Roadster 12V system is down.

[driver_EV]

as MLAUTO wrote before: yes, it's the switchpack underneath the dash-panel, charging the 12V aux-batt.
i can highly recommend you NOT to put just a bigger/stronger battery in your car wich maybe pulls eventually more current while charging, because those switchpacks are quite vulnerable and - besides expensive - a pain in the butt to change (you have to undo half of the dashboard).
when you buy a new 12V battery please contact Gruberpower, they have sufficient alternatives !

i had twice before a 12V failure on a roadster. one was similar symptoms as yours but easy to fix:
when i opened the pem, the red positive cable coming from APS-connector to the board (see pic) had dry solder and lost contact.

View attachment 423439

the other one there was a burned cap on the same board but in the powerline of the PEM-fan. that car didn't loose 12V but complete comms to the CAN-bus.

well, in my eyes that bloody Molex-12V-plug is a highly suspected candidate as well...

anyway, good luck !

Thanks for the info and pics. It is always good to see and understand issues that have happened in the past with the Roadster.

I have completed the repair of the Molex PEM connector. Fortunately, mine was just mechanical damage to the external portion, and the pin condidtion looks fine. I damaged the latch feature because the rubber seal was in poor condition expanded preventing normal operation, causing me to fight with it.

The parts were an exact match, thanks to PV-EV for that ordering information.
Not having experience with this connector, I spent a little time discovering the method of disassembling it.
Photos show the spot to pry on to pop the white part free. That is the gateway to freeing the pins as well.
You need to pop a small tab up and slide the white part out with a small flat screwdriver.

 

[driver_EV]

The pins are easy to release at that point, and assembly is easy.
I coated the rubber seal with a little rubber protection product (shown) hoping for some insurance there.

Put back together it looks and feels like new.

 

[warden_clyffe]

little advice: when you order the crimp-contacts lookout for gold plated

 

[driver_EV]

So I thought I would post an update now here in November.
The upgraded battery is still serving to replace the missing ESS 12V supply via the added fusebox tap I installed.

The car does not die, since control units are provided the missing 12V ESS supply from the 12V battery.
Also, I still have to repeatedly connect the trickle charger, to keep the vehicle drain from discharging the battery while the car is not being driven.

It seems that when driving the car, there is not enough power/voltage level to fully top up the battery. The external trickle charger just keeps the battery voltage from dropping further (below ~12.5V) once the initial charge cycle finishes. So, after driving, I connect the external charger, that goes through it's program to charge the battery, but does not keep it at a fully charged voltage.

Next, I am planning to add a DC boost driven by the APS, that will more fully charge the 12V battery when the key is on and the APS has been activated. I found an appropriate looking board that I can use for that.

To do this I can add one more tap in the fuse box on the APS supply, power the DC boost add 1V or so to get the 12V battery to top up, at least when the key is on.

This board works better than I thought it would. You give it a DC source, and use the trimmer to dial in the desired output. It has two buttons, the first just toggles the digital display on/off (the output remains while it has input power), and the second button toggles the display to show input V or output V.

 

[driver_EV]

As an alternative, I may consider just using the APS fuse connection to charge the 12V battery directly through a rectifier, without the boost.

But I really do like this panel allowing me to see in realtime the 12V level(s) without getting the Fluke meter out and fiddling to check it.

Also, I have been able to remotely check the 12V battery using OVMS:

 

[markwj]

Also, I have been able to remotely check the 12V battery using OVMS

The voltage shown there is DC-DC converter voltage, not the voltage from the little 12V lead acid battery. OVMS is not powered from that battery. Also, not calibrated (although you can calibrate it with parameter system.adc, instance factor12v, default value195.7).

 

[driver_EV]

The voltage shown there is DC-DC converter voltage, not the voltage from the little 12V lead acid battery. OVMS is not powered from that battery. Also, not calibrated (although you can calibrate it with parameter system.adc, instance factor12v, default value195.7).

Understood Mark, However, since the (non-APS) DC-DC converter is dead on my car, the only source of continuous 12VDC now is the 12V battery circuit! I have patched the 12V battery in (fusebox tap) to take place of the failed ESS 12V DC converter.
So, for my car, the 12 volt battery does currently power the OVMS while APS is not active.

-and the external trickle charger I plug in keeps the battery from draining.

Thanks for the information on calibration.

 

[markwj]

UI have patched the 12V battery in (fusebox tap) to take place of the failed ESS 12V DC converter.
So, for my car, the 12 volt battery does currently power the OVMS while APS is not active.
Thanks for the information on calibration.

Ah, ok, that makes sense.

FYI: The calibration functions as following:

  // Allow the user to adjust the ADC conversion factor
  float f = MyConfig.GetParamValueFloat("system.adc","factor12v");
  if (f == 0) f = 195.7;
  float v = (float)MyPeripherals->m_esp32adc->read() / f;

You can set to 1 to get the raw value from the ADC, then divide by your measured voltage, to set the conversion factor. 195.7 is what we measured as average across a bunch of devices, but there is some variance in the resistor divider network that can be calibrated on each individual module.

 

[driver_EV]

Ah, ok, that makes sense.

