Help! I bricked my wife’s Prius! (HV Battery Swap Gone Bad)

Warning - long thread. Please wait for multiple posts before replying.

Background - I’m a retired computer systems engineer. In 2008, we bought my mom a new Prius, and when Mom passed, my wife inherited it. So, a “one” owner, low mile (70K) prius driven gently by two classy mature ladies, 100% dealer maintained, never an electrical or mechanical problem. (We’ve towed it on a dolly behind our RV since 2014, so the back wheels are higher mileage). A few months ago, we noticed declining mpg and occasional, worsening park-green-start-purple events and startup (ICE only) hesitations. I called our respected dealer, and they confidently predicted a relatively near future death triangle event (and no offer of any goodwill or discount on a OEM new battery). I decided on a DIY pre-emptive HV battery replacement before we head for the hinterlands, and to spend a few extra $ on a “Gen 2 Battery Rebuilt From Gen 3 Cells”. I then started driving, testing, and “excercising” the Old battery while waiting for the replacement battery. This is when the confusing and frustrating stuff started....

 

Now, this sounds a bit arrogant, but it’s literally true that I have design, engineering, diy maintenance, rebuild, and/or repair experience on bicycles, cars, trucks, boats, computers, small and large airplanes, and I’m safety trained and certified on electric railways on 3 continents. So a Prius HV battery swap should be a walk in the park (with a little help from my friends at PriusChat), as I glibly explained to my wife and other skeptics.

The first bit of confusion - the old battery system started behaving noticeably better exactly while I was on the phone placing my order with the battery vendor. Why does Prius HV battery life/performance depend a bit on driving style and test history? First, everybody knows that the closer the mechanic, the more intermittent the problems, right?.

Next, I downloaded Dr. Prius (iPhone, beautiful application) and plugged in my ELM327 for some in-depth testing. The OBD connection seemed a bit flaky, but a quick email exchange produced “Apple recommends a network reset for reliable WiFi elm operation”. A bit confusing because I use it reliably all the time in non-Prii. It did then produce lots of test results, but no bad module suspects, or other signs of imbalance.

So, how can I estimate the useful life left in my old battery and the expected gain from my “new” rebuilt battery?

I ran 3 “life expectancy” tests, improving my technique and the indicated capacity each time (43%, 48%, 51%, the last with no warnings).

Another quick exchange with Jack at DrPrius produced this - “Exercise NiMH can get it to rearrange the crystal structure and increase capacity temporarily, however don’t forget the actual capacity decline with each full-cycle, in other words, don’t exercise too much”. BTW, Jack is a true hero. PriusChat should strike a medal for him and the Dr Prius iPhone app.

Notwithstanding the testing, I was still seeing “Park green, start purple” battery events, and conveniently, the nicely crated battery arrived just when I finished stripping the interior. Then the confusion ramped up again...

 

While testing, I read that there’s really little or no difference between “gen 2” and “gen 3” battery modules.

What’s the complete definition of “gen2” and “gen3” as applied to Prius battery modules?

Apparently Toyota didn’t publish details of any change or upgrade to them associated with the gen 3 model year. (But I do notice that there are multiple battery parts fitment changes during the gen3 model years?) So, (I guessed) the only practical way to define a “gen3” module is to look at the serial number on each one, which has the manufacture date encoded in it. Note that I originally decided on the “whole battery pack” approach to avoid getting into battery disassembly and per-module balancing work, to get my wife back on the road ASAP. Now I had to do the “QA” function to be sure the rebuild actually contained “gen3” cells. It turns out that each has a QR code on it, so I could scan all and copy to a spreadsheet (all on my trusty iPhone), translate the dates, and compute the age in years.

The average was 2012, in the middle of the gen3 years, and 27 out of 28 were exactly as expected. Unfortunately, one was clearly anomalous. The date code was actually Sept, 2006, 2 years older than our old battery pack!

How did a battery made in 2006 stay lively enough to stand tall with the gen3 crowd in 2019?

And a series-connected battery is only as good as its weakest cell!

So, does age really have any effect on battery life expectancy? I emailed the battery vendor and called my Toyota dealer tech and management to learn the following:

“Date doesn’t matter, only a full charge/discharge capacity test and IR measurement matters. That was the best battery we had. Go ahead and install the battery, it’ll last a long time”.

“Toyota has never published details of battery life and test specifications, but if you had the death triangle, brought it in, reminded us that you’re a 3-new-Toyota customer, we’d comp you a new battery.” (damn, I wish I’d asked that way on my first call!)

“Our current core value is $1300 and all cores go only to Toyota certified hybrid (recycling?) shops.”

“If you must know, capacity (AH)>=6.5=“gen3“, and <=5.0 is junk. No per-module or whole battery test data is kept. The paper whole-battery charge-discharge curve is all you get. The one lower-voltage module line visible there is because one voltage sense/compute includes the safety disconnect voltage drop. Please install the battery. It will give you long life”.
...

 

By this time , my lovely wife seems to be tiring of the noisy, 2-seat weak-battery configuration, so time to shoot the engineer and get on with it.

With a little help from my grandson, a plywood ramp, and some paracord, we got the swap done with no sparks, scratches, or dents. Reconnect +12 and safety connector for first test, and no joy! Death triangle and P080D means we forgot to push down on the safety lock. Click, done. Then the real troubles started...

