Gen II Prius Individual Battery Module Replacement

117 pages.....i hope someone has a map for the thread. If not i got my bed time reading for the week.

 

For discharge/recharge cycling modules the default has always been to charge at at leisurely and safe 2A.

However, there have been posts where charging has been done at 6.5A.

I’ve just set up three chargers to cycle seven modules at 5A:
- iCharger 106b+ (one module)
- Turnigy Reaktor QuadKore (four modules)
- iCharger 406 DUO (two modules)

Discharge capacities:
iCharger 106b+: 20W
Reaktor QuadKore: 20W
iCharger 406 DUO: 80W on each of the two channels

Note: The iCharger 4010 DUO (which I haven’t got, but wish I did) has a discharge capacity of 130W per channel (twin channel), giving close to 20A discharge on a normal power supply (ie without regenerative discharge).

I’ve set all seven modules to charge at 5A, with the maximum discharge possible on each charger (ie 20W (2.7-3A) on the iCharger 106B+ and Reaktor, 80W (around 10A) on the iCharger 406).

I’ve set up a pedestal fan to blow air over the top of the modules.

I’m charging alternate modules (ie 1, 3, 5, 7, 9, 11 and 13).

Is 5A going to cause any problems? Or, should I set it back to 3A or 2A?

Overview - the pedestal fan is blowing air over the top of the modules.







The 80W per channel discharge capacity iCharger 406 DUO means that it gets through modules much quicker than the Reaktor QuadKore, but only two at a time instead of four for the QuadKore:



The Turnigy Reaktor QuadKore shortly after starting the first discharge:



The iCharger 106b+ at the start of the first discharge:



I bought this 2004 Prius from the original owner last year, with the intent of rebuilding the battery once I learned how. Techstream was showing faults in blocks 8, 9 and 10. There was a lot of corrosion on the copper busbars and voltage harness:

 

This clip covers removing the hybrid battery from the car, replacing the pack of modules and then refitting the hybrid battery to the car:

This clip covers analysing modules, and the process for deciding which modules to replace:

- For replacing modules, remove the whole pack from the metal casing first. It’s much easier this way.

- You’ll most likely find a heap of corrosion on the copper busbars and nuts. Although the copper busbars can be easily cleaned, consider buying some nickel-plated busbars and stainless steel nuts (ie M5 flange). You can find nickel-plated busbars on eBay US, or on Alibaba.com (eg Zhejiang Rhi Electric Co., Ltd. - copper busbar, flexible busbar).

- Figure out who does this regularly (eg for a living) and pay extra attention to their posts (eg strawbrad, ericbecky, S Keith) and anyone else with their own workshop.

- The Turnigy Reaktor series of chargers was previously the general favourite, but there was a software change in them a couple of years ago and the current ones won’t show the module capacity. The Charsoon Antimatter chargers are apparently identical to the old (ie good) Reaktors and can be found on http://www.banggood.com or AliExpress.com - Online Shopping for Popular Electronics, Fashion, Home & Garden, Toys & Sports, Automobiles and More. An iCharger DUO (ie 306, 406 or 4010) is also a good choice.

- Clean the blower fan! Heat kills batteries.

- For replacing modules, you want ones the same age and capacity as the existing ones. Each module has a serial number - the manufacture date is in the first four digits.

 

No way are you going to get one matched to your car. You will get luck of the draw.
Everyone uses different specs and tools to measure their modules.

 

Hi All,
I am still trying to get my arms around the basic information needed for successfully conditioning individual modules using a hobby charger.

What I am discovering from staying up late and trying to read years of past posts is that there seems to be no real consensus on some fairly basic specifications and procedures. The problem with looking back through the archives is that some of the recommendations may have been proven wrong by now, but it is not always obvious when hopscotching through past posts.

So my question is: Is there what amounts to a consensus on both charge and discharge values and a proven methodology for conditioning modules? This seems like a topic that would deserve a "sticky post" from a group of folks that really know what is going on from lots of first hand experience. The area I find most confusing is the range of recommendations for minimum discharge voltage level for a module and the maximum mAh charge threshold.

