Troubleshooting HV ECU burnt connector (causes P0AFA, red triangle, 'replace hybrid battery' etc

I am keen to troubleshoot a problem with my 2008 Prius ('Gen 2' / NWH20 / XW20). I hadn't used it much for a few weeks, then when going to use it, it kept charging the battery, and the battery went from full, to one bar - over a few minutes. The red triangle had been on, and stayed on. I attached my OBD-2 wifi tool and looked up the readings. Basically, HV battery blocks 1-2 (which would be 1-4 of the 28 actual block) were flicking back and forth between +15.5volts and -15.5 volts. The rest of the pack was at the same (+) voltage. Force charge (accelerating in drive) was not possible. The car was still driveable. After a short while, the car would no longer start properly; the red light comes on, and 'Ready' flashes on and off 3 times briefly immediately, then stays off. The car won't go into drive.

Some research led me to a number of threads, in particular this thread (please read - it has a LOT of info, without which I couldn't have got as far as I have with my issue; it related to Gen 1 NWH11 cars technically but is beginning to include Gen 2).

The battery fires at ECU sense connector thread | PriusChat

I've posted some stuff on there also (and will link to this thread). This has been happening to 'Gen 1' (NWH11) Priuses, but has now happened to my 2008, so this problem likely covers all NWH11, NWH20, probably the Prius C, and maybe many more.

In my case I expected to find a burnt connector - I got into the battery compartment urgently in case it was to catch fire. I attach files of what I found:

The battery modules were still balanced right before I took the ECU out - the car is parked up, the ECU is in my house for testing.

I attach an (apologies - large) picture of photos of the connector (both from the motherboard side and from the external side of the connector:



You need to expand the image to see the labelled pins. However, image 'A' shows the outside of the connector. Not in great shape. However, all the pins are there. Pin 22 is the most corroded. Of particular note, note of the pins on the outside are actually pitted or worn away at all - just corroded.

Image 'B' shows once view of the connector from the inside. This shows the two major issues - pin 11 (the one above 22) has vaporised at an area near the base Also, pin 5 has really vaporised - not much is left.

Image 'C' shows that pin 22 has almost completely melted but not quite. It is covered in white ?metallic vapour deposits. The top section is still the correct shape, but the lower section is quite melted.

Image 'D' shows pins 5 and 11 again. Image 'E' shows the top of the connector, and you can see that pin 5 is mostly gone, and pin 11 is damaged. The parts that are left are still unmelted, interestingly.

This car was running fine, and I removed the ECU the day after the failure.

 

At the current time, it is unclear why this happens. This connector is the one that the high-voltage battery ECU uses to measure the voltage of the prismatic cell packs. It therefore carries high voltage across the terminals. After entering the ECU, there is a 1000 ohm resistor, which means that the current being drawn is very low in normal circumstances.

It has been noted that pin 22 seems to corrode much more often than the others (so much so that often the pins are pristine, except pin 22). pin 22 is often one that has burnt out, too. This problem with corroding / shorting has caused a few battery fires, so is a serious issue.

An interesting thing about the failure on my car is what has happened to pin 5. It has completedly melted away (thus sustained a high current). It could not have got this current from the ECU (due to the 1000 ohm resistors among other things). Yet, the upper part is undamaged, and doesn't even show signs of heating damage. It seems to have melted from the circuitboard, up. This seems to be similar to what is happening on several other ECUs, but seems to have occurred without lots of corrosion (or without being next to pin 22), which is the 'normal' way this fault occurs.

Looking at pin 5 (specifically, looking close to the missing 'vaporised' piece), there is a white film on some components nearby, almost like a metal vapor deposit. I made a spreadsheet to figure out the voltage differences, to see what may have caused a short-circuit:


This spreadsheet has pins 1-11 at the top, and 12-22 at the bottom. I have made notes about the condition of the pins (on the motherboard side). Red means the pin is broken, orange / yellow describes melting or thinning.

Pins 5 and 11 are two that have vaporised and now have gaps. Looking at their neighbours;

For *pin 5*, pins 6, 15, and 16 show some melting around the base. Looking at 'Image B' earlier, it can be seen that pin 6, next to it, has 'shrunk' at the lower half, with heat damage. Pin 16 (the one underneath / behind pin 5) has also taken a lot of heat damage and partially melted. Pin 15 also has some melting at the base, but is difficult to see.

For pin 11, the pins nearby are also damaged. Image 'C' above shows that pin 22 (the pin behind / underneath pin 11) has taken severe thermal damage - almost completely melting away.