After a couple tries, the best achievable status was grim:

What is the root cause of death triangle, red !, flashing gas gauge (first one bar, then all bars), VSC, no engine start, and codes cc100 and c3000?


(Did I mention that the hood had never been opened for any battery work to date?).

Note that the above codes are not listed in any thread here or any other google I tried.

Almost all functions except the engine worked normally, including forward and reverse on battery. The engine even started up for a few seconds once, then shut off when we closed the rear hatch!?!?!? And then no more engine.

Checked the ECU connections twice, reset codes via 12v, OBD, and hold start for 45 seconds. No change. Contacted battery vendor, tried a few more things, then put old battery back in and repeated all the above results identically. So, swapped batteries for the THIRD TIME!

I’m a complete believer that when things go wrong like this, the beginning of complete disaster is when I expand the work list to as yet untouched parts, but I found a 2008 mk2 fuse diagram, and went thru it to identify all fuses that might affect the engine. When I opened passenger compartment fuse panels, the layouts were not as shown.

How many different fuse panels did Toyota install on gen2 priuses?

So, we confirmed that at least one function on each suspect fuse was functional.

I read one post implying that these codes might implicate the CAN bus, so rechecked the connections on the battery ECU visually and electrically multiple times. OBD bus resistance is correct for 2 parallel terminating resistors.

At this point, both batteries showed only 1 purple bar, so we started preparing to use the last bit of energy to drive onto the tow dolly and tow it and a bag of cash to a dealer. Meanwhile, we’re getting hungry because my wife has no favorite car to go buy groceries


So, if you’ve had patience this far,

What’s the magic bullet to avoid letting a dealer or other hybrid mechanic throw some expensive parts at this brick?

 

...

Since it took a while to write all this, spoiler alert - we got it running for under a buck!

I went back thru the fuse diagrams and located a few evasive fuses to check further. The first one I put a multimeter on had a 14 volt drop! Yes, it was the AM2/IGN 15A fuse under the hood, and yes, it came up ready. A short 12V disconnect, and NO CODES!

So,

1. How did swapping a battery take out the AM2 fuse, or what caused it to blow?
2. Why can’t the ECU gurus do better than a couple undefined codes? If I’d have been on the interstate when it blew, it would have been costly and unsafe.
3. Why doesn’t Toyota provide owners with better tech data?

We put ducts n vents back in, drove it around for a while, and it was easy to see that the new battery is stronger.

Cleaned and Reassembled a few dozen interior parts, then did the final test - Dr. Prius life expectancy test ( remember old battery at 40-50%, near end of life). The test took about 30% longer to complete, and I almost froze from keeping the discharge rate in the green with the A/C, and life rating 125%compared to a new battery. So, yes, Virginia, there really is a “gen3” battery improvement available! Thank you Mark at Falcon Hybrids.
BTW, the posters here and some other threads (some pretty senior members, I notice) are way more hostile to both “gen3” and Falcon than my very diligent experience would support. I thought a moderated forum would produce better information. If I’d read that hostile stuff, I’d probably have chosen (unwisely, in retrospect) to equip and fuss with module level repair and rebalancing.
We should reserve our hostility for the Ovshinsky/GM/Texaco/Chevron (and yes, in this restricted context Japan/Panasonic/Toyota, and USPTO) conspiracy that deprived us of this valuable technology for decades and left us with a proprietary, secretive, restricted parts market.

With only a little luck, I’ll probably get much more than the warrantied one year life.

But, now I have to wrestle with that AM2 fuse mystery, and hopefully respond to some of the above helpful posts...

 

1. I had 3 price steps - “gen 2”, “gen 3”, and dealer new. Each about $400 apart, with due attention to core and shipping costs.
2. There has been some risk expressed that “new” may have some issues (old module dates, balance, and, of course actual capacity.
3. Age is much less a factor than tested capacity, and average/peak temps during storage or prior use are also a factor.

So I chose the middle step. In retrospect, new would not have helped me at all, but only the next hopefully 4-5 years will close the case.

 

This complexity is one reason I wanted to avoid module-level fun and games.

 

Cause I tend to be at large in America all summer, a long ways away from my garage, and potentially a long ways from either HFT or Toyota level civilization.

 

Bought one. It seemed so unfriendly to my iPhone that I decided further oldfart diagnosis (voltmeter on fuses) would be more fun. In retrospect, not sure it would have done better, but I would like to read a system analysis on this...

 

Sorry if I wrote imprecisely. When I started the “project”, the only solid problem was park-green-start-purple, and occasional hesitant acceleration because low hv battery soc. I did have to charge the 12v for a few hours only once, and thorough testing later indicated battery ok. All the ICE crap came up (apparently purely by coincidence) during the first battery swap. I had a long talk with a senior toyota tech, and he couldn’t offer any better strategy on the am2 fuse than “drive it til it blows again”. I’ll probably add occasional “touch the fuse after a drive” tests, and if I burn my finger, I’ll dig into the inverter pump. It does seem to be running better ( fuel economy) day by day, presumably as it relearns detail parameters after all the resets.

 

Here’s the fuse document...

Note number 31...
Here’s the dash...