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

[MPaulHolmes]

Wait on applying battery voltage. Make sure we are on the same page for contactor control.

First, it goes like this:

+12v ---------------------- contactor in (on board)
contactor out(on board) -------------- contactor coil PLUS

contactor coil ground ------------- +12v battery ground.


You will also need to run the angle offset test, and stuff like that.

[arber333]

Originally Posted by MPaulHolmes

Wait on applying battery voltage. Make sure we are on the same page for contactor control.

First, it goes like this:

+12v ---------------------- contactor in (on board)
contactor out(on board) -------------- contactor coil PLUS

contactor coil ground ------------- +12v battery ground.


You will also need to run the angle offset test, and stuff like that.

Well... i have it like this now:

+12V-->EV200 coil+ --> EV200- -->small relay contacts-->GND
while i command small relay trough PCB relay ===> relay PCB+ --> relay PCB- -->GND
==> is actually single wire from inverter, i am low on wires so i connected gnd to inverter alu box.
The same for precharge relay. I measured continuity on all contacts and it checks out. It works. I had to add intermediate relay since EV200 needs a clear GND to latch. I guess economiser makes trouble. I saw the same problems with Johannes inverter. There i use one NPN transistor to open collector trigger small relay which in turn latches EV200. DOH!

Since there is another relay in the wiring i can use it as a safety and wire EMG button to remove GND from relay so EV200 goes out.

[arber333]

Argh! i totaly missed your photo from instructables about desat connection!
Now i will have to dissassemble everything up to PCB. Can you tell me if i can just connect all high transistors to common + rail? I would go from one IGBT to another and finally out to + rail with single wire. It would made for much cleaner installation.
Lower IGBTs will be connected to E like on your picture.

EDIT: I soldered a desat sense wire along lower PCB side. Black for lower transistors to E and red for higher transistors to + rail. The latter i just connected to one another and finally i put one wire out to + rail. When i connected PCB +12V system worked, no desat warning...
I will get my 40Vdc and put them on!

[kalidasbala]

Paul, what is the battery pack size for your go-kart? How fast the solar charger is able to directly charge the batteries - charge rate?

[MPaulHolmes]

Battery is 5.6kwh, 93v A123 cells from a spark EV. Capable of about 100kw! Maximum charge rate during the right time of year should be around 50 amps from the panels.

[arber333]

ARGH! 3x IGBTs are busted. One battery connector was reversed and i didnt figured until a little smoke came from main board.
Smoke originated from gate resistor on low transistor no 3. Main board reported undervoltage error.
When i connected main board again no errors were reported and precharge and main conatctor were latched normaly. Of course i didnt connect main voltage this time.
When i took it apart i found totaly shorted IGBTs. All three. They were test pieces but still i feel bad. I dont know about drivers yet.

Paul can you tell me the procedure that can show me if drivers work ok?
Could i just connect main PCB and run some routine that would give 50% duty on all gates so i can measure with scope?

I do have spare IGBTs, but i want to know if everything else is ok first.

EDIT: I powered up main PCB again and measured all contacts. What i found:
IGBT no 1 = -6V
IGBT no 2 = -2.3V !!! Huh?
IGBT no 3 = -6V
IGBT no 4 = -6V
IGBT no 5 = -6V
IGBT no 6 = -1.2V !!!! huh?

I also measured at zener voltage splitter and i get 11.75V between D54 and D55 and 10.2V between D50 and D51. Other voltages are 7.2V
Still need a way to check drivers?

EDIT2: DOH! I see why values are strange... i had my lab PSU set to 12V when i debugged PCB. It was late and i was not in a great mood....
Today i will put propper 24V on it and run "debug 1" command.

tnx

[arber333]

Hi paul

I have replaced drivers. I havent put PNP NPN transistors on yet. I wanted to try circuit without them first.
I connected EVERY emitter to common collector like you said. Unless there is more to it...
Well i put your debug code in. I started it and typed "debug 1" into console.

I measured driver output... I still get -8V on output! So drivers keep output down. Into drivers i see a steady square weave from dsPIC.
Is there a method to just run square weave output from drivers regardless of wiring? I want to know if there is other trouble there somewhere.
When powered i get cca 0.35A draw on 24Vdc. OK?

