Just Another HV Battery Thread and Experiments

Here we go. I mentioned in another thread that I was going to do some experimenting over the next week or so. Here's some background.

A fellow forum member in Tennessee with a 2005 Prius was having some HV battery problems. He and I got together in Atlanta on Saturday, 10/7/2017 and we installed a TMR-JWAP Gen 2 battery built with Gen 3 modules. The battery we replaced was a Falcon Hybrid Solutions rebuilt/refurbished battery that was reported to be about 18 months old, and purchased by the previous owner. If anyone from FHS can verify this, contact me and I can provide the numbers on the seals. After the swap, we each headed back to our home turf.

There were some new threads on the forum about batteries again which made me start thinking about this battery. I've decided to do some experiments, as this is a great opportunity to show a simple module swap and also show "reconditioning". Personnally, I believe both methods can be successful, if enough attention is paid toward matching likely capacities. Here's my current plan, which may change as time progresses. Feel free to provide input.

1. Visually inspect the battery
2. Check the torques on the module and electronics fasterners
3. Document module voltages (after the battery has sat for 24+ hours)
4. Install the battery into my car
5. Drive it a bit to observe
6. Load test it and post graphs
7. Remove it and replace any bad modules. I'm going to just grab random modules from stock that haven't been tested, but match near the voltage and age. Perform no other module work or balancing, etc.
8. Install in car again, do about a 25 mile drive and load test again.
9. Remove battery, cycle test all the modules "recondition"/balance/reassemble with matched pairs.
10. Load test again.

I'll be posting all graphs and info as it occurs.

If there's something you would like to suggest, let me know.

 

The visual inspection was a bit disappointing. It was actually worse than the Dorman battery I swapped out a few months ago. I pulled the cover off the case. The case was pretty beaten up. A strip of tape went the length of the battery to label the blocks. All the modules had marker writing on them, the foam rubber seal on the sensor wire side was not installed, the sheet metal where the HV cables enter was bent back for some reason, of the three thermal sensors under the modules, the one furthest from the ECU had both mounting tabs broken off and was just laying on the bottom of the case, the middle one was only connected by one of the two tabs and the one closest to the ECU had the wires bent down at a 90 degree angle. Neither of the wire supports were installed. The copper bus bars are nickel plated. Minimal corrosion on the terminals, just a bit of green here and there. I did not disconnect any wiring plugs to check for corrosion. One module mounting bolt was broken off in the module. I verified the torques on the battery terminals and got anywhere from 1/8 to 1/4 turn on each fastener in order to "click" my torque wrench. (48 in-lb)

 

Here are the initial module voltages. Notice one of the modules in Block 9 is a bit lower.

 

The battery was installed and here's the Techstream display for it. Take notice of Block 9.

 

Took it for a drive and the battery display bars were all over the place. It would go from green to purple to blue a few times over the 5 miles or so that I drove it. Did a load test after I got back. Here's the chart that shows module voltages during the discharge, a chart that shows delta V between the highest module and lowest and than a chart of the data. The discharge was going ok when the capacity display suddenly dropped about 30% and the engine kicked on. Each block has it's own voltage curve displayed. Note Block 9 again. In a perfect battery, all 14 lines would be exactly on top of each other and would appear to be a single line. Although delta V is pretty wide, they're all tracking with each other until block 9 just falls off a cliff. As soon as that happened, the triangle and associated alarms/battery failure codes appeared.

 

Here's what your dash will look like. Disregard the maintenance light. My oil change came due about 50 miles into my drive to Atlanta.

 

If you have the Hybrid Assistant app, this is the screen you'll see. If you have techstream, the second photo is the freeze frame data stored in the ECU. Again, take note of Block 9 voltage.

 

I agree. And yes, my back is feeling it, lol.

That Falcon pack is now removed and one of my TMR-JWAP battery packs is back in again.

The next step will be to remove the failed module from block 9 and install a semi-random module from my stock. Slap it back into the car and try it all again. The pack is on the workbench, so hopefully I may be able to complete that tomorrow. Why isn't it ever just a Block 1 module?

 

My mistake on Post #8. That is one of the screens from the OBD Link LX app, not the Hybrid Assistant app.

 

Here we go on day two. I took today off from work as a vacation day due to having a Dr appointment, so I have some spare time to play.

I started the disassembly. These are the fasteners from the non-sensor side of the modules. Strange that 4 are OEM and 24 are aftermarket serrated style. Notice the corrosion. The OEM style, when torqued correctly seem to stay very clean on the surfaces that are crushed together. These....not so much.

 

Spin the battery around to work on the sensor side. Noticed some repair work on the wires. NBD as long as it's done well. These are the fasteners removed from the sensor wire side of the modules. Again, strange that one is OEM and 27 are serrated style. All bus bars are nickel plated, in case that hasn't been mentioned. And of course, no foam seal on this side.



.

 

I was planning to just remove the bus bar/sensor wire assemblies, but wasn't able to because the grounding strap was actually fastened UNDER the ECU fastener mount, which required the ECU to be removed. After removing the ECU, I found one of the relay safety covers. During the battery swap, after seeing the TMR battery, the owner had mentioned that he had never seen the plastic covers before. (he had accessed the relays previously to install a grid charger) The second safety cover was nowhere to be seen.

 

So..I was trying to do this in a way I thought the average home DIY person would do it. Then I decided to get back into a more normal (to me) routine. I removed the busbar assemblies and the electronics section of the case. Removed the 2 nuts that hold the pack to the case floor and removed the pack from the case floor. I then stood the pack so the fan end is down and ECU end is up. This makes it easy to take off the thermal sensor wire harness. As noted previously, the three bottom sensors weren't installed very well.

 

Now....check this out. Here's a photo of the bottom of the pack with sensors removed. The 2nd and 3rd modules from the top are the Block 9 modules. See anything odd in the photo?

 

After removing the clamps and sliding the modules out to get to Block 9, this is what I saw after Block 8 modules were removed...



The Block 8 modules were unmarked and had no visual evidence at all that anything had happened to Block 9. Also note how all the soot marks flow downward, exactly how the cooling air flows.

 

Here's another example of why a module voltage doesn't tell the entire story of its overall health.

 

I grabbed a module off the shelf that has been sitting there for a few months. It was from a 2005 pack I was parting out on ebay. The modules had all been cycled and individually load tested. I sold about half the modules, but never renewed the listing when it expired. The remaining half of the pack has just been sitting on the shelf waiting on me, lol. I just got busy with other things. Although it is very likely better than the typical ebay module, I'm going to reassemble the pack using it. I have the test results from it somewhere, I'll see if I can hunt them down and post them also. I don't even think I'm going to match the voltages. The modules in this experiment pack average around 7.8, this replacement is around 7.6x, so we'll see how it goes.

 

I've been assigned puppy baby sitting duty for a bit, so I'll have to end it here for a little while. Next up will be reassembly and installing it back in the car. May not get done until tomorrow, as the wife wants to spend some time together today.

We'll end this today on a happy note.........the first 2 pictures are from about an hour ago, the last 2 pictures are right now.



The future King of the Cul-de-sac. 6 weeks and 5 days old.