Mainboard V2 troubleshooting

[Esbro]

Well spotted uhi. Although, I have the probe grounded directly to the chill-plate and all the grounds go to that for now. I'll change the probe and check the connections to ground.
Yes johu, I have a 5v feed into the driver boards -9v and 15v from the DC-DC. The 11.1v was on the signal feeds into the board socket - the reading was taken from the scope. ...... I'll check the ground and probe. Thank you both for your replies - I'll report back shortly...Steve

[Esbro]

Well spotted uhi22, I've now checked grounding. When I disconnected the breakout board from the Main Board,to isolate each section, it looks like I missed a ground connection - now fixed. I now have absolutely no scope trace - except for what I think is some sort of "static". - see photo. The lasterr command shows "1" overcurrent, but there is nothing on the bus and the current sensors' output voltages are 1.79v,1.82v. johu has told me, previously, that the inverter had shutdown due to this error which I had hoped I had reset, by 'set lasterr 0'.

Hi johu, yes, I do have the voltages 5V to DC-DC and -9v,15v out and I have checked the resistors - no burn and the values are ok.. The TVSs are all the same with (using diode test on multimeter) 2.3ish volts one way and 0.79v the other. As you can see from my comments above, I've come to a stop again. I don't seem to be able to get over the error 1, this would I guess give me no signal at the driver boards.
My guess is that I have only the one problem (maybe) but I can't fathom out what it is.
Thank you both for your replies... Steve

[Esbro]

sorry for duplication - I thoughtI had lost the original, so re-wrote it

[johu]

About the overcurrent error: I just saw you set really low gains/dividers for ilXgain. Revert them both to 4 for now. That should get rid of the overcurrent error.

Do you see -9V on all gate driver outputs?

If not try removing a TVS. What you measured seems odd but it might just be the 10k pull-down.

set lasterr 0 does nothing at all, it is not writable

[Esbro]

I seem to be having trouble posting. I have posted 2 replies and they are not shown. Maybe it's something to do do with me adding attachments. sorry.... Steve

[Esbro]

Having trouble posting - my last posts did not show

[Esbro]

OK I think It's ok now. Thanks johu, I'll try setting the gain to 4 and see what happens. I have -9v & 15v from the DC-DC but nothing to the gate outputs. This is what prompted me to suspect the optos. The TVSs don't show a short across them, by multimeter buzzer setting.
There is no output from the inverter board, but resetting gain may fix that. I'll do what you suggested now and report back... thanks... Steve

[johu]

Please remove a TVS. Do it now.



Then check again for 9V

[Esbro]

Post didn't go again ! Did as requested removed TVSs I have 9v on all outputs. Outputs on DC-DC ok too. No input on Driver Board input.
From the posts that did not arrive - the gist is.... have -9v on Top,Ebot, have fixed error1 ( ocurlim set at -1000, now set at +1000. There is now an error 10 pre-charge, but I am running manual mode (start 2) and there 0v on bus. I have a list of parameters but can't find one for precharge volts.
Hope this makes sense .... Steve

[johu]

Aha!
udcsw
But you can ignore that as long as you're in manual mode

[Esbro]

OK. From that, do I ignore error 10 ? I still have no signal at the driver inputs. Could this be a problem on the Main Board ?
Thanks .... S

[Esbro]

Tried checking again yesterday and got a square wave at the driver board inputs. Then I connected up the boards and lost the square wave.
"errors" gives me -
15: DERATE TMPHSMAX
500: STOP PRECHARGE
24847: WARN HICUOFS2
24857: STOP OVERCURRENT
Get "lasterr" gives 1.00 (overcurrent). il1gain and il2gain are set to 4. OCURLIM 1000

Checking JP3 pins show 3x 5v supplies, 3x gnd, but no outputs on PWM1,2,3 Top or PWM1,2,3 Bot. I guess, because inverter is shut down.
Maybe the problem is with my parameter settings ?? .... but don't have a clue which ones - except from the error msgs.

TMPHS shows 84.25 (it's 12C) I changed the sensor, still no difference. Open circuit gives -25C (I did try a sensor from my m/meter (13kohm) this gave a near temp 17C)
Precharge (suggested to ignore this on manual start)
HICUOFS2 No warning noticed.
OVERCURRENT - no voltage/current on bus. I'm running the latest firmware.
Any Ideas much appreciated ... S

[johu]

Yes TMPHSMAX is because of the high reading of tmphs. Are you using the JCurve sensor that was provided with the kit? Because that is what you set snshs to. Otherwise just leave it open circuit for now.

HICUROFS2 points towards the current sensor signal not making it to the processor. Where R13 meets C7 check if you still see your 1.7V there (check also at junction C5/R9)
OVERCURRENT is probably a resulting fault.