There is a considerable range of voltages that are recommended as the "ideal" discharge voltage. I understand the need to not go too low and damage individual cells in a module, although this number is also debated. My limited experience in discharging modules with a headlamp that draws approximately 3A is that there is a point around 6.5 volts where the modules "avalanche" for the lack of a better term. I have been drawing down modules with this ~3A load to about the 6.5V level and then using the automatic discharging function of the charger to slowly discharge at ~5W down to 6.2V. This allows me to get an idea of the capacity of the module and speeds up the first discharge cycle. My approach has been to do two discharge/charge cycles with 90% of the first discharge cycle done as the load test and then let the charger run in automatic to complete the first two cycles. This allow the module to "rest" before doing the last discharge/charge cycle which I have been doing fully automated. For now I am going through the entire HV battery to get all of the modules to the two cycles stage. My hope is that there is some better settings or method that I can use for the last cycle. There are recommendations for discharge voltages as high as 6.5V and as low as 5.5V or less. Is there a consensus on this? Is the minimum voltage variable as a function of the age and condition of a module or is there an "ideal" number based on the chemistry of the NiMH cells?

Several people have posted charging data charts that suggested their modules were charging up to the charger's set limit but may have been shy of reaching the module's total capacity. The recommendation was to increase the upper charge limit to between 8000 and 8500 mAh. The logic was that the Delta V feature in the charging circuit would protect from overcharging the module and having a higher charge threshold would account for the current (power) that was being lost to heat in the charging process. From an electronic circuit perspective this makes sense to me. I don't see how the charger can tell the difference between "useful charge current" that is charging cells from metered current that is converted to heat from the battery module's internal resistance.

Independent of the threshold question there is an issue of charging current and overall heat build up. Lower charging currents and longer charge times would seem less likely to develop damaging levels of heat in the module.There is mention of using fans and avoiding overheating but what is a critical temperature? How come more people are not using the temperature controlling features built into many of the hobby chargers?

In later posts I came across warnings about using charging thresholds above 7500 mAh. Related to this is the charging cycle power setting. Some folks have cautioned to use only 1A of charging current and elsewhere there is mention of the need to charge at a higher rate (up to 5A) to be sure that the Delta V function of the charger is triggered. The logic being that at low current rates the charger may not reach full charge or somehow be fooled. I have been using 2A for charging and am limited by my charger to a maximum of 5W of discharge power, ~.6-.9A depending on the voltage levels throughout the cycle. I am using an upper current limit of 8200mAh and seeing ranges for fully charged modules on either side of 7200mAh.

It appears that there is a point of diminishing returns on the number of charge/discharge cycles and there does seem to be a consensus that 3 cycles is a usually a good number. I am guessing that there is also an optimal process for doing the first, second and third charge/discharge cycles for a typical module with respect to reforming the cells. Is this true and if so, what is the consensus?

Thanks in advance for any feedback or suggestions. I am always impressed by the level of help and comradery that members of this forum have for each other.
Best regards,

Will

 

I feel your pain. My saga started a couple weeks ago.

2006 Prius here with about 125k on it. I'm the original owner. We hadn't been driving it very much over the past year and didn't drive it at all for around two months. The 12v battery died so I charged it up and turned the car on for a few minutes to make sure it was good (didn't get any errors). We then didn't drive it again for about another month and the 12v battery got low again. I charged it up and this time I took it for a drive and got the red triangle of death. I knew this day would come, but I'm honestly surprised this is the first error I've gotten in 13 years. I'm certain letting the car sit for those long periods caused my traction battery issues to accelerate.

Fast forward a week and I had my copy of Torque Pro, BAFX ODBII reader, and about 20 hours of YouTube/web research under my belt. I was getting a P3012 error, which was a bad module in block 2. I pulled the battery and quickly determined that one of the cells in module 3 was no bueno (stuck around 6v). Did a quick open voltage test on the others and they seemed fine (7.6v+).

I foolishly ordered just one module on eBay and had it installed a few days later. Took the car for a drive and it lasted about 30 minutes before I got a P3013 block 3 error. Damn. Picked up an "HTRC" dual NiMH charger on Amazon ($75) and started going to work. Quickly determined that module 5 (in block 3) voltage would drop big time as soon as any load was put on it. Ordered a few more modules on eBay and then started testing the others.

After trying different techniques and researching endlessly for days, here is what I've learned and what works for me (in no particular order of importance):

- I've read that these modules can be safely charged up to 3.25A, but it seems plenty of people charge at 5A and I'm doing this to save time. The NiMH charger auto adjusts the voltage.

-The modules are still compressed in the battery block while charging/discharging to handle swelling and I keep a fan on them.

- I'm only charging two modules at a time and they are at least two modules away from each other (every other module seems a little too close).

