Gen II Prius Individual Battery Module Replacement

Re: "Individual cell voltages cannot be determined" - FYI: they can - on cylinder type NiMh batteries used in 1997-1999 Prius and in Honda Hybrids. Not the point.
The point is: In Prismatic NP1 & NP2 modules used in Prius/Lexus/Camry hybrids - although you can not directly access individual cells in 6-cell module - what you can do and what I observe all the time is: THE CURVE. When the discharge curve down to 6V has a steep slope - almost falling down in a free fall - that means one of the cells in that module is hugely degraded. Those modules could still show something over 2 -3 AH capacity but you should not use them. They are "dead". Have a look at this sample - I have published it over a year ago and already mentioned that before - see the first chart where I showed two dead modules (#12 & #18) one is over 2.2AH and another one over 3.3AH @ discharge current 2.3A (those were replaced by two good ones and that battery is still working for more than a year now)

20 Channel High Voltage Battery Analyser - HYBRIDS

There is no point of wasting your time to charge to 7.5Ah. Quick temperature raise to 35-40 Celsius (ambient 17-20 Celsius) in second hand batteries (with ~ 3.5AH remaining capacity) during charge @2A from about 6V could happen at about 4AH pumped in.
I just set the temperature alerts alongside with AH cut-off and other alarms.

As for safety - I control the Complex remotely and not only observing software performance and alarms online but also have the video feed from the Complex to see if there is any smoke (kidding), to see if there are any curious cats (serious) got into the premises. Wish to avoid cats being testing how many lifes they have It is so easy to set up - I do not even need to explain how-to, right?

Thanks Eric. Tesla Battery has been comissioned last Friday. I can now see the stats and as promised will be sharing monthly on our facebook page here: Toyota Hybrid Battery Experts

 

Just wanted to thank everyone except for S Keith for all of the information and discussion in this thread. I learned a ton, and started reconditioning and test my (seemingly) perfectly good ~150,000 mile pack today. It'd been a quite while since I drove it last and it's SoC at the time was very low, and all of the cells had stabilized around 7.5V (two at 7.48, most at 7.5, a handful at 7.51) so I'm hopeful of simply refreshing their capacity and perhaps discovering if anything is abnormally weak and should be replaced while I have things apart...I might make one of those fancy spreadsheets and update this post if I find anything interesting along the way.

I wonder how michaelk87 made out...

 

Hi.

I've been balancing my 2005 Prius battery for the last month or so. Today I put it back in the car. Things were fine for about five minutes then it popped a code for module 12. I opened it up again and tested the internal resistance (20mohm), and resting voltage. Both seemed fine. The car seemed to be driving normally as well despite the code. The only thing I noticed was the amount of battery charge indicated on the screen seemed to be jumping around. It would be full, then almost empty, then full again, only a few seconds apart.

Here's the discharge/charge data for the module. i ended up cycling it 7 times, but it got up to comparable capacities to the other modules:
D1 C1. D2. C2. D3. C3. D4. C4. D5. C5. D6. C6. D7. C7
1644 3837 2757 4206 2956 4738 2296 5036 3605 5546 3525 5856 4620 6384

Should I replace this module? Should I replace other modules that needed a large amount of cycles? One I ran 8 cycles on, another 7 cycles, and 3 I ran 6 cycles.

Here's a link to all my charge discharge data:

 

You need to load test them to see if they are bad or not, just testing voltage and balancing will not produce good results unless you get lucky.

 

Here

Could you explain how to load test or point me to instructions elsewhere? I have the reAktor 300w, which can discharge at 20a. That should work as a load yes?

Edit: nm found it.

Just load tested all modules, One minute at 20a. All modules had voltage drops between .30 and .37. Module 12 was .35

Where do I go from here?

(Added load test data to linked spreadsheet above.)

 

Do you have Techstream? Run a charge and discharge test in the car while monitoring live battery data.

Charge all the modules to full. Let them sit for 24 hours. Discharge at 20 amps to 6 volts. That's a good measure of capacity.

Where are you starting the module count? Techstream starts the count of PAIRS opposite the battery ECU.

Brad

 

Does techstream work for Mac? That's the only computer I have. I started my module count on the ecu side of the battery pack I had the code read at an autoparts store. The code was p3022.

I already did individual module capacity tests. Would doing it with techstream provide me different information?

 

PO3022 Means block 12 is weak. The block codes are P03011 to P0324 for blocks or pairs of modules 1 to 14. Again the codes readers start the count of pairs opposite the battery ECU. So block 12 is the 5th and 6th module from the ECU.

No, Windows only.

At what discharge rate? A 1 amp rate is almost meaningless. Use the 20 amp rate the Reaktor can handle. Then you are measuring capacity at a rate the modules experience in the car.

Techstream will not measure capacity. It can show live block voltages as you drive. The weak blocks clearly stand out.

Brad

 

I did my charging and discharging at 20a. Here's my data:


How do modules 5 and 6 look to you in that dataset?

 

Ouch! slamming a 20 Amp rate until full is just mean. Stay away from full and it's fine. The car does much more but not to full. What did you use to terminate the charge? Delta V or delta temp? What where the settings?

Number 6 looks suspect. It's initial discharge was low and it took a big step backwards in column "L".

Brad

 

I used delta v. So should I just replace it?

 

I would run another cycle on it.