FYI: The calibration functions as following:

  // Allow the user to adjust the ADC conversion factor
  float f = MyConfig.GetParamValueFloat("system.adc","factor12v");
  if (f == 0) f = 195.7;
  float v = (float)MyPeripherals->m_esp32adc->read() / f;

You can set to 1 to get the raw value from the ADC, then divide by your measured voltage, to set the conversion factor. 195.7 is what we measured as average across a bunch of devices, but there is some variance in the resistor divider network that can be calibrated on each individual module.

Thanks again for that detail. I am planning to study the OVMS manual again. In particular may later want to get an alarm or notification of the battery voltage dropping below ~12.3 or so. That would be an early warning that it is not charging as it should be. This telemetry is very useful.

 

[markwj]

Thanks again for that detail. I am planning to study the OVMS manual again. In particular may later want to get an alarm or notification of the battery voltage dropping below ~12.3 or so. That would be an early warning that it is not charging as it should be. This telemetry is very useful.

You could do that with a small javascript script function. Hang it off one of the timer notifications (such as timer.600 to run every ten minutes), check the voltage metric, and then use the notification system to alert if necessary.

 

[KeithBoxer]

So you have error ID 3003 SHFT: Warning Lost Comms VMS and error ID 3004 SHFT: Warning Lost Comms DMC. Also, you can probably ignore that error ID 1554 as that typically shows up when booting up the 12V system.

Those two errors are both coming from the shifter which could indicate the shifter is having an internal problem, or the shifter's status messages aren't reaching the VMS to indicate all is well. I don't see the correlation to the 12V system issues though. I'll keep thinking about it and see what I can dig up.


That is correct.

I am experiencing the same fault messages as mentioned in the above thread : ID 3003 SHFT: Warning Lost Comms VMS and error ID 3004 SHFT: Warning Lost Comms DMC on my roadster 2.5. #344

I was trying to follow the thread to find out what the eventual fix was, but I seem to be lost. I would be grateful if anyone can provide advice about what needs to be checked to find the cause?

 

[PV-EV]

See posts 20 and 28. If you do not have 12v on the mentioned wire then that may well be your problem also.

 

[KeithBoxer]

Maybe unrelated but when I was having problems all my gearshift buttons also illuminated. Lots of errors and car would not drive or charge. Changing PEM APS plug resolved all errors. No problems in two years now.

Thanks for your latest comments. I noticed this in your thread. I have also similar symptoms with all gearshift buttons silluminated, car will not drive or charge, and various fault messages. How did you identify the the PEM APS plug was the issue? I have tried removing it and I cannot see any signs of damage or wear. Is there any way to test it?

 

[PV-EV]

Thanks for your latest comments. I noticed this in your thread. I have also similar symptoms with all gearshift buttons silluminated, car will not drive or charge, and various fault messages. How did you identify the the PEM APS plug was the issue? I have tried removing it and I cannot see any signs of damage or wear. Is there any way to test it?

the latch on mine was broken and the internal bulkhead connector was cracked. The cost to replace the whole thing myself was low so thats what I did. The results were better than expected. I believe the connector was loose due to the broken latch and cracked body and that was what was causing my problem. The actual pins didn’t look bad. What has Mr Gruber reported?

 

[KeithBoxer]

the latch on mine was broken and the internal bulkhead connector was cracked. The cost to replace the whole thing myself was low so thats what I did. The results were better than expected. I believe the connector was loose due to the broken latch and cracked body and that was what was causing my problem. The actual pins didn’t look bad. What has Mr Gruber reported?

I have sent Peter the vehicle logs and am waiting for his reply. But I notice that on the VDS readout it lists the most recent fault , dated October which I understand is due to a glitch with the GSM, but the rest of the fault messages are earlier see photo....also on the vital stats the temp data from the motor and pem are not registered, does that give any more clues I wonder..?

from August lat year.

9

 

[KeithBoxer]

the latch on mine was broken and the internal bulkhead connector was cracked. The cost to replace the whole thing myself was low so thats what I did. The results were better than expected. I believe the connector was loose due to the broken latch and cracked body and that was what was causing my problem. The actual pins didn’t look bad. What has Mr Gruber reported?

 

[KeithBoxer]

Is this the place to test the 12v as mentioned in earlier thread?

 

[slcasner]

But I notice that on the VDS readout it lists the most recent fault , dated October which I understand is due to a glitch with the GSM, but the rest of the fault messages are earlier see photo....
from August lat year.

I think you meant GPS rather than GSM. It is a limitation of the firmware in the Garmin GPS sensor that causes it to deliver a date 19.6 years in the past if the sensor has reset due to loss of power. A date 19.6 years in the past appears to be a date 137 days in the future if the year number is not displayed and the current year is assumed. I suspect that the faults dated August 16 are not from last year, they are actually from April 1 this year.

 

[KeithBoxer]

I think you meant GPS rather than GSM. It is a limitation of the firmware in the Garmin GPS sensor that causes it to deliver a date 19.6 years in the past if the sensor has reset due to loss of power. A date 19.6 years in the past appears to be a date 137 days in the future if the year number is not displayed and the current year is assumed. I suspect that the faults dated August 16 are not from last year, they are actually from April 1 this year.

Thanks, yes you are right I did mean the GPS rather than GSM, and Many thanks for clarifying. I found online that Gruber Motors have a fix to update the GPS date.

 

[slcasner]

As you may have read in other threads, I made a cable that allows updating the GPS firmware by connecting to one of the two big cables that plug into the VMS in the passenger footwell.