HV battery help needed. Where I am at with DIY repairs

The module pairs are numbered 1-19 with #1 being adjacent to the control electronics. It is module pair #18, near the opposite side.

Could you list the voltages you measured and identify the one closest to the control electronics?

How are you measuring the code, specific make and model of scanner?

My first thought is to pull the pack and remove the four end modules. Inspect the inner two closely, wash and dry all four, and reinstall swapping the pairs. If you get P3029, you know the outer pair or at least one of the modules is bad and needs to be replaced. You could split the #18 pair and if P3028 or P3029 shows bad, you'll know which one is bad.

If you have or get a smart charger, see RC hobby shops, you could measure the Ahr capacity as they can go through a charge/discharge cycle. This would also identify the specific module that has failed.

You might check over my experiments:
Prius Battery Photos

There is a high probability that the existing battery modules are aged and suffer from low Ahr capacity. You might consider the "Re-InVolt" solution. Rebuilding the pack on your own is possible but my experiments suggests this needs to be done carefully with tools to test and balance the modules.

Do you have spare transportation or is this your only ride?

Bob Wilson

 

All it means is that the observed state of charge of that module is significantly different than the observed state of charge of the other modules.

It's possible. It certainly hasn't been ruled out.

Remove the busbars and check them, as well as the module voltage sensing wires, for corrosion. This is the long way around, but in the absence of access to live data, it's the first step.

 

That suggests a high charge, higher than normal. It moves the ambiguity group towards the sensor wires and/or ECU. This reading was just after driving the car or after letting it sit overnight?

Does the multi-function display show the battery state of charge show as high level, medium or low level?

The reason I ask is the battery ECU would be reporting the State Of Charge. If you're seeing high voltages while the display show a low level SOC, I would suspect the battery ECU is confused from either a bad sensor wire and/or ECU problem.

I don't know what they charge for a reading but it would be a good idea to ask them to read out all 19 module pairs and compare their voltages versus VOM reading. If the ECU thinks or sees the voltage too low, a broken or corroded sense wire or other fault, it would lead it to try to try and overcharge the pack. This would be consistent with your symptoms.

It also makes sense to ask them to read the temperature readings. The schematic shows five sensors.

The NHW20s, '07, only have 28 modules so they would have combined modules from two for the 38 needed for your NHW11. One way to check would be to read the serial numbers of the modules looking for breaks in the series.

Always worth asking. I suspect another answer but no harm trying. If you're at a Toyota service center, you might ask them to read the codes and all module pairs.

Owch! That is too high.

I think you've already identified a likely fault as likely one of the sensor wires used to measure the 19 module pairs may be defective or possibly the battery ECU. Ohming it out will be tricky ... caution, very high voltages are concentrated where the sense wires collect.

The safe way is to remove the bus bar from the batteries, both sides, and then ohm out the sense wires. This would remove the HV risk. If you decide to check the sense voltages with the battery in the pack NOT RECOMMENDED be careful.

If a sense wire was broken, we would expect to see P3030.

If the voltage sense lines are good, then it would be the ECU. One of the things "Re-InVolt" mentions is replacing the sense lines with a corrosion resistant, gold-coated sense lines. The KOH solution 'eats' metal. I had some copper wire jumpering two modules together and came back a couple of months later to find the wire nearly eaten away.

Bob Wilson

 

The warranty on the replacement battery is 1-year (like any Toyota replacement part), so if it was replaced in 2007 it no longer is covered under warranty.

Your measurement of module voltage near 8V when the car is IG-OFF is not proof that the modules are OK. The question is what happens to the module voltage when the battery is under load, where the current flow is 50A or more. If a module has unusually high internal resistance, the voltage will drop as the current flow increases.

I recall that when I owned my 2001, traction battery voltage routinely got into the 300V range, so I would say that reading is normal.

Given the tools available, I agree with your thought to swap around the last four modules and see what happens. While you are doing that swap, carefully inspect the bus bars and sense wires, and clean off corrosion if present. It would also be a good idea to measure resistance of the sense wires leading from the last two module pairs to the traction battery ECU wiring harness connector, just to make sure that's not a problem.

If the wiring is good and the DTC remains the same after swapping the module pairs, that is proof the battery ECU is bad. If the DTC shifts, then you will know the fault is within the module pair.

