openinverter forum

Anyone with persistent overcurrent just flip the sign on ocurlim.

Jack- So you’re saying if it comes with -2500 value, swap that to +2500?


Retrevnoc-

Not sure if this is related, but when I first hooked up a pedal I used One from a chevy volt. It had 5v, high reference, low reference in the diagram (instead of signal and ground). The pedal didn’t work so I pulled one from one of my parts sprinters that had the traditional 5v, sig, ground and it worked just fine. Not sure how the Prius pedal is setup but it’s something to consider.

Ok. I received an email last night requesting assistance with a persistent overcurrent issue. Before I get to that let me once again reiterate that I do not provide one to one support of any kind for these or other boards. I simply don't have the bandwidth. The reality is that if I answer every question and solve every problem then nobody will have an impetus to learn or support others and once I eventually stop selling boards it ends there. That's not the plan. Case in point everyone wants a nice user manual but no one wants to do it. Well, I won't be doing it. So step up , step back or go to a commercial vendor.

Now to the question of what happens if we see a persistent overcurrent error.

1) Powerup with only 12v. No hv. Is the error there? if yes then change the sign of the parameter ocurlim. Some versions of firmware require this to be negative. Newer versions it must be positive. When testing I flip this sign to simulate an error and sometimes forget to flip it back before saving parameters.

2)If the error is present when only 12v is applied regardless of ocurlim parameter polarity then one of 2 things are at fault. The logic board or the inverter. To determine which is the problem requires some diagnostic testing of the signals coming from the current sensors to the logic board and the value of voltage at 2 test points on the logic board.These test points may be found in the bottom right hand side of the board adjacent to R45 and R58. Using a multimeter on DC volts measure the voltage between each test point and 12v ground. At idle (no hv no throttle no run signals) a voltage of approx 1.65v should be present here.

3) If the error occurs when starting the inverter , then remove hv , set udcsw to 0 and try to start with no hv. If the error occurs then most likely cause is a faulty inverter power section.

4)if the error occurs only with hv present then disconnect the inverter from the motor and try again. If the error is still present then most likely cause is a faulty inverter power section.

I could go on with various other tests but thats whet I can think of right now. Wouldn't it be great if this got added to the wiki or the user manual ....

Jack Bauer wrote: ↑Wed Apr 01, 2020 11:31 am Wouldn't it be great if this got added to the wiki or the user manual ....

I have added a Trouble Shooting page to the wiki and added the overcurrent procedure described above. I'm hoping to go back and add more of the trouble shooting procedures that have been discussed as time allows.

much appreciated:)

Thanks for the information. I will test the points on the board, test the switching without HV and let the forum know the results.

In the future I would be willing to make a comprehensive user and install manual pdf once I become more expert on all the features of the controller.

wigman wrote: ↑Wed Apr 01, 2020 5:45 pm In the future I would be willing to make a comprehensive user and install manual pdf once I become more expert on all the features of the controller.

Sounds great... maybe start with the Wiki?

I’ve completed the testing and here are my results-

Ignition on (no hv, no start signal)-
Test point #45 = 1.76vdc
Test point #58 = .202vdc
error codes-
Pwmstuck
Hicurofs1
Tmphmax

Ignition on, start signal no hv-
Error codes-
Tmphmax
Current limit
This time (for the first time) it didn’t trip the main contactor with an overcurrent code

Ignition on, start signal with hv-
Error codes-
Tmphmax
Current limit
This time (for the first time) it didn’t trip the main contactor with an overcurrent code

Put it in drive- it spins! Throttle works

Put it back in neutral.

Out of curiosity, in the the run state I recheck the test points with a multimeter-

45 = 1.9vdc
58 = as soon as I touch my multimeter lead to the point the main contactor Opens and error code overcurrent appears.

Redid the test three times and every time I try to test #58 point the contactor opens. #45 always reads the same voltage without any issues.

Could I have a bad trace/solder point or device causing no voltage where there should be voltage and causing the recurring fault?

Please advise. Thanks!

wigman wrote: ↑Thu Apr 02, 2020 1:45 pm 58 = as soon as I touch my multimeter lead to the point the main contactor Opens and error code overcurrent appears.

Redid the test three times and every time I try to test #58 point the contactor opens.

Can you make sure all your connections to the inverter connectors are solid especially the grounds? Can you video your setup so that we can review what you are doing?

wigman wrote: ↑Thu Apr 02, 2020 1:45 pm Could I have a bad trace/solder point or device causing no voltage where there should be voltage and causing the recurring fault?

Anything is possible, if you purchased this board from Damien then it will not have been tested in a LDU.

A high resolution photo of the inverter board might help us spot a problem.

Here is a short video of the setup

Please ignore the sloppy wiring, I left everything long for going in the vehicle.


I checked and all my connections are good. O ohMs from the 12v battery grounds to the wires that go in the 23 connector.

Attached is a hdr photo of the board. Tested pin again and now it has .037v and it’s back to overcurrent stop as soon as the start signal is sent.

wigman wrote: ↑Thu Apr 02, 2020 8:01 pm Please ignore the sloppy wiring, I left everything long for going in the vehicle.