- Yes, it is taking forever with only one dual charger, but my goal is to put as little money as possible into this. I have plenty of time on my hands.

- I've read that the modules should be discharged down to everything from 5.8v to 6.4v. My charger doesn't let me manually set just any discharge voltage. Instead, you can select from pre-defined discharge "cell count" settings. The "6S" (six cell) discharge voltage is actually under 5v, so I have mine set to "7S" which is 5.6v. The "8S" option is 6.4v, which I think is a little too high. That said, there definitely is an "avalanche" effect where the voltage quickly drops from 6.4v to 5.6v usually within about another 500mAh of discharge.

- My charger lets me discharge at (up to) 2A, but I have yet to see it discharge at anything higher than 1.4A. As such, discharging takes much longer than charging.

- I have the charge capacity cutoff set at 7500mAh, but the problem with this is that my charger throws a safety error and cancels the auto charge/discharge program when it's triggered. I've tried playing with the delta peak settings, but it'll sometimes throw false alarms and not charge the battery as much as I'd like. As such, I keep delta peak very high and I manually run each charge / discharge cycle and log the numbers.

Now that I've settled on the above techniques, I'm starting the process over with each battery.

I believe there is a law of diminishing returns with it comes to the amount of time spent reconditioning the battery array. Yes, you can spend quite a bit of time getting the batteries back up to 6500mAh+ each, rearranging them so that capacities match in each block, and then balancing the open voltage of the entire block. Or, you can just try to get your car back onto the road as soon as possible, hopefully for another year, before you may get another error and have to replace a few more modules. I've even read that the Prius balances the block voltage by itself once reinstalled and driven for a while. You'll only get a red triangle if a block's voltage is off by more than 1.2v, I've read.

Having something like Torque Pro running full time while you drive so that you can monitor the voltage of the blocks is pretty helpful for predicting coming failures. I'm definitely going to keep a couple spare modules on hand since they will get harder and harder to find. And at some point, though, I'll probably bite the bullet and spend the $1600 on a new set of cylindrical batteries. The car only has 125k and I plan on keeping it until the wheels falls off.

Best of luck...

 

Hi Jonas,
Thanks for all of the detailed information. I fussed around at first doing high current discharges through a headlight but have decided that the 5W maximum discharge that my charger does for automated cycles seems to get the same results, although much more slowly. I have decided that I will do a battery resistance test and final load test at the very end after normalizing voltages of all of the reconditioned modules. I had one obviously bad module with a shorted cell and a several low capacity modules. All of the lower capacity modules have shown steady improvement with multiple cycles but I don't know yet if they will have a larger internal resistance or self discharge problems.

I am still trying to find a known reliable procedure for load testing and measuring internal resistance. My understanding is that the IR test requires a precision high wattage resistor for making the Delta V measurements. A light bulb has a variable resistance at different temperatures and currents. Have you come across anything that you would consider definitive for either test? I am planning to go through the whole pack and try to get things as dialed in as possible. The car was purchased as a project and I am looking forward to driving it and deciding if it will replace my 2000 Saab 95. If I decide to keep the car long term, I will also most likely go with new cells if the battery needs rebuilding again. I am hoping that the demand and increasing market for prius modules will drive costs down and provide some more options. We'll see. Good luck with your project. Best, Will

 

I would love to find a way to decrease the discharge time, but as it stands I can't get my discharger to get anything above 1.4A (even at the 2.0A setting). It takes 4-5 hours to fully discharge down to 5.6v, which is pretty frustrating. I broke down and bought a second dual charger to speed up the process.

I wouldn't stress too much about internal resistance. In my experience, if a module's voltage drops big time as soon as a load is put on it, I replace it. Otherwise, it seems the voltage drop rate can improve as you recondition the battery. I've been having really good results draining my 13 year old modules down to 5.6v. I've had modules that in the first round had a discharge of only 2500mAh, and I was able to get them up to 6500+ (a few I got up to 7400mAh after 4-5 cycles).

I've also been playing more with the delta peak setting, and 8mV seems to be the sweet spot for reconditioning. Some batteries triggered the delta as low as in the 4000's on the first round, but I was eventually able to get them to take 7500mAh without triggering (but BOY they sure got puffy).

Chuck the modules with obvious bad cells and those whose voltage drops below 7v within a minute. Those need to be replaced for sure. The rest should be cycled a few times and then you can do the "one minute test" to see how you feel about the voltage drop. I discharge multiple batteries at the same time and watch to see if any drop faster than the others.

