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

[thingstodo]

Originally Posted by MPaulHolmes

I think I found out what was wrong in the code that caused the weirdness. I'm going to go test it, in the 3,000,000degF shop.

So you have not buried 300 feet of copper tube and tried the circulation pump/car radiator/fan cooling we discussed?

There are 2 sets of PI values. For kp-id ki-id, and kp-iq ki-iq. I had kp-iq and ki-iq set to weird values. It works to set them to the same as kp-id and kp-iq. If you don't want to wait around for me, type something like this:

kp-id 1000
ki-id 1
kp-iq 1000
ki-iq 1

run-pi-test

OK - something to try tonight. I'll maybe hold off on using the wire cutters on burnt MOSfets.

So, each time you change the values, change it for both of them. I'll change the code so that it happens automatically. I just wasn't sure if there was ever a situation where kp and ki for Id and Iq needed to be different. It never had to be for the 2 motors I've tested so far.

Was that something you did to allow the controller to work with PMDC or switched reluctance type motors, without changing the code?

I'm going to send over another hex file. I also had set Id to 75amp, and Iq to 0amp. It would make more sense for Id to be 0 and Iq to be 75.

Whatever you say!

[MPaulHolmes]

I got me some good people working on this! Isn't collaboration fun!?!

P.S.: I wouldn't worry about the controller blowing up. An explosion never hurt anybody... Killed, yes. but never hurt. haha

[thingstodo]

Originally Posted by MPaulHolmes

I got me some good people working on this! Isn't collaboration fun!?!

Actually, I quite ENJOY this sort of thing. It's NEVER BORING! And I'm learning stuff all of the time.

Education ALWAYS has a cost ... your time .. the occasional electronics .. it's ALL GOOD.

P.S.: I wouldn't worry about the controller blowing up. An explosion never hurt anybody... Killed, yes. but never hurt. haha

I've had many electronics blow up in my presence. I have been the likely source of the blow up in about half of those cases.

There have been LOUD NOISES (ringing ears kinda hurt) and letting out the smoke (bad smell, ozone sharp smell, varnish burning) and BRIGHT FLASHES (you see spots for a while, but it doesn't actually hurt) .. some sizzling .. dark scorch marks .. melted copper and aluminum ... no deaths so far!

[thingstodo]

My results are ... confusing ...

Paul - Here's hoping that it makes sense to you!

There is output for kp 2000 and ki 0, kp 1000 and ki 0.

Changing ki to 1 gives extreme output - that's the best description I can come up with - EXTREME

kp 500 and ki 0 gives no output, like the output is disabled .. somehow .. but if you DONT cycle power and go back to kp 2000 and ki 0 .. there is output again.

kp 20 and ki 0 gives no output, or just noise. It does not drop over time and get there eventually.

Here are the data files and a description of what I did at each step.


file 200. kp-id 2000 ki-id 1 kp-iq 2000 ki-iq 1. Extreme graph. The controller must have hit current limit.

file 201. kp-id 2000 ki-id 0 kp-iq 2000 ki-iq 0 There is no output - just noise.
cycle power on high voltage, then 12V.

file 202 kp-id 2000 kp-iq 2000 ki-id 0 ki-iq 0 - results 528, drops to 145 at 38, 28th sample. Should be good? Nice and slow.
But 200 shows insane output, with ki 1

file 203. try 1000 and 0. It's slower and never really gets to 0

file 204. try 1000 1. extreme output again.

Did the multiply by 8 work?

file 205. cycle power and do 500 and 0 - no real output, just noise

file 206. shut off high voltage and watch it bleed away. Leave 12V on. Then turn off 12V at 114V. Just a trial
Nope - no difference. No output

file 207. shut off high and low voltage, drain high voltage with 10 ohm 25W resistor.
Nope - no output at all.

re-run 202 - 2000 and 0 - file 208. Start by discharging caps on high side. It works!

file 209 - try kp-id and kp-iq 200 and ki-id and ki-iq 0. no output
dont cyclepower.

file 210 - go back to 2000 and 0. It worked?

file 211. One last trial - kp 20 ki 0. Don't cycle power between. Does not work, but perhaps output is not disabled. Hmmm?

[thingstodo]

Paul sent me some new code, but I have not gotten out to test it so far.

I hope to get the code loaded up and do some initial testing Friday night.

[thingstodo]

Update for Sep 4

hex file dated Sep 2, 2015 was loaded into the AC controller
- plug PICkit3 into AC board ICSP header
- plug usb cable into laptop
- start software to load hex file into controller
- while that's running, turn on 12V to controller (not high voltage)
- connect to controller via PICkit3
- erase controller
- load hex file from USB stick
- program controller with hex file
- disconnect from controller
- exit software
- turn off 12V to controller
- remove USB from laptop
- disconnect PICkit3 from controller

Start hyperterm at 115,200 baud, n, 8, 1 with no handshaking
Start capture to 220.txt
type - nothing is echoed?
check serial cable - not seated well. Push the connectors all the way together and try again
file 220 is listed below

AC controller firmware, ver. 1.0
//220
kp-id 2000
ki-id 0
run-pi-test
// no response. Check serial cable
run-pi-test
// still no response - cycle power, both high and low voltage
AC controller firmware, ver. 1.0
// precharge is definitely not 5 seconds
run-pi-test


That was the end of the test. Besides echo of the commands, the program appears to do nothing with these parameters. But this used to generate an output.

Shut everything down. Discharge the high voltage bus. wait for 2 minutes. Power up and start again. File 221.txt


AC controller firmware, ver. 1.0
run-pi-test

Again, no response. Did I kill the controller somehow? How can I tell? The programmer shows that the program was successfully loaded. The PICkit3 and software respond to the erase command.

Go back to the last hex file I have, burn it into the controller again ... just to make sure that SOMETHING still works.
- Repeat the procedure above with the older hex file.
After the hex file is burned, disconnect the PICkit 3, connect with hyperterminal and capture data to 222.txt, which is attached. The first part is reproduced below.

AC controller firmware, ver. 1.0
run-pi-test
+0497
+0528
...
+0544
+0512
+0528
+0512
+0544

The output is not great, but the controller reacts to the run-pi-test command, so I don't think I killed anything.

It's almost like the hex file was corrupted somehow ... but now my dim little brain recalls that Paul's email included a warning -

Hello! I got the controller up and running out there, and hooked up 120v nominal batteries (to be like your setup) and did a fairly significant update to the code. Here's the most recent hex file. Do you have a throttle attached? I have a high pedal lockout setup right now.

for those of us who missed it the first time , here's a recap ...

I have a high pedal lockout setup right now.

Oh yeah. That would make the controller completely ignore me until the throttle comes into range .. too bad it took me until I was writing the update before I figured that out!

I guess I'm set for testing on Tuesday, when I get back home!

[G-Max]

I'm loving reading all this. Its' like a brilliant electronic engineering drama.
Hey, on the controller board there are 6 sets of two diodes mounted vertically with their cathodes soldered together. What are you using them for?

[MPaulHolmes]

Hello! Those are 5watt 17v zeners, back to back. Right at the gate to emitter. Like this:

emitter -----> <------ gate. That way, if gate goes outside the range of -17v to 17v, it gets clamped very hard. zeners clamp more harshly than transient voltage supressors.

[G-Max]

Is that so you get nice sharp square waves on your PWM output signal, or for a safety purpose?..

[MPaulHolmes]

It's just for safety. A 600v 600amp IGBT can dissipate like 4kW of heat, but if the gate to emitter goes outside the range [-20v, 20v], you will destroy that sucker.