all that jazz. Hopefully that all will happen within a week or so.

[steiner]

Paul,

About midway through this thread you were working on an AC controller that used mosfets in the power section. I haven't been able to find anything on the wiki in regards to a mosfet based power section. Do you have any of these boards for sale or do you happen to have the PCB files/schematics? I already have the mosfets and capacitors.

Thanks...
Rick