Once you've identified that the module pair is bad, then if you need to identify the failed module you could place a load on the module pair. For example if you connected two automobile halogen headlights in parallel, that would draw ~10A load. See which module shows lower voltage. That should be the bad module.

Or, as Bob suggested, you could be tricky and move only one of the suspect pair to a different position within the battery case, and see whether the DTC shifts or remains constant.

 

Such readings can be quite normal, as can be a reading of 7.9V per module. In fact, they are fairly common.

 

One technique, risky, is to throughly block the wheels and put the car in reverse. In a perfect world, with the rear bumper also against an immovable object ... large diameter tree or cement wall. This will put a significant load on the traction battery. But there might be an easier approach.

If you have a non-contact thermometer, you might be able to detect a weak module by measuring their individual temperature after a drive. However, we need to first make sure the battery ECU and sense lines are OK.

The best approach is to have someone with a scanner read out all 19 module pairs so you can compare it to the manually measured voltages. A Graham scanner can read out the minimum and maximum module pair and give the location. Send me a PM if you need to rent one, it will take a week to get it to you. Faster, you might want to check with your local shops on Monday. But Tomfreed reported the Autoenginuity worked.

Bob Wilson

 

Re: UPDATE Where I am at with DIY repairs

Most commercial code readers read only engine codes.

Disconnecting the 12V battery is often recommended on this forum, but it doesn't fix anything. Fortunately for you, your code(s) are resetting immediately. Knowledge is power.

Rather than reset it, you might want to let it remain until you can collect the data that the vehicle has logged. This data is designed by the manufacturer to help you fix the vehicle.

You did make great progress. If the warning symbols bother you, just push the "info" button and select a screen; the large symbols should be replaced by one that is a bit less annoying.

A temporary workaround, of course, but you may find it useful in the interim.

 

Re: UPDATE Where I am at with DIY repairs

Without the codes, we're still blind. Can you return the 'cheap scanner' and/or test others until you find one that shows codes?

Since the next problem showed up after the 12 VDC reset, we know a problem remains. However, three power on-off cycles and the code indication should go away if it is an intermittent. The codes will remain in the ECUs but at least it won't be 'in your face.' Note "should" as it hides the codes if they are no recurring. Don't forget that the previous problem can mask another.

If you were closer to Berkeley or San Francisco, I would recommend a trip to Arts Automotive or Loucious Garage. Both have high reputations.

There is the "jumper" test that can cause the car to flash some of the vehicle ECU codes on the indication lights. At the base of the dash is the OBD connector, DLC3. The one in my car is 'loose' so it may be possible to pop it out so it is easier to see. Amazon sells an OBD extender that brings the pins out so you don't have to hang looking up in the driver footwell:

• short side - also has two small 'tits' smaller than the pin areas and a slightly larger tang that spans two pins.
• long side - has two small 'tits'
• Ts - counting short side from left, this is pin #3
• Tc - counting short side from left, this is pin #4
• CG - counting long side from left, this is pin #5 and should measure the same as vehicle ground
• Tie Tc to CG - read out the flashing codes from the dash warning indicators
• Tie Ts to CG and switch to ACC or LOCK - should erase the codes

Bob Wilson

 

The battery voltage scale is on the right edge. These were tests where the traction battery voltage was recorded during high-speed descents. In both cases, it started at 300 V and ran as high as 350 V but they were under heavy charging currents, ~18 A. Given the 6.5 Ahr rating, this is called a '3C' rate of charge. However, on the flat, the car rapidly switches to a discharge mode to bring the voltage down.

You can see more battery voltages in this chart that shows a forced charge and reverse discharge:

Again, the scale is difficult to read but the 'orange' line is the traction battery in 10 volt units and is read from the left axis. So initially the traction battery varied but holding a mean closer to ~295 V. Then I started the forced charge and the voltage peaked at ~345 V., seconds 2500. I sat still for a brief interval just at seconds 2700 and you can see the voltage was dropping. Finally, I started up the hill in reverse, seconds 2800 to 3100 bring the traction battery voltage just under 300 V.

Sorry I didn't post this two year old data earlier. Regardless, I still see a resting voltage of over 300 V. as being too high. After sitting, it should come closer to 270-280 volts.

What happened to the two bad modules?
Any interest in selling them?

Bob Wilson