First comment... I see you're using EXRAD shielded cables and the shield is not stripped clear of the core conductor (most obvious on the first contactor visible in the video). Are you sure the shield isn't shorting to ground somewhere or even providing an intermittent contact for power?

Other than that, I can't see anything and hopefully someone with LDU experience will step in now

wigman wrote: ↑Thu Apr 02, 2020 8:01 pm I checked and all my connections are good. O ohMs from the 12v battery grounds to the wires that go in the 23 connector.

Second comment... It looks like you're using the Tesla wiring loom. Are you sure that loom is compatible with the open source inverter pinout? I thought it had changed

Edit: OEM and OS pinout comparison;

https://github.com/damienmaguire/Tesla- ... NEW_V2.pdf

Third comment... in the video the 23 pin connector appears to be floating in space (not screwed down)... are the two metal tabs on the 23 pin connector PCB ground connections?

1. The “shielded” cable is not shielded. I ran out of 2 OT wire terminals and Autozone only had a 2 gauge terminal so that’s positive wire that wouldnt fit in the terminal. I thought it would be fine for testing since the amp draw is much lower.

2. I used the wiring diagrams in the beginning of this thread and corresponded the numbers on the harness not wire colors with the components. I’m pretty sure every thing is connected fine because every other function (d,n,r, ac pedal, ground contactor, ignition switch, main contactor, precharge relay, the board power) works correctly. I didn’t change the stock encoder wires but I double checked to make sure the numbers lined up with the diagram and the one time the motor spun the accelerator worked as it should in D and R.

wigman wrote: ↑Thu Apr 02, 2020 1:45 pm 58 = as soon as I touch my multimeter lead to the point the main contactor Opens and error code overcurrent appears.

Measuring the voltage on R58 should have no impact on the inverter operation. Can you try another meter to ensure this isn't a faulty meter? Can you also check what happens if you move the negative meter lead to different ground connections?

Sounds like the current sensor input is floating if touching with a probe has any effect?

There are two smaller tabs on the bottom that ground to the case cover when it’s put back on. There is a longer tab on top that attaches to the inverter via a allen head machine screw. The top tab connector is attached to the inverter.

Due to the Covid-19 restrictions some areas are not accepting postal shipments. This is causing delays in my shipping timetable for Tesla and other boards to customers. Where possible alternative arrangements will be made with private couriers (DHL etc) to effect delivery. If no delivery service is available then the item will be retained until normal service resumes. If you are waiting for an item and feel this may have affected you then please contact ino@evbmw.com to check with me.

Looking at the schematic and pcb layout here are my thoughts-

The signal from each current sensor comes in connector 13 to IC13
Then the signals are sent to the current filter section to IC11
Both IC13 and IC11 are mirror devices, as in they have a common 5v and gnd and they run an A circuit on one side for current A and a B circuit on the other side for current B using the same exact circuit configuration.

In my case current B isn't giving the voltage reading on the test point but current A is.
I would assume the 5V and gnd is working fine because circuit A has correct readings.
That would leave four possibilities-

1. One of the two IC devices are bad on the B current side
2. The signal incoming signal from current B is not there or whatever generates that signal is broken
3. One of the resistors or capacitors in the filter circuit on the B side are bad
4. the ground connected to R57 is bad

There are test points above and below IC13 that would tell me values for each circuit. If anyone knows what those values should be that would help in my diagnosis.

thanks

also, I'm not an electrical engineer so if I'm way off please correct me

I figured out the issue.

After testing the current sensor A circuit I discovered pin 2 on harness 13 does not have 5V.

For proof of concept, I jumpered pin 10 on harness 15 that provides 5V for current sensor B to pin 2 and all the values on the test points for the current sensors were equal and the motor would start and spin without throwing an overcurrent code.

I still get tmphmax derate and current limit derate codes

After I took the board off I was able to see why there was no 5V on the pin. The connection between the harness socket and the board was not connected as is displayed in the video below. Hopefully I can get a new board that will solve all of these issues.

Excellent work fault finding. This is the sort of thing I'm trying to encourage because although painful it ensures a level of knowledge about the circuit that will be invaluable for future support. It's kinda hard to see for sure in the video but is the pad lifted from the board or is the solder joint between the pad and pin broken?

Regards the derate errors are they warnings or errors? Warnings would be normal as not all motor temp sensors are implemented in the code.

wigman wrote: ↑Sun Apr 05, 2020 1:24 am I figured out the issue.

Congratulations

I believe the pad is still connected and the solder joint is broken, but I will take a close look today and confirm. Here are photos of the exact errors I’m getting even when jumpering the bad connection.

For the temp, I can see it toggle between 100C (maxed out) and 23C (actual temp) so I’m assuming the overtemp code is because of the toggling to max temp every couple seconds.

Hmmm...sounds like you may have another open solder joint on one of the heatsink temp sensor pins. It would be worth checking on the igbt driver plug on the logic board.

Also check the mux select pins are both toggling if you have a scope. This looks like issue I had, the randomness was from the temp values changing rapidly as the wrong inputs were read in - then being sampled for the graph at a lower rate.