Gen II Prius Individual Battery Module Replacement

Prius Battery rebalance thread | PriusChat

HV battery issues | PriusChat

Traction battery rebuild question | Page 6 | PriusChat

Hi twocell,
You're in for quite an adventure. Good luck.
I don't feel qualified yet to answer your technical questions but it may be helpful to pass on a few other things.
I started by reading this thread twice. Took days. But it at least gave me a little background vocabulary that was very helpful and gave me a glimmer of hope that this really was possible for us NOOB's to do.
Then I found these threads that proved to be helpful as well:
Prius Battery rebalance thread | PriusChat
HV battery issues | PriusChat
Traction battery rebuild question | Page 6 | PriusChat
In this last thread, YY4U tells the story of what he went through. Very heart warming saga of the trials and tribulations that you may face. I know I ran across many of the problems he had. But he was ultimately successful through his unrelenting determination and the help and encouragement of everyone who took the time to give him advice.
I am blown away by the depth of knowledge expressed by those on this forum. Very humbling.
The one piece of advice I do feel qualified to give is to go very, very slowly. If you're not sure if touching something is safe, assume it's not, and put on your HV safety gloves before going forward. And did I say to go slow?
All the best,
John

 

Thanks. you gave me one thread I was not aware of. I hope it has enough for me to build a procedure.

I will take my time and share anything that may be additional information. Thanks again.

 

Decided to play it safe and Dchg the modules manually. Used two headlights to Dchg two modules at a time to cut down on the overall timeframe. Tried to bring each one down to the 7.0V to 7.2V hoping that they would rebound to somewhere below 7.5V. Was largely successful except for a few instances. I found that Dchg to 7.2V was not as successful as Dchg to 7.0V. I've posted the Dchg results in the attached spreadsheet.
It looks to me as if I need to further discharge #4, #6, #8 to bring their resting voltage down closer to the 7.5 target I set. I'm afraid I may have ruined #14 and #19, especially #19. I thought I had the timing down but these seemed to accelerate the Dchg during the last few minutes of the Dchg process. I wasn't watching closely enough and was relying more on the timer to alert me to end the process. Seemed to have reached the far end of the Dchg curve (post #658) more quickly than anticipated. Watched them a lot more closely after that.
So #4,#6,#8 are too high but I can easily bring them down some.
Can I just run #14 and #19 through another Dchg>Chg cycle using Dchg to 6.0V and the Chg using a capacity cut-off of say 6500mAh, see where the voltage is and then manually Dchg using the headlight apparatus to the 7.0V that then seems to be returning a rested voltage in the 7.4V - 7.5V area?
Or...
Can I just equalize all the modules by placing them all in parallel? At what point, i.e., what's the maximum difference between the high and low voltages does it become unsafe to parallel equalize? Module #4 is high at 7.625V and module #19 is low at 7.113V. The difference is .512V, a little over half a volt. I'm going to go back through some of the threads on equalizing and see if I can get a better handle on this point. I would sure like to hear anyones opinion on this, however.
Thanks
edit: I've been using the term balance in this post instead of equalize. Will be more careful in the future

 

Forgot to attach the spreadsheet......

 

I started by building proper tools to test the results. Took months.
Few years after first HV Battery Analyser was up and running I joined this forum to see what issues DIY folks are faced with.

There is too much fudge and a lot of nonsense around battery rebuild. Too many unfulfilled promises, misconceptions, wrong expectations and misinterpretation of seems to be straightforward principles of physics and logic.
My tools and methodology are honestly revealing the reality, bluntly exposing the facts of the matter.
If you wish to start honest battery re-build business - come to New Zealand to get full training and bring back home complete battery testing complex, see Testing Complex - HYBRIDS

P.S. up till now - the one little thing I added to my tool set based on this thread is a 6 amp load for testing 1 module alone. Normally I test the whole pack in one go with the high current loads.

 

Hi Alex, I am about to start this process using the b6ac. Do yoiu mind sharing the settings you are using. Thnx

 

Hi Raysa

First of all if you want overcharge the modules because it's impossible to turn off delta peak sense which always stop the charge before overcharge, you should set it to maximum which is 15mV/cell, to change that setting you should go in the memory section where you can save/load your charging settings.

I charged mine at 0.7A with 9100Ah cut-off, i used low charge rate since time does not matter for me, but you can use 2-3A charge rate for faster charge.

I used 9100mAh cut-off like John(Britprius) recommended in this thread somewhere on last pages you can find
Discharging down to 6V.
Actually that's a same settings that everyone uses here, nothing special in B6AC charger.


Alex.

