openinverter forum

Very much looking forward to getting my board! Thanks Tom

bulletbug wrote: ↑Wed May 27, 2020 6:48 pm

Bryson wrote: ↑Wed May 27, 2020 6:27 pm Interesting that your AC current at both 120 and 240V bounces between either 10.67 or 19.20 amps. I’d start with the AC current calc and work backwards. If I’m understanding correctly, AC current is back-calculated from DC current?

Something that's not shown in these logs is that I've seen AC current (Input side) showing at 2.13 amp when AC line is completely unplugged. I'll start looking at your suggestion. I'm looking forward to getting Toms feedback as well.

Thanks for the input.

How have you grounded your charger, the line voltage gets reported correctly however there is no frequency reported atleast as expected.

Also the charger jumps into a state and stays there, right after startup, you would expect something a bit more of a stair case like.

tom91 wrote: ↑Wed May 27, 2020 7:22 pm

bulletbug wrote: ↑Wed May 27, 2020 6:17 pm
Ok. I've rerun logs using the higher voltage charger. The EVSE is a Juicebox brand with wifi reporting. Watching the related app I'm able to get what I consider to be more accurate charging stats, voltage, amps, etc. Update cycle for the app seems to be in the sub-second range since it reports the HZ of the AC line

Observed Amps: 0.8 amps stable

Initializes and moves directly to .8 amps and holds steady. I've rerun it several times, resetting the board, unplugging charger etc. and get the same results.


SavvyCan527.csvSettings527.rtfSerialLog527.rtf

Please report on accurately I see you mention AC read outs from the Juice box.

So the Juicebox tells you the AC current correct, and this is 0.8 Amps is it not?

Can you please see what it draws when you set the AC limit setting to 10 instead of 5. I might be the stupid command formatting is changed.

Please also do not speculate how values are calculated as all displayed feedback values are purely translated from canbus there is no other math happening. So this looks like an exercises of rescaling/checking all values reported AND sent.

With these settings:

Settings Menu
1 - Auto Enable : ON
2 - Modules Enabled : 2
3 - Can Mode : Off
4 - Port Type : 1
5 - Phase Wiring : 1
6 - DC Charge Voltage : 400V
7 - AC Current Limit : 5A
8 - CAN0 Speed : 500
9 - CAN1 Speed : 500
a - Can Debug : OFF
b - EVSE Debug : ON
t - termination voltage : 450V
q - To Quit Menu

Juicebox shows:
Status: Charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

Next steps:
1. Unplug j1172 plug from socket (juicebox confirms standby and charge amps go to zero)
2. Through serial commands change only item 7. from 5A to 10A
3. Press control board reset button...confirm serial monitor timestamp numbers have started over
4. Plug in j1772 plug to socket

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Volrage: 247

tom91 wrote: ↑Wed May 27, 2020 8:28 pm

bulletbug wrote: ↑Wed May 27, 2020 6:48 pm

Bryson wrote: ↑Wed May 27, 2020 6:27 pm Interesting that your AC current at both 120 and 240V bounces between either 10.67 or 19.20 amps. I’d start with the AC current calc and work backwards. If I’m understanding correctly, AC current is back-calculated from DC current?

Something that's not shown in these logs is that I've seen AC current (Input side) showing at 2.13 amp when AC line is completely unplugged. I'll start looking at your suggestion. I'm looking forward to getting Toms feedback as well.

Thanks for the input.

How have you grounded your charger, the line voltage gets reported correctly however there is no frequency reported atleast as expected.

Also the charger jumps into a state and stays there, right after startup, you would expect something a bit more of a stair case like.

Grounding:
Ground line from AC j1772 is connected to the charger case via original grounding point.
Negative terminal from 12v Battery is attached to the same grounding screw as the AC line on the charger case
Ground line that comes from Molex external connector is currently held in place under one of the control boards mounting screws.

I've checked continuity between all these combinations, and that is good. That might not be enough to prove this part is solid but that would be beyond my current knowledge.

tom91 wrote: ↑Wed May 27, 2020 8:28 pm

bulletbug wrote: ↑Wed May 27, 2020 6:48 pm

Bryson wrote: ↑Wed May 27, 2020 6:27 pm Interesting that your AC current at both 120 and 240V bounces between either 10.67 or 19.20 amps. I’d start with the AC current calc and work backwards. If I’m understanding correctly, AC current is back-calculated from DC current?

Something that's not shown in these logs is that I've seen AC current (Input side) showing at 2.13 amp when AC line is completely unplugged. I'll start looking at your suggestion. I'm looking forward to getting Toms feedback as well.

Thanks for the input.

How have you grounded your charger, the line voltage gets reported correctly however there is no frequency reported atleast as expected.

Also the charger jumps into a state and stays there, right after startup, you would expect something a bit more of a stair case like.

