openinverter forum

joromy wrote: ↑Tue Oct 29, 2019 6:57 am

DrJeff wrote: ↑Mon Oct 28, 2019 1:55 pm Help needed - I'm open to any ideas no matter how wild - regarding these symptoms.

I would drop the evse and test it in manual mode, connected directly to mains without contactors.
Then you can try without ground from the mains (if mains ground is a problem)

This is good advice... eliminating the EVSE will tell us a lot about the issue.

Kevin Sharpe wrote: ↑Tue Oct 29, 2019 9:32 am

tom91 wrote: ↑Fri Oct 25, 2019 9:34 amThe EVSE signal interpretation is not the issue, if the logic board says it is sending a 16A per module command the modules only see canbus commands that says 16A.

Tom, are you aware of anyone who has tested >32A single phase?

No I am not, but with an EVSE signal generator, a crude one, I did measure the full range and look at the control board response and found no wierd behaviour.

tom91 wrote: ↑Tue Oct 29, 2019 12:54 pm

Kevin Sharpe wrote: ↑Tue Oct 29, 2019 9:32 am

tom91 wrote: ↑Fri Oct 25, 2019 9:34 amThe EVSE signal interpretation is not the issue, if the logic board says it is sending a 16A per module command the modules only see canbus commands that says 16A.

Tom, are you aware of anyone who has tested >32A single phase?

No I am not, but with an EVSE signal generator, a crude one, I did measure the full range and look at the control board response and found no wierd behaviour.

Many thanks

Kevin Sharpe wrote: ↑Tue Oct 29, 2019 9:35 am

joromy wrote: ↑Tue Oct 29, 2019 6:57 am

DrJeff wrote: ↑Mon Oct 28, 2019 1:55 pm Help needed - I'm open to any ideas no matter how wild - regarding these symptoms.

I would drop the evse and test it in manual mode, connected directly to mains without contactors.
Then you can try without ground from the mains (if mains ground is a problem)

This is good advice... eliminating the EVSE will tell us a lot about the issue.

I can easily enough connect the charger to 120VAC directly, getting a 240 line will take a little effort. But I think the 120Vac should do the trick as far as eliminating EVSE issues.

Jeff

Kevin Sharpe wrote: ↑Tue Oct 29, 2019 9:24 am

DrJeff wrote: ↑Fri Oct 25, 2019 4:13 amSo I stuck an oscilloscope on the control pilot line from the EVSE (with a reference ground back to the GND on the charger). When the EVSE was not activated the amplitude of the square wave was 20V (??? expected 12V). When the EVSE was activated the square wave amplitude fell to 17V. I only use the oscilloscope once in a blue moon so please correct me if I'm reading this wrong.

Something wrong here. Can you measure the pilot signal and ground at the J1772 connector without any connection to the charger? You're looking for a solid +/- 12V signal.

See Basics of SAE J1772 (here)

When connected to the Proximity line on the controller card, I'm getting +9-12V which is consistent with the EV Connected Ready signal. When I disconnect from the Gen2V3 card (so EVSE connected to nothing I get +12V - no square wave. I used the GND connection from the EVSE to baseline these measurements. Tomorrow I'm going to check the consistency between the GND connections in the garage (I saw a slight difference when measuring yesterday which used a GND from a separate line. All the lines are all feed from the same sub-panel.

Jeff

DrJeff wrote: ↑Wed Oct 30, 2019 3:04 amWhen connected to the Proximity line on the controller card, I'm getting +9-12V which is consistent with the EV Connected Ready signal. When I disconnect from the Gen2V3 card (so EVSE connected to nothing I get +12V - no square wave.

As you say those are states A and B and are correct (see image below). Can you also confirm that you get a 1kHz pilot signal with a swing +6V / -12V during charging? Also verify the duty cycle during charging is 66%?

DrJeff wrote: ↑Wed Oct 30, 2019 3:04 amTomorrow I'm going to check the consistency between the GND connections in the garage (I saw a slight difference when measuring yesterday which used a GND from a separate line. All the lines are all feed from the same sub-panel.