EDIT: I connected only Cn and En together and tested only Driver n, without the rest. It seems it is not good for rest of emitters to be connected to rest of capacitors at least for testing.

EDIT2: I dont get it... now i got 0.5A draw and driver U21 and transformer 1 are also hotand, but mosfet driver U24 is really hot! Well i tried weave form on mos driver and sure i get slanted weaveform of a current. Something is drawing current from transformer. Should that mean driver is shorted? I just changed drivers! I also saw one zener diode was very hot, while other not!

EDIT3: Well i found 3x 100R lower collector resistors burned. On sight they were ok, but when i took them off they fell apart. hm... I replaced them with DIP wire resistors for testing. Paul says wire resistors 100R shouldnt be a problem.
D51 is hot as hell... rest of the circuit is ok. I took driver resistor off. Amp draw is 0.35A again... D51 checks out and activate at 17V, grr... i missed that! Its supposed to be 24V diode. DOH! I have another 24V zener somewhere...
That did it! Now when i command "debug 1" i get nice square function on driver resistors. I will just add totem pole transistors and try my 450A IGBTs that i luckily didnt put first in. Until tomorrow then...

A

[arber333]

I ran the inverter with 40V. everything was good, so i tried to start your pi test. It was succesfull for 40V. Then i ran offset test... Well motor doesnt want to spin so should i leave it there? Motor jumps a little but it doesnt move much. The sound of water flowing is disconcerting. I dont have reference whether pwm is working or not. I dont have any cooling for inverter yet...

Is 40Vdc good voltage or is it too small. I just took a pack from front of my car since it was the easiest thing to do. My friend with 160V car will stop by tomorrow and i can connect inverter to 160Vdc. Would that be better?

What to do next?

A

[thingstodo]

Originally Posted by arber333

I ran the inverter with 40V. everything was good, so i tried to start your pi test. It was succesfull for 40V.

Sounds good. If I remember, the motor tuning values change a bit when you change voltages. But the numbers you got for 40V should work OK with higher pack voltage. running the PI test at higher voltages will get you more aggressive acceleration.

Then i ran offset test... Well motor doesnt want to spin so should i leave it there? Motor jumps a little but it doesnt move much.

I don't remember an offset test. Is that for the encoder?

The sound of water flowing is disconcerting. I dont have reference whether pwm is working or not. I dont have any cooling for inverter yet...

So the PWM sounds like water running? I had a couple of different sounds. A whine was the first one, then Paul put in the random frequency and it sounded more like a white noise generator. But no running water noise.

I ran the inverter with no water cooling during my tests. It will shut itself down on high temperature and on high current. I was not moving a car or anything, but I did have some load on the motor.

Is 40Vdc good voltage or is it too small. I just took a pack from front of my car since it was the easiest thing to do. My friend with 160V car will stop by tomorrow and i can connect inverter to 160Vdc. Would that be better?

I ran a bit higher - I think around 50V (48V lead acid pack) at first. The rest of the time I was running at 128V, which is my pack voltage. I don't know if running the offset needs more voltage or not ...

Paul - what does the offset do?

Arber333 - do you have a list of what you have done, so far, written up? I will be working with the new AC controller board in the new year, connected to some older and smaller IGBTs ... I guess I should go through the thread again and see what's been happening lately!

[arber333]

Originally Posted by thingstodo

I don't remember an offset test. Is that for the encoder?

Paul - what does the offset do?

Arber333 - do you have a list of what you have done, so far, written up? I will be working with the new AC controller board in the new year, connected to some older and smaller IGBTs ... I guess I should go through the thread again and see what's been happening lately!

Well sure! I am keepin a blog https://leafdriveblog.wordpress.com/
Offset is the angular value of encoder misalignment with motor magnetic field. Even Leaf motor has this written under the mottor. So that they can run another motor with the same inverter or vice versa. I just dont know how they put the code in.
Paul in your software, does the angle show difference from north pulse or just from any pulse? In that case the motor just spins into the blue until it sees Z marker yes?