 

I'll try to bring everything up to date.
6/11 - Brought the voltages of #4, #6, #8 down some and placed the modules in parallel and left them to equalize overnight.
6/12 - Did another load test. John(Britprius) was right. It's a little easier to capture the voltage readings with the module voltages in more of the flatter part of the discharge curve than it was before. Here is where I'm still a little fuzzy when it comes to looking at the results. Instead of looking at the calculated internal resistance for anomalies, (I'm still not convinced of my ability to capture the data correctly) I looked at the difference in the voltage at the beginning of the load test and the voltage at the end of the test while the module was still under load. Based on those numbers, #19 again stood out at .951V difference and #14 at .711V difference. All the rest were in the .4 - .5 'ish range. I've uploaded the spreadsheet with the test results. It's the tab Load Test June 12.
My question now is do I replace #19 and #14 or just go with them, put these Gen3 modules in my 2005 Gen2, and see what happens?
Again, any advise or thoughts on this would be very welcome.

 

Took some time to go cool off in the mountains. High was around 115F in Phoenix and it was nice to go someplace cooler.
6/20 - I pulled the HV battery out of the 2005 after its 24 hour session with the Hybrid Automotive Prius Grid Charger. After looking at my Techstream data I didn't see any (obviously) bad modules so I thought I'd try this piece of equipment out and see if I could correlate with the Dchg>Chg cycles I plan for these Gen2 modules after the swap with the Gen3 ones.
Probably have more to say about that later.
My concern now is how to pair the Gen3 modules into blocks as I'm installing them into the 2005. I agonized over replacing modules #19 and #14. I finally decided that maybe they weren't that bad and looked at just pairing them with stronger modules.
How to decide which were the stronger ones?
On 6/11 they were all equalized. Showed about 7.494V. I had recorded the voltages again on 6/12, about 12 hours after equalizing. This is V(6/12) in the "Gen3 Voltage Pairing" spreadsheet I've attached.
On 6/24 I measured the voltage again, V(6/24) in the spreadsheet, to give me an indication of which module was dropping more quickly compared to the others. 12 days between readings would hopefully be enough time to show any real differences. Modules #14 and #19 showed the lowest voltage readings, (highest drop in voltage over the time frame).
So I sorted all the voltage readings so I could pair the modules with the highest voltage drop with those with the lowest.
My thinking is that code p0a80 is signaled (I think) when the HV battery ECU detects a 1.2V module pair (block) difference. So perhaps this pairing will prolong that event. My highest combined voltage reading of 14.977V for module pair #10 + #9 is 0.161V higher than my lowest combined voltage reading of 14.816V for module pair #12 + #19. I don't have a clue how the HV battery ECU finds its 1.2V threshold, but 0.16V seemed to me far enough away from 1.2V That everything should be OK.
I really am shooting in the dark, here. So please someone let me know what your thoughts are about this way of pairing modules and the reasoning(?) behind it.
Tomorrow I'll be placing the Gen3 modules into the Gen2 battery pack as outlined above. My plan is to then have those Gen3 modules spend some amount of time on the Grid Charger and then to fire the 2005 up for a test run. Wish me luck.....

 

The voltage difference is only 0.3V and that is measured under load/discharge. The difference of the max voltage and min voltage blocks are compared. I believe the voltage differential has to be sustained for a continuous period of time (secs), but off the top of my head I can't recall the exact number for that.

Good luck.

 

dolj,
Thank you for the clarification. I picked up that 1.2V differential from another post that I read.
2005 Prius P0a80 | PriusChat It is post #10 by jdenenberg. I really respect his expertise. I'm sure I just misread or didn't understand something.
I've attached two items to this post. One is a copy of the post by JeffD and the other is a Techstream screenshot of my 2005 Gen2 taken on 6/21/15 at 10:00 AM. It shows a difference between the Battery Block Minimum and Max Vol to be 18.42-18.13=0.29V . Does this mean my Gen2 was close to throwing a code? I don't have a clue how long this situation persisted. All of the modules in this Gen2 HV pack are being replaced by the newer Gen3 ones so there's no real cause for alarm. I'm just trying to further my understanding a little. Thank you for the time you spend helping me and others.
John

 

John,

The 1.2 volt differential happens when one of the 6 cells in a module is shorted (usually due to being discharged beyond 0 volts). This is a permanent failure of that module. The 0.3 volt module pair (Battery Block) differences can often be corrected by module balancing (getting all 6 cells to the same SOC) and equalization of all modules to the same voltage.

JeffD

 

Jeff,
Thank you. I knew it was just something I didn't quite get. I appreciate your help.

 

Possibly, but not necessarily as the 0.3V difference has to persist for a period of time. Unfortunately, I can't recall what the period of time is. In any case, a fleeting reading will not throw a code.

 

Thanks, dolj.

 

Correct. Cheers, mate. I guess with the snow in Welly this morning, you have parked your Prius for a while

I've done live tests on Prius - on new and old packs - the max difference on pairs (during MTG start) was 0.5V and individual module voltage dropped below 7V (you can't capture that with Techstream - TS is it quite slow to do that). And no error codes were presented.
Duration was quite short - just the time for the ICE to kick in.
Those live tests I have conducted render a lot of mind-blowing claims and conclusions made in this thread pretty much useless.