On your idea regarding the stair case idea...I have noticed that a timer initialization is commented out at line 135:
// Timer3.attachInterrupt(Charger_msgs).start(90000); // charger messages every 100ms

I don't know if this is related but it seemed odd.

bulletbug wrote: ↑Wed May 27, 2020 8:31 pm With these settings:

Settings Menu
1 - Auto Enable : ON
2 - Modules Enabled : 2
3 - Can Mode : Off
4 - Port Type : 1
5 - Phase Wiring : 1
6 - DC Charge Voltage : 400V
7 - AC Current Limit : 5A
8 - CAN0 Speed : 500
9 - CAN1 Speed : 500
a - Can Debug : OFF
b - EVSE Debug : ON
t - termination voltage : 450V
q - To Quit Menu

Juicebox shows:
Status: Charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

Next steps:
1. Unplug j1172 plug from socket (juicebox confirms standby and charge amps go to zero)
2. Through serial commands change only item 7. from 5A to 10A
3. Press control board reset button...confirm serial monitor timestamp numbers have started over
4. Plug in j1772 plug to socket

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Volrage: 247

Thank you, the feedback I am decoding from the canbus is not great, in the status it seems the charger phase says its at a voltage limit (possibly due to scaling issue), if you increase the charge voltage request this is the quickest way of figuring out if the request voltage scaling is incorrect for your charger.

Set it to 550V (If the commanded voltage should scale with the reported back voltage) and do not forget to increase the termination voltage too.

Did you manage a capture of it drawing/putting any current into the battery? (sorry if I missed this)

Please do not worry about code, it pumps out the right messages to make a charger work. I checked the content there and nothing appears wrong or missing.

tom91 wrote: ↑Wed May 27, 2020 8:41 pm

bulletbug wrote: ↑Wed May 27, 2020 8:31 pm With these settings:

Settings Menu
1 - Auto Enable : ON
2 - Modules Enabled : 2
3 - Can Mode : Off
4 - Port Type : 1
5 - Phase Wiring : 1
6 - DC Charge Voltage : 400V
7 - AC Current Limit : 5A
8 - CAN0 Speed : 500
9 - CAN1 Speed : 500
a - Can Debug : OFF
b - EVSE Debug : ON
t - termination voltage : 450V
q - To Quit Menu

Juicebox shows:
Status: Charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

Next steps:
1. Unplug j1172 plug from socket (juicebox confirms standby and charge amps go to zero)
2. Through serial commands change only item 7. from 5A to 10A
3. Press control board reset button...confirm serial monitor timestamp numbers have started over
4. Plug in j1772 plug to socket

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Volrage: 247

Thank you, the feedback I am decoding from the canbus is not great, in the status it seems the charger phase says its at a voltage limit (possibly due to scaling issue), if you increase the charge voltage request this is the quickest way of figuring out if the request voltage scaling is incorrect for your charger.

Set it to 550V (If the commanded voltage should scale with the reported back voltage) and do not forget to increase the termination voltage too.

Did you manage a capture of it drawing/putting any current into the battery? (sorry if I missed this)

Please do not worry about code, it pumps out the right messages to make a charger work. I checked the content there and nothing appears wrong or missing.

I set item 6. to 550v and termination voltage to 650v.

Reset the board with timestamp starting over.

No change to Juicebox readings.

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

In the serial monitor I have never seen the state go to 2. It moves between 0 and 1 as I would expect based on code...but I've not seen it go to state 2.

I believe it is charging...just at the wrong scale.

I did have to change the DC volt scaling number (mentioned earlier) to get the voltage to report correct pack voltage. Bringing it up in case you may have not seen that mention.

bulletbug wrote: ↑Wed May 27, 2020 8:54 pm
I set item 6. to 550v and termination voltage to 650v.

Reset the board with timestamp starting over.

No change to Juicebox readings.

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

In the serial monitor I have never seen the state go to 2. It moves between 0 and 1 as I would expect based on code...but I've not seen it go to state 2.

I believe it is charging...just at the wrong scale.

I have not seen a can log showing me it is actually charging, no changes in Battery voltage. You mention a few posts back that you can get it to do a 5Amp charge at the beginning of a day got a can capture of that?

I would need to see the reported status transition during that charge session to understand if there is an issue in the setup (grounding ect.) or with canbus scaling.

tom91 wrote: ↑Wed May 27, 2020 9:00 pm

bulletbug wrote: ↑Wed May 27, 2020 8:54 pm
I set item 6. to 550v and termination voltage to 650v.

Reset the board with timestamp starting over.

No change to Juicebox readings.

Then Juicebox shows:
Status: charging
Current: 0.8 amps (steady, with occasional move to 0.7 amps)
AC Voltage: 247

In the serial monitor I have never seen the state go to 2. It moves between 0 and 1 as I would expect based on code...but I've not seen it go to state 2.