Looking at the spreadsheet, it seems that the pins surrounding the ones that blew also took thermal damage. Some interesting points:
- The damage to the adjacent pins is limited to the areas of the primary pin that 'blew'. The best example is pin 11 (which blew), and pin 22, underneath, which severely melted, but the melting damage is only adjacent to the area on pin 11 that melted.
- On pin 5, the damage to adjacent pins is limited to the lower parts of the adjacent pins
- Interestingly, pin 4, next to pin 5, has some white (metal vapor?) deposits on it - from whetever happened to pin 5. But, it has no other heat damage or melting etc. Pin 6 is quite damaged, as are other surrounding pins. Pin 4 is not actually connected to anything, so cant form circuits with anything.

All of this points to a short-circuit happening suddenly, inside the ECU. My ECU has the classic pin 22 damage but, pin 5 is interesting as it is isolated. It is the original ECU and the car is not a used import.

 

If you got this far, I would be keen to hear your thoughts on what could have caused this. It seems closely related to the HV ECU fires and maybe to a lot of the 'sudden HV battery 'failures'' that we are seeing.

To me, it seems that the adjacent pins caused a short circuit - yet there is nothing 'rattling around' in the ECU. It seems the adjacent pins (the ones that were energised, anyway) were involved - yet other pins (e.g. pin 4) took no damage at all - and this pin isn't connected to anything at all.

My thoughts are that a brief episode of electric arcing is occurring. The short circuit on pin 5 could not have been caused on the actual circuitboard itself, but something caused it - but there is no obvious reason. If an electric arc was to occur (e.g. at the bottom of pin 5 to one of the other pins), then the arc could continue. Other pins could even become involved. Given that each pin has a 31 volt potential difference from the ones next to it (and around 50 volts diagonally), there is enough voltage. And there is certainly enough current from the battery!

I'm waiting for a 'new' (eg second hand) ECU + battery, and will see what state my actual HV battery is in. I want to have a proper look at the 'failed' ECU under a microscope, as I wonder whether this could be related to the coating of the connectors combined with the high voltage + current...

Would be grateful to hear your thoughts / experiences / ideas for further testing!

 

fair humidity, little salt in the air. I'm focusing on pin 5 as that one has no obvious corrosion to make it do what it did!

The car itself, totally rust free, no issues with corrosion inside or out, and all other HV ECU connectors are immaculate!

I really want to figure out this mystery!

Actually, looking back, on the connector photo, on the 'right hand side', the hole for pin 5 is immaculate on the 'loom' connector, so the cause of it shorting so badly is confusing!

 

Quick question - is it safe to unplug the orange connector from the HV ECU - or do you have to remove the bus bar first? I am pretty sure it is still ‘live’ so you really should dismantle the bus bar...

 

Thank you for the info; I still can’t figure out how the pins melted so badly inside the ECU, yet leaving the ECU pristine - and the connections between the pins (even the vaporised ones) is immaculate underneath - no heating damage etc.

I am guessing the HV ECU is toast though from what you see, and also the HV pack itself is destroyed?

Is there any chance I could repair the ECU by removing the corrosion and replacing the vaporised pins?

I removed that orange safety plug before anything else, but what worries me is I know the orange connector has no current limiting components (hence this whole problem), and also, as it senses the battery voltage, I am still worried that it could be ‘live’ even when the orange plug is removed? I will treat it as live until I get a better look at the whole battery - I treated it as live when unplugging it.

 

Actually, following this up; I don’t think my hybrid battery itself has failed because voltage readings were consistent (including across pin 5 which is interesting) - exception being pins 11,22 but it was obvious why this was. I have no doubt the bus bar will be very corroded because it is original, untouched from new!

I am also confused as I don’t think that my battery pack is ground shorted, but if it is was, or if the connector had a short, that couldn’t damage the pins from the connector to the circuitboard? Am still trying to figure this all out!

 

Thank you! I do also find it interesting that edthefox5 has come across many ECUs with just a pin or two blown - something odd is going on in these ECUs!

 

I think a moderator has to do it; I was using my own terminology as am new to this meant 'high voltage electronic control unit' whereas I see HV ECU is the Hybrid ECU (I assume). Will call it battery ECU from now on (when I remember).

What I can do, in fact have done, is picked up a 'spare' hybrid battery. It is sitting in the back of my car (EV). First thing is a battery ECU inspection, then removal of top plate (of the spare). Am tempted to 'swap it in' if that looks good - just have to hope it hasn't drained too much (would use OBD-2, and stop if it seems properly low - and use grid charger instead).

 

OK - I have a dilemma now. I have an immaculate battery computer that came from a second hand battery. That battery has been sitting for 4 months - was in working condition. If the whole replacement unit is charged up enough to fire up the engine, how dodgy is it to then charge the battery in the car (as opposed to grid charging?) - it used to be the recommended way but it seems not any more?

Would be awesome to get my car back on the road, then explore the faulty battery / computer combo in more detail!