Fixing up a neglected Prius, HV battery investigation

We have a 2004 Prius that kept getting handed down across the family through various events. Went from my wife, to my sister-in-law, to her first son, and now it's waiting for her second son to get his license. Unfortunately they parked it on two flats at my mom's house, key in the cupholder, SMART still enabled. Aux battery was flat. Gave it a last-ditch charge attempt and it managed to pass the test at Autozone (though that guarantees nothing).

Brought it back to my mom's and reinstalled it, hoping the car would start. No good. Everything lights up and you hear various relays and contactors clicking, but the dreaded red triangle. I managed to remember my OBDII reader and got two codes -- P0A80 and P3030. Grabbed the extended PIDs and I could see that blocks 1 and 2 are both below 2 volts, block 8 is 13.8v, and all the rest are 15v.

Pulled the hybrid traction battery out and brought it home with me. Hoped it would just be a damaged sensor wire. Plenty of corrosion but it definitely looks like I have at least four bad modules. First four modules are at 1.2v or less, and module 13 is at 6.4v.

I assume I'll simply need to replace the first four since they're near-zero. I'm more curious about the single 6.4v module -- will this one need to be replaced, can the Prius recharge that on its own, or will I need to manually recharge it? Naturally, I upgraded my RC truck to LiPo and no longer happen to have a NiMH charger handy.

 

Yeah I suppose you're right. The extra $25 for a fifth module is worth it to avoid dragging this home again, given the car is nearly an hour away.

Thanks for the tip on the ECU connection. I was planning on cleaning up the ring terminals and verifying their integrity anyway, but now I'll check the other end of the harness. I remember reading about the fan harness corroding, too.

 

I'm on the fence about it.
On the one hand, this is kind of a surprise gift for my brother who thinks it just needs an aux battery and it'll be good to go.
On the other hand, their son doesn't get his temps for another 6 months and there's more work to be done on it once we know the powertrain is ok.

If I still had a NiMH charger I'd be more inclined to do the full refresh, but as it stands I'm hesitant to purchase a charger that is otherwise useless to me, to possibly refresh batteries in an abused 19-year-old car I don't even own with 180k miles.

I might change my mind after doing more research on the topic. Right now I'm having issues with most of the guides being written a decade ago and discuss RC chargers that are no longer available, plus half the posts in threads being slapfights or people advertising a $500 grid charger.

 

I decided to at least make an attempt at all this, but keep finding conflicting information over the years as people try different things with mixed results.

For discharging, the consensus I've seen is to aim for 6v. I set up an Arduino + relays + 50 watt headlamp bulbs so it could time 4 modules and automatically disconnect the bulbs when 6v is reached.

It's recharging the modules I'm a little more concerned about, and it makes me hesitant to discharge my 'good' modules. I don't have a use for a NiMH RC charger, but I do have a use for a benchtop regulated power supply. From what I understand, I'd want to set it to 325mA and ~8v times the number of modules in series (So for 6 modules, 48v). Does this sound right? Once the power supply reaches the target voltage, it switches from Constant Current to Constant Voltage and the amperage drops.

Do you know offhand, typically how long it should take to recharge a module from 6v to full at a constant 325mA? Not looking for an exact time, but is it like a 'couple hour' thing, overnight, weekend...? I've only tested on my 'bad' modules so far, had them charging for ~24 hours. Two dropped below 6v within 30 seconds, the third is still running after about an hour.

 

Yeah, I meant discharging. I knew all three were bad, when I pulled the pack they were at 0.3, 0.6, and 1.2v. All three charged to a fake 8.1v and two of them discharged immediately to 6v under load (unsurprisingly). The third module ran a load for an hour or so before I realized the Arduino locked up. With an initial 1.2v maybe it still had one good cell inside, but the module itself is obviously toast and only good for testing the Arduino programming.

It's supposed to keep track of how long the program was running, but I must have messed something up with keeping track of a large number. I'll have to figure something more graceful than "Increment timer variable by 1 each loop", but the Arduino doesn't have any type of clock. Maybe slowing the program to 1 cycle per second would help.

What I'm concerned about is if the bench power supply is even really recharging them or just doing that 'empty voltage' deal like on dead modules, so I end up discharging all my modules to 6v and they all show 8v after recharging but all immediately drop back to 6v once the car tries to start the ICE.

And just to verify, the ~35 hours is true whether I'm charging 1 module or 7, as long as the current is 325mA and the voltage is multiplied by the number of modules?

 

Woke up to find the 7 'good' modules (pulled at 7.6v each) on the charger reached the target 57v and current actually dropped to 0mA. I wonder if the bad modules were still technically charging because it would continue to show ~150mA at target voltage.
No matter, the bad modules are just for experimenting and testing the Arduino programming before risking my good modules (like yesterday when it locked up).

Specifically went with a 60v-capable supply so I could charge 7 modules simultaneously, conveniently 1/4 of a pack. Likewise, wired the Arduino to discharge-time 4 so it's 1/7 of a pack (plus I ran out of analog inputs).

Got an early Christmas present of an ultrasonic cleaner; a little vinegar and brass brushing yielded some clean busbars.

 

Hmm, any thoughts on the highest safe voltage I could charge each module to? Started discharging the good modules and during recharge, it only takes about 18 hours before the power supply reaches target voltage (57v for 7 modules, 8.14v each) and drops off to 0mA.

For the first discharge cycles, numbers are all over the place. Only tested half the modules so far but worst is 64 minutes, best is 153, average is 121. The 64-minute was on the same block as a faulty module. If it doesn't reach at least 110, I'm going to replace it.

 

Bumped it up to 60v, the maximum that power supply can handle (8.57v per module). After 40 hours it was still putting current in, though it tapered off from 325 to 242mA.

The modules felt cool to the touch, so I started removing the clamps to start charging another batch. As i loosened the clamps, it seemed like the modules were trying to bulge out. Tightened the clamps back up, guess I'll let it sit under the fan for a couple hours and try slowly unclamping again. Maybe the middle is warmer than I expected.

Had it clamped with 2x4s at each end, also resting on 2x4s to get airflow underneath, in a cool basement in Wisconsin, with an M18 fan blowing down the length of the modules. Honestly didn't feel warm to the touch at any point but that's not saying much. Wish I owned an IR camera

 

Checked again this morning, site won't let me edit the previous comment. Didn't swell this time, everything looks under control. Went ahead and 3d printed a copy of those orange busbar holders -- twice now, I've had a nut land between two other posts and spark while removing the busbars.

If it's beneficial for anyone else, here's the STL: Prius HV Busbar Holder by XirallicBolts - Thingiverse

 

Whelp, after many weeks of discharging and charging.... I got burned by eBay. Two of the five modules are in just as bad of condition as one of my bad modules. Consistently discharged to 6v faster than all the rest. Considering these modules are marked 2017, they must've been out of some high mileage taxis.

I don't know where I'm going to go with this. The future owner just got his learner's permit on Friday so it'll still be a few months.

I guess I'll have to get ahold of two MORE modules. Honestly wish I had left negative feedback instead of neutral. I at least mentioned that it's inappropriate for them to send such a wide range of modules -- a 2007, three 2017s, and a 2018 module. No end-user wants that kind of variance and they should be matching modules on a specific order.