I believe it is charging...just at the wrong scale.

I have not seen a can log showing me it is actually charging, no changes in Battery voltage. You mention a few posts back that you can get it to do a 5Amp charge at the beginning of a day got a can capture of that?

I would need to see the reported status transition during that charge session to understand if there is an issue in the setup (grounding ect.) or with canbus scaling.

May I clarify some assumptions I have to see if the issue may be "baked" into one of those?

From the serial output line:

Phase 2 Feebback // AC present: 1 AC volt: 248 AC cur: 27.73 DC volt: 380 DC cur: 0.00 Inlet Targ: 60 Temp Lim Cur: 15 25 26 EN:1 Flt:0 Stat:1001101

I'm assuming that:

DC Volt: 380 represents the power modules observation (and reporting via internal can messages) of pack voltage and should match the measurement I would see with a simple voltage meter on the positive most and negative most terminals of my battery pack.

Or should That DC Volt be telling me the voltage that the charger is putting out into the battery?

I have had the battery pack in two locations as I've worked on it. The only situation where I've been able to observe (through clamp meter on negative most post) is not the same location that I'm working now. I have not been able to see anything over a 1 amp reading since moving.

Another Question: Should I see more than 1 red light on the active module at any time? I have one solid light on during our conversation. Earlier today (prior to any logs you've seen) I had all 3 modules activated and one of them flashed an additional light on as a relay click happened. Additional light turned on and off as relay click was heard.

bulletbug wrote: ↑Wed May 27, 2020 9:13 pm May I clarify some assumptions I have to see if the issue may be "baked" into one of those?

From the serial output line:

Phase 2 Feebback // AC present: 1 AC volt: 248 AC cur: 27.73 DC volt: 380 DC cur: 0.00 Inlet Targ: 60 Temp Lim Cur: 15 25 26 EN:1 Flt:0 Stat:1001101

I'm assuming that:

DC Volt: 380 represents the power modules observation (and reporting via internal can messages) of pack voltage and should match the measurement I would see with a simple voltage meter on the positive most and negative most terminals of my battery pack.

Or should That DC Volt be telling me the voltage that the charger is putting out into the battery?

This is the voltage measured at the output of the charger.

Was your corrected scaling around 0.0076293945313 ?

I am really struggling to figure out what is going on with your charger, i would almost get to the conclusion the output voltage sensing circuit is damaged in some way, along with the AC current sensing.

bulletbug wrote: ↑Wed May 27, 2020 9:18 pm Another Question: Should I see more than 1 red light on the active module at any time? I have one solid light on during our conversation. Earlier today (prior to any logs you've seen) I had all 3 modules activated and one of them flashed an additional light on as a relay click happened. Additional light turned on and off as relay click was heard.

You need to hear relays click yes when plugging in the EVSE. as the charger stages precharge.

tom91 wrote: ↑Wed May 27, 2020 9:28 pm

bulletbug wrote: ↑Wed May 27, 2020 9:13 pm May I clarify some assumptions I have to see if the issue may be "baked" into one of those?

From the serial output line:

Phase 2 Feebback // AC present: 1 AC volt: 248 AC cur: 27.73 DC volt: 380 DC cur: 0.00 Inlet Targ: 60 Temp Lim Cur: 15 25 26 EN:1 Flt:0 Stat:1001101

I'm assuming that:

DC Volt: 380 represents the power modules observation (and reporting via internal can messages) of pack voltage and should match the measurement I would see with a simple voltage meter on the positive most and negative most terminals of my battery pack.

Or should That DC Volt be telling me the voltage that the charger is putting out into the battery?

This is the voltage measured at the output of the charger.

Was your corrected scaling around 0.0076293945313 ?

I am really struggling to figure out what is going on with your charger, i would almost get to the conclusion the output voltage sensing circuit is damaged in some way, along with the AC current sensing.

My scaling number that brought the DC Volt to the correct measurement compared to my pack voltage is 0.007804311

tom91 wrote: ↑Wed May 27, 2020 9:34 pm

bulletbug wrote: ↑Wed May 27, 2020 9:18 pm Another Question: Should I see more than 1 red light on the active module at any time? I have one solid light on during our conversation. Earlier today (prior to any logs you've seen) I had all 3 modules activated and one of them flashed an additional light on as a relay click happened. Additional light turned on and off as relay click was heard.

You need to hear relays click yes when plugging in the EVSE. as the charger stages precharge.

So when any one module is happily charging along it will have one red light or 2 on the module board?