 

Alright. I'll give it another go.

 

Folks,

I’ve mentioned on many occasions that just capturing numbers like in the many posts here is good to know but not good enough.
Without seeing module performance during charge/discharge it is not easy to identify faulty modules.

Remember that 6 cells inside the module are not aging similarly - meaning that one of 6 cells could have degraded and made the module almost obsolete. While charging or discharging that module you would observe some capacity readings but will not be able to identify failed module.

Very steep charging or discharging curves under the constant current load is the way to properly identify those faulty ones.

If it were that simple I would not come up with my High Voltage Battery Analyser Complex which apart from providing essential Control and Alarm functionality also Visualise the testing process.

Just by observing the Charge Graphs I can tell without the need to conduct the Load Tests which modules are dead.

Last night uploaded new 5 min video about that. Watch the Charge Graphs on #18 & #19 at the end and compare with the rest.
That is the follow up video from the previously published Part 2 (Load Testing 20 modules simultaneously).
It also shows one of the ways to balance charge the pack, although our methods rarely require doing that.
Enjoy:

 

Hello everyone:
I am new to this thread although I have read a lot of it. I have another thread going on the main forum about the results of converting a 2005 Prius to a light Duty pickup truck. That effort went well, a link to the newspaper article is provided below, but that is not what I am writing about here.

What do you call a Prius hybrid combined with a pickup? The Pruck (w/video) | Tampa Bay Times

Last week, my daughter's 2005 Prius with 177.5K miles started showing Traction Battery Block failure DTCs. So I bought a used battery that had been previously balanced by Todd Thompson at TampaHybrids.com and installed it in her car. Took the core home with me, as well as three good modules I bought from Todd. I am now in the process of balancing that battery using a Hitec x 4 charger. The procedure I am using are those published by ryousideways on page 2 of this thread. I have read other shreds of information in this very long thread, some of it conttradictory. Such as his procedure of equalizing all the modules by wiring them in parallel. Later on there is information that this is useless . I have three questions that I can think of right now:

1 Somewhere in this thread, is there a condensed version of what the most probably correct cycling parameters and technique is?
When putting the pack together again, there is no rule that says they have to go back in their original position. Someone suggested that the modules at the end run cooler, so when re-assembling the battery, put the original end modules in the middle and the original middle modules at the ends. This leads to question #2
2. Since the battery and hybrid ECUs evaluate Traction Battery health on a "block" basis rather than "per module",(2 modules in a block), is there any merit to pairing a slightly stronger module with a slightly weaker module to come up with a roughly equal distribution of capacity across all 14 monitored blocks?
3. Ryousideways' criteria for a well balanced battery is that module voltages should be within .02 volts, and module capacity within .2 Ahr (200 MAhr). Looking at my discharge/charge data so far the latter does not seem to be a realistic/achievable goal.

Looking forward to hear from the experts. I certainly am a novice on battery issues.

 

Nice job on the ute ruudh it looks great, very tidy work. best and most recent battery/rebuild tips are in this thread from about page 50 onwards.
good luck!

 

Hi. Good question. I too have been reading this increasingly complex post of a mere 75 pages and taking notes. However, I used just 1 HiTec x 4 charger and charged/balanced 28 modules. Figure it out: an average of 19 hours per four module / 28 modules = 7 separate charge/discharge routines = 133 hours, assuming you are up all day and all night for just over 1 week. Since we need some sleep, assume that the 7 cycles take 24 hours each, then it would take exactly 7 days for 3 cycles, not counting some modules that might require a 4th, or even 5th cycle. Hope this helps.

Thank you for the comment and the advice on how to read this thread. I am slogging through all the 75 pages of posts to filter out relevant technical information. I have been looking for consensus but I have found some, not much of that. A little bit time consuming, but then it is worth knowing/evaluating information from contributors who really know this stuff. Certainly I don't.....Ruud

 

I test the pack in less than 40 minutes to identify failed moduiles. Can be done during charge or discharge mode. Easy to spot.
Re; some of your questions:
- Multiple CHRG/DSCH cycles is just waste of your time;
- Parallel balance - do not do this. Dangerous and prohibited by the nature of NiMH battery design;
- Different capacity modules mixed in pairs - that could make sense but only if you can test the Delta V all way through within the working range of the battery which you can't with you RC Toy charger;
Based on multiple assumptions and misconceptions people have made on this forum I've compiled few videos (3 to 7 min long) to help you visualise what's going on when charging/ discharging the pack.
Just watching 3 of them will give you idea of how modules behave under load and during charge. My site is the other source - I've published there (as well as on FB page) numerous test results (chanrge/discharge) conducted on the individual modules simultaneously under the same load/charge current.
Remember that if you've tested for example 2.5Ah - 2.7Ah @ 1A down to 6V with your toy charger - that does not give you assurance that the tested module is usable. I can distinguish between usable and not usable modules of the same tested capacity just by looking at the graphs.

 

no worries I hope you did not take me the wrong way there, I just mean from about page 50 onwards the information is more fact than guess/trial.

cheers

 

Cmon man, why not just say don't DIY this at home unless you have one of my $10k test setups.


If anyone is interested the way I rebuilt mine all the way back on page 11-13 in this thread has worked splendidly to date and still going strong. Looking to add a grid charger setup from Hybrid Automotive to it soon as well.