In particular those useless "capacity tests" with RC cycle chargers: e.g. measuring capacity with RC toy chargers are useless as they do not capture individual cell failures (one of 6 cells in the module).
It is self-explanatory from the graphs I have posted on my site.
Nevertheless I found from the questions I am receiving regularly - it is "too hard busket" for some to comprehend.

Last weekend I've tested another pack - straight away 7 modules defected. You would never pick that with RC cycle charger, voltmeter or Techstream. ..Posted here - first screenshot from the top:
20/40 Channel High Voltage Battery Analyser - HYBRIDS

As I've mentioned earlier - I invite peopel who wish to start testing and rebuilding pack professionally to come to the "Emerald City" (Auckland, New Zealand) and get hands-on experience from the Wizard. But even if you don't - as Dorothy suggests in the clip below - you "would not be any worse than now..."

 

Kia ora, kiwi, kei te pehia koe? It is cold atm, but not cold enough for snow here in the city, despite what you might have seen on TV. The Pri is always parked in the garage out of the weather, but ready and able to get me where I need to go. LOL.

 

When you just change "module or two" and test battery pack on car – it may look successful for a while, and the car may even be drivable for a week or even few months. But that quick fix will not last for long and the process need to be repeated. If you DIY and have time to do it non-stop in between the trips – you can do it.

Here are few reasons why High Voltage Battery Packs (failed on you or obtained from wrecks or amateur rebuilders) need to be professionally tested. I am talking here about the Load Test to identify bad modules and to capture Usable Remaining Capacity of still good ones and another know-how metric crucial to the battery performance. That kind of test can't be done in principle with Voltmeter, RC Cycle Charger or even professional Techstream tool. Nor those devices allow rebuilding the pack professionally.

Recent load tests on 5 Toyota packs (40 modules each) identified a lot of failed modules which can’t be reused. And it was always more than just one or two bad modules. In two packs there were 7 bad modules in each of them. Perhaps some would still consider those “good” but my standards disallow reusing them.

See the sample screenshots of the tested pack where 7 bad modules have been identified and those can’t be (should not be) re-used. BatteryAnalyser's uploaded images - Imgur

Note that Voltmeter will show those modules as normal and RC Cycle chargers will show some capacity left in those modules and will not capture modules with one failed cell inside.

 

I have a 2005 Prius with 83k miles. 2 weeks ago it started throwing P0A80 frequently, along with an occasional P3019 or P3020.

I have an updated scangauge II (FW v4.12) permanently installed in my car. I cleared the DTC codes several times but they kept coming back.

I configured Xgauges for battery current, state of charge, and battery fan speed. Although I could still drive my car by frequently clearing the recurring P0A80, performance and gas mileage (as expected) was very poor (30 mpg). Whenever the car thew a new POA80DTC code the battery fan would immediately speed up (6) and the battery amps + or - would be nearly 0. The ICE ran all the time. Clearing the P0A80 would allow modest battery amps (perhaps +/- 15 amps), but the SOC would go all over the place. A new P0A80 would usually happen within 10 miles.

After reading this excellent thread I decided to take out my HV battery and test the individual modules. I immediately found 1 module with a hard failure ( 6.5 volts). Counting from passenger side it is module #18 in block 9.

My next step was to order 4 tested used gen II modules on eBay. The seller marked them for mAh capacity. They ranged from 5100 to 6200 mAh.

After swapping out #18 I load tested the entire HV battery on my bench. The only other suspicious module was #7 in block 4. It kept reading about 10-15 mV lower than all the others. So I swapped it out too.

At this point the HV battery is installed back into my Prius and it is driving like a whole new car! No DTC codes.

Battery fan speed stays on 0, 1, or 2. Regen braking current up to 100 A. Starting up from a dead stop the Prius will use 35-45 A to assist the ICE. While driving around town the average battery current stays within +/- 25 amps most of the time. SOC stays between 59 and 65% on my Xgauge.

On the Prius MFD display, all the blue bars are lit up and occasionally the first green bar. In comparison the Prius MFD bar graph was all over the place with my failed HV battery. One minute there would be only 1 or 2 blue bars, 5 minutes later it would show fully charged (2 green). I didn't often watch the bar graph in the past. But I suspect my HV battery had been getting weaker and weaker for the past 6 months. My mpg had been gradually dropping for the last 6 months even though it is summer now in the US midwest where I live.

So I think I've done some good by changing out 2 modules.

After I test drive my Prius for a couple of weeks I'll start a new thread here to discuss my analysis and testing methods. They are a bit different than what others are using. I should mention that I'm a working EE with considerable experience with both SMALL NiMH batteries (like in a cordless telephone handset) and with medium size NiMH batteries (handheld power tools). This is the first time I've worked on anything as big as the Prius HV battery.

More to come, EB

 

Wow! Thank you for this thread, everybody.
I have the P0A80 and P3017 codes. I think I will avoid the DIY and go to the dealer with this one. I am guessing approx. $3200.00 ?

It is interesting that the 12v battery is still the original and still ok.