When I activate all 3 modules under these settings:
Settings Menu
1 - Auto Enable : ON
2 - Modules Enabled : 3
3 - Can Mode : Off
4 - Port Type : 1
5 - Phase Wiring : 1
6 - DC Charge Voltage : 400V
7 - AC Current Limit : 10A
8 - CAN0 Speed : 500
9 - CAN1 Speed : 500
a - Can Debug : OFF
b - EVSE Debug : ON
t - termination voltage : 403V
q - To Quit Menu

Module 3 has a second red light click on...stays on for 2 seconds or less...I see an up blimp in the amperage output from Juicebox...then relay click sound is heard and light goes out. Amperage goes back down to near zero.

That is just module 3, to do all modules it is 123.

Please try get a can capture.

So from capture 527, just module 2.
You can see the Vbatt is un scaled. However you can see the VboostTarget at 450V, which I believe is a firmware limit.

Due to something not working correctly, most likely there is no current starting to flow for some reason. You see the Vboost shoot up, spike and then a flag of voltage limit achieved being set. Thus most likely something is properly broken.

Here is a can capture using the Module 3 only. Module 1 and 2 do not exhibit the additional relay clicking attempts.

joromy wrote: ↑Mon May 11, 2020 7:07 pm

bulletbug wrote: ↑Mon May 11, 2020 5:35 pm Line 822 (for phase 1) : dcvolt[0] = ((frame.data.bytes[3] << 8) + frame.data.bytes[2]) * 0.0105286; //we left shift 8 bits to make a 16bit uint.

This is very good finding!!
I have the same voltage conversion with correct voltage reading, but with an European model.
Maybe we should make a list with prod version serial number....

Joromy, did this get the charger working?

bulletbug wrote: ↑Wed May 27, 2020 9:50 pm Here is a can capture using the Module 3 only. Module 1 and 2 do not exhibit the additional relay clicking attempts.

Module3Only.csv

Same behaviour unfortunate, but does reset and retry which is not good.

If the DC voltage needed to be scaled to correct it, I'm thinking this implies my charger has either:
* Different CAN messaging at the power module stage or
* Different hardware for detecting voltage and current than other chargers.

Here's a picture of the tag on the charger:

bulletbug wrote: ↑Wed May 27, 2020 10:05 pm If the DC voltage needed to be scaled to correct it, I'm thinking this implies my charger has either:
* Different CAN messaging at the power module stage or
* Different hardware for detecting voltage and current than other chargers.

Here's a picture of the tag on the charger:

20200527_160157.jpg

But you tried that, by bumping the charge voltage and termination voltage and it did nothing....

So I've been seeing basically the exact same symptoms as Bulletbug on 2 different chargers. I have tried adjusting the DC scaling number with no change to behaviour, but I have not yet tried changing the output charge value. I can try that in the near future.

I only have one of those chargers remaining, but it seems odd that mine and Bulletbug's chargers are all broken in the same way.

If I can record CAN captures of my unit's behaviour, would that be of help for further diagnosis? At least then we can see if it's a common error among certain versions of the Gen2 charger.

I do also have a working unit that I've been using for a while. When I swap the open source boards around the working charger stays working, and the non-working one stays non-working.

Cheers,
-Ryan

Cyco90 wrote: ↑Wed May 27, 2020 10:33 pm So I've been seeing basically the exact same symptoms as Bulletbug on 2 different chargers. I have tried adjusting the DC scaling number with no change to behaviour, but I have not yet tried changing the output charge value. I can try that in the near future.

I only have one of those chargers remaining, but it seems odd that mine and Bulletbug's chargers are all broken in the same way.

If I can record CAN captures of my unit's behaviour, would that be of help for further diagnosis? At least then we can see if it's a common error among certain versions of the Gen2 charger.

I do also have a working unit that I've been using for a while. When I swap the open source boards around the working charger stays working, and the non-working one stays non-working.

Cheers,
-Ryan

If you could do a capture of the working one charging and then a capture of the behaviour of the none working one I will happily have a dig through it.

tom91 wrote: ↑Wed May 27, 2020 10:38 pm

Cyco90 wrote: ↑Wed May 27, 2020 10:33 pm So I've been seeing basically the exact same symptoms as Bulletbug on 2 different chargers. I have tried adjusting the DC scaling number with no change to behaviour, but I have not yet tried changing the output charge value. I can try that in the near future.

I only have one of those chargers remaining, but it seems odd that mine and Bulletbug's chargers are all broken in the same way.

If I can record CAN captures of my unit's behaviour, would that be of help for further diagnosis? At least then we can see if it's a common error among certain versions of the Gen2 charger.

I do also have a working unit that I've been using for a while. When I swap the open source boards around the working charger stays working, and the non-working one stays non-working.

Cheers,
-Ryan

If you could do a capture of the working one charging and then a capture of the behaviour of the none working one I will happily have a dig through it.

Tom - You're awesome!!

Thanks so much for all the time you've spent with me today. Hopefully we can keep things out of the trash pile that don't belong there...and increase confidence with people that they can make a working charger without it being what appears to be such a hit or miss gamble.