[Esbro]

Hi johu, Thanks for your reply. No, I'm using KTY83 (in my case KTY83/120) sensor, specified on instruction sheet, as there was no sensor supplied (or shown on BOM) with my kit. I'll leave it open circuit for now (-25C).
I have checked voltages on il1,il2 they are 1.82 and 1.80. I also checked junction voltages R9/C5, R13/C7. Thes showed as 1.82, 1.80v.
How close do the il* voltages have to be to 1.7v, as my values are high. I don't think this is meter inaccuracy, as I've compared the meter to other devices. It's about 0.25v out. Tanks again.... S

[Esbro]

Latest - I disconnected Temp Sensor from board - there is now square wave on all driver inputs. Unfortunately, although the is a signal at the optos, there nothing out. The TVSs removed have been replaced, so I think I'll replace optos on 1 board first, then check again...S

[Esbro]

A final check of the day revealed that there Is a sq wave on the Opto outputs And at the gate outputs but this is only about 1 volt (read from scope scale).

[Esbro]

Update - I now have clean square wave to driver boards. I checked volyages from DC-DC 15.5, 0, -9.2 (all boards similar). Checked voltages and waves on Driver outputs (photo -blue trace). Not sure where to take output readings from, so did Gtop - Etop, Gbot - Ebot got +/- 24v on scope on both (photo - yellow trace). Am I right thinking this is OK ? Thought I'd check before connecting to expensive IGBTs .... Thanks

[johu]

Very good

[Esbro]

Thanks johu. I'll connect up and see if everything works, then connect to the motor.... S

[Esbro]

Hi, I've set everything up and given it a run with the motor connected. The motor doesn't run (or at least I can't hear or feel it running).
I've scoped 2 of the phases and get what looks like sine wave to me. I've also tried a combination of various values for fweak and fslipspnt, ampnom remained at 100. Nothing I do seems to make the motor spin.
I note from the scope, that the voltage on the phases stays at a peak of 9.3v. This looks quite low to me, as the bus voltage is 40v.
Have you any ideas what I can look for next ? Many thanks....Steve

[johu]

Set boost to 10000 and fweak to 20 or so. Assuming it's a 400V motor?

[uhi22]

The scope screenshots say "AC" for the input divider setting. This leads to measurements which are centered around 0V, even if the physical voltage would be unipolar. This makes the interpretation more difficult than necessary. The yellow graph of 16.Nov. shows square between +12V and -12V, this is not the reality, because as I understand the driver is supplied by +15V and -9V.
I propose change the scope setting to "DC" for both channels.
And care needs to be taken where to connect the scope ground. Especially if measuring two phases. E.g. connecting two scope grounds at two different E_top would create a massive short cut, you do not want this. I would propose to use batt- for scope ground when measuring the phase voltages.

[Esbro]

Thanks johu, I'll try that. My motor is 208v.

Hi uhi22, thanks for your comments on the screenshot. I hadn't realised that the setting AC or DC would give a different interpretation.
This is my inexperience using this scope. I will certainly change the settings and try again. Regarding the grounding - I am using the heatsink plate for ground and, at the moment, the -ve from the bench power supply is connected to it. The scope is battery driven.
I hope that these arrangements are suitable. Thanks again for your input. I need any help I can get just now - I've been fighting with this for over 12 months now ! ... Steve

[Esbro]

Hi again, johu 1). I tried to run with boost 10000 and fweak 20/30/40. The result was that the waves were as before, the motor didn't run and, over about 5 minutes, the heatsink (chillplate) rose in temperature to around 35C (at which point I panicked a little and shut it down, as there's no cooling connected). The 10A fuse I had in the battery supply melted (a bit), but the 15A replacement was ok. I let the heatsink cool down over night while the batteries were recharged.

2) Hi again uhi22 ..I changed settings on the scope and, to me, I couldn't tell the difference. The trace still seemed to be centred on '0'. This time I only used 1 channel for safety.

Today, I tried various settings and got a barely discernible hum/vibration from the motor, so I have decided to drop the sub-frame/motor out of the car to inspect the motor on the bench. The wave on the phases is on the photo. settings fweak 40, fsilpspnt 120, ampnom100 and boost 3000. These are all arbitrary numbers - as in what the hell.....
Taking the the motor out was my least favourite option. Why ?.. see photos. but it looks like that's what I have to do.

Thank you both, so much for advice and hints

[johu]

Well fslipspnt=120 ain't gonna do much, it dumps 120 Hz onto the stationary motor and the latter converts it to sound Put 3 Hz or 5 Hz and slowly increase as the motor spins up (if it does)