I'd recommend using a single car battery to power the BMS and charger. I know this seems like a none issue but the two things I can see that are different in your setup are power supply and 40A single phase

I haven't been able to get me tesla gen 2 charger working. I've set the ac current to 5amps, single phase, and termination / dc voltage to 380v. The pack voltage is 340v. When i plug in my j1772 charger, the 3 modules light up and I hear a click, but then a couple seconds later, there is another click and the 3 modules no longer light up. The j1772 charger says it is charging, but no current flows at all. Any ideas on what it could be? Faulty charger?

PROGRESS!!! 2/3, but still better

Thanks guys for the help - finally some progress that points towards a cause !!!

I replaced both PSUs (feeding the 12v/GND for the Charger AND 12v/GND to the Contactors controller) with a single car battery, feed through a switch and a 15A fuse. This allows 2 out of 3 modules to charge (any two modules concurrently charging, but not three). So the charger must be doing some kind of check (current leakage? power availability?) that if not passed prevents the requested charging from initiating. As additional modules are requested either the sensitivity of the test is increased or more likely whatever 'error' the charger is testing for passes a threshold. It could be that swapping from PSU to Car Battery made the difference or just messing with the cables and connections was sufficient.

The odd need for a ground connection between the charger/HVJB cases (grounded to the EVSE) and the GND from the car battery is still necessary for the Gen2V3 card to be able to tell if the EVSE button is pressed, connected, or unconnected. When that ground is missing - the Gen2V3 card oscillates from Pressed to Connected. I suspect that the GND connection from the EVSE is connected to the Model S grounds just after the charge port. So while this looks like an odd ground connection in my setup, it is normal for the Model S (EVSE ground connects to car's body ground rather than to the HVJB case).

Also, I have a laptop with USB connections to the Gen2V3 charger card, and also to my Battery Contactors controller. The 2/3 charging symptom occurred whether or not the laptop was also plugged into 120v power (i.e. battery vs wall plug).

So some things to follow up on next...
1. Further cleaning up / increasing cable size on the 12v lines.

Correctly grounding the charger case, 12V ground and EVSE/AC ground to the same point is curcial for this all to work.

All EVSE signals are ground referenced, also the modulels have leakage detections and earthing checks.

DrJeff wrote: ↑Sun Nov 03, 2019 3:10 am PROGRESS!!! 2/3, but still better

That is good news

Does your EVSE have the ability to generate a pilot signal below 40A? Some have configuration switches for installations that can't support that much current. Would be interesting to know whether the system behaves the same when connected to 30A EVSE.

tom91 wrote: ↑Sun Nov 03, 2019 1:58 pm Correctly grounding the charger case, 12V ground and EVSE/AC ground to the same point is curcial for this all to work.

All EVSE signals are ground referenced, also the modulels have leakage detections and earthing checks.

Yeah, thats critical

Do we have a way of detecting the Tesla power module firmware version or board revision via CAN?

Rx7FD wrote: ↑Thu Oct 31, 2019 6:52 pm I haven't been able to get me tesla gen 2 charger working. I've set the ac current to 5amps, single phase, and termination / dc voltage to 380v. The pack voltage is 340v. When i plug in my j1772 charger, the 3 modules light up and I hear a click, but then a couple seconds later, there is another click and the 3 modules no longer light up. The j1772 charger says it is charging, but no current flows at all. Any ideas on what it could be? Faulty charger?

What happens if you set AC current to 10A? 5A is below the J1772 minimum threshold and you might be hitting a restriction set in the Tesla firmware.

Roadstercycle wrote: ↑Thu Oct 24, 2019 7:17 pm Are we still rolling the dice on buying a Gen 2 charger and crossing our fingers it works? I'm just about at that point and I'm not overly thrilled with running an ELCON charger.

Damien has two controller boards for me and i'll try to spend some time testing the chargers that we have next week.

Some good news on board supply. I have now acquired a much more reliable pcb manufacturer here in Ireland and boards will be in stock for immediate dispatch. No more silly delays and boards are being built on an automated smd line so quality is much improved.

tom91 wrote: ↑Sun Nov 03, 2019 1:58 pm Correctly grounding the charger case, 12V ground and EVSE/AC ground to the same point is curcial for this all to work.

All EVSE signals are ground referenced, also the modulels have leakage detections and earthing checks.

SUCCESS!! All 3 modules Charging.

Some additional changes this morning - now all 3 modules are charging...
1. Created a 'ground bar' to which all the GNDs connect (EVSE, HVJB Case, Charger Case, Battery Mgt System [Contactors], Charger inputs]
2. Wired the EVSE ground directly to the ground bar (not to the HVJB case)
3. Cleaned up the connections from the 12V car battery to the Charger controller

A few little things to follow up on...
1. The Gen2V3 controller is still oscillating between button pressed/connected, but charging still works? (I guess it catches a connected state). This may say something about the ground connection between the charger controller card and the charger.

YouTube video here

So, all along it was the mess of your testing setup?

DrJeff wrote: ↑Sun Nov 03, 2019 8:04 pm SUCCESS!! All 3 modules Charging.

Congratulations

tom91 wrote: ↑Sun Nov 03, 2019 8:26 pm So, all along it was the mess of your testing setup?

That's a bit harsh... I can't find anything in the Zero EV user guide about the importance of the charger grounding scheme

Clearly we need to improve our documentation... maybe a charger Wiki would be helpful?

I have solved the missing ground on the board by soldering a small wire from one of the ground screw holes to ground on the board.
On the original Tesla board, all screw holes are ground.
But I guess this is fixed in ver. 4 boards

Kevin Sharpe wrote: ↑Sun Nov 03, 2019 2:04 pm

DrJeff wrote: ↑Sun Nov 03, 2019 3:10 am PROGRESS!!! 2/3, but still better

That is good news

Does your EVSE have the ability to generate a pilot signal below 40A? Some have configuration switches for installations that can't support that much current. Would be interesting to know whether the system behaves the same when connected to 30A EVSE.

Unfortunately, there's no config with my EVSE, it's 40amp or nothing.

tom91 wrote: ↑Sun Nov 03, 2019 8:26 pm So, all along it was the mess of your testing setup?

Yep, sure looks like it - one of the downsides of learning as you go. Of course trying to 'debug' something like this is a bit of a crap shoot... is it the charger version (couldn't get a clear steer on whether anyone else had my version working), potential firmware differences, singlephase/type1 potential issues, etc.

But in the process I've learned a lot. I understand the charger code in detail, got to test my battery control module many times, I understanding the charging process. All good because that's what I set out to do building the TesLorean - i.e. learn a lot. Now that I can charge the battery I can go on to testing the drive unit, PTC heater, coolant heater, AC, etc.

What has been great is the support and ideas from the forum.

Thanks
Jeff

joromy wrote: ↑Sun Nov 03, 2019 11:30 pm I have solved the missing ground on the board by soldering a small wire from one of the ground screw holes to ground on the board.
On the original Tesla board, all screw holes are ground.

I don't understand... are you saying that the open source PCB is missing some ground connections that are present on the original Tesla board?

joromy wrote: ↑Sun Nov 03, 2019 11:30 pm But I guess this is fixed in ver. 4 boards

I don't see any bug reports on Damien's github so unless he's a mind reader I suspect all versions of the board are the same.

Kevin Sharpe wrote: ↑Mon Nov 04, 2019 12:53 am I don't understand... are you saying that the open source PCB is missing some ground connections that are present on the original Tesla board?

The original Telsla board has ground in all screw holes. That is normal on most boards with a metal casing.

joromy wrote: ↑Mon Nov 04, 2019 7:44 am

Kevin Sharpe wrote: ↑Mon Nov 04, 2019 12:53 am I don't understand... are you saying that the open source PCB is missing some ground connections that are present on the original Tesla board?

The original Telsla board has ground in all screw holes. That is normal on most boards with a metal casing.

I've opened an issue on github (here) so that Damien is aware and can implement a change or ignore as he wishes.

Firstly I'm delighted Jeff has his charger up and running. We spent a lot of time going back and forth as to what would be causing the problem.

On the subject of grounding. I have mentioned this multiple times on multiple platforms. The Tesla chargers are very sensitive to grounding. The case MUST be connected to vehicle 12v ground AND evse earth/ground when charging. Otherwise problems arise. The open source logic boards do not implement ground on the screw holes. Yes I do know that Tesla do. My reasoning is born of many long days and nights spent chasing ground loops and emi. I have 2 Gen 2 chargers in 2 of my cars charging almost daily with no problems. Is this exhaustive proof of a working design? No, of course not. If people feel this is necessary then please by all means download the design files, make the modification, make some boards , do some testing and report back.