 

Prolong® Battery Module Load Tester - Hybrid Automotive

Not cheap, but makes load testing easy.

I use that and the timer on my iPhone to run a headlight globe on each module for two minutes.

All the numbers go into a spreadsheet on my laptop.

 

Thanks for the link. I have heard mention of a "two minute test" with a halogen bulb drawing ~5A but didn't realize that it was part of the Prolong system. It seems that a prerequisite would be a uniform or leveled starting voltage for all of the modules. It is not clear to me what the criteria is for disqualifying a module. The information on the Prolong site is a little vague.

I have just started entering my charging data in a spreadsheet from numerous pages of notes. It will be nice to be able to scan rows and columns and see the differences at a glance.

What is your thinking on the "too high" modules? Is the higher capacity or voltage indicative of a particular fault? I am discharging to 6.3 volts based on what I thought looked like an rapid drop in current below that level. I was worried about over taxing the cells but maybe I am not stressing them enough? The whole delta peak thing remains a bit of a mystery to me. My charger does not show any indication of it entering the peaking phase. I have left the settings at default which I believe is 5mv.

I am about halfway through conditioning my battery pack now and plan to load test all of the modules at the end. I know that I have at least 1 bad module and a couple that could be marginal based on capacity. I would still like to figure some good way to measure IR to see how it correlates with higher fills or larger voltage drops under load. I bought a DC amp meter and that may replace the need for a precision resistor. I will let you know if I come up with a procedure that has repeatability and accuracy in calculating IR.

 

The criterion for a good load test is quite simple. Start with all modules at the same voltage. Put each under the same load for the same duration monitoring the voltage. The drops should all be small and about the same. Each should recover to the same voltage after the test. Discard any modules that drop significantly more than the others.

JeffD

Posted via the PriusChat mobile app.

 

I would say they should actually be at similar state of charge. As opposed to voltage.
Same voltage can be very deceiving. Especially if it is unloaded voltage.

 

I’ve had one particular 2008 Prius sitting idle for about two years now, so finally put together a hybrid battery for it.

And got the red triangle straight away.

No worries… turn the car off… back on… red triangle.. turn car off… back on… red triangle…

Time to get the laptop and Techstream out, to find a P0AFA code. Which I’d never heard of.

Searching through PriusChat indicated that this code has something to do with the voltage harness.

I assumed that it was a problem with the lower black cable so sourced a replacement.

It wasn’t until later that I noticed that there was indeed a missing tab on the voltage harness, but that particular tab gets hidden behind the metal tab coming off the black cable.

You can also see in the lower harness the nickel-plated busbars that I’m using.







Block 1 was showing a much lower voltage than the rest. Something’s wrong there.



Revving the engine for a while got the block 1 voltage up but not quite up to the level of the other blocks.



And then I got a negative voltage. Time to remove the battery and see where I’d gone wrong.



Ah. Hidden behind the tab for the lower black cable should be a voltage sensor tab. Now I know to double check this in future.

 

Yep. That's a classic for a corrosion surprise.

Easy to pinpoint if you have diagnostic tools to help you hone in on the problem.
Nicely done!

 

Let me pile on with some of my experience. Perhaps your feedback will make it even more useful for the community here. This is what I hoped to find in a search but figured out the hard way. I hope it isn't already posted somewhere because I would rather had learned this the easy way! This has taken me weeks to figure out. That downtime is okay because I my autocross/track car is still street legal-ish.

Most charging cycles end up at the charging capacity limit set in the 'user settings' of the charger. I tried a few settings with maximum of 8200mAh and almost always reached the cutoff instead of detecting a full battery. This occurred using both chargers listed below. This is a hassle because each D->C cycle must be started manually and one brand of charger does not readily show the discharge performance. The chargers hit this capacity limit in D->C cycling, C->D cycling, manual charging, and auto charging. I tried messing with delta peak settings to no avail. I really don't know what I'm doing when changing that setting from default.

I started cycling my modules with a Prophet Sport 4x100 Charger. This charger appears to be reporting higher discharge mAh than what is actually occurring. My 2005's modules with 110K miles are all showing 6200-6850mAh discharge to 6V. This charger is good for cycling modules and comparing results, but is not accurate enough to compare these results anywhere else. I opened a support case with them to see if there is anything I can do about this.

I got four of the Turnigy Accucel-6 80W chargers, discharged them to 6V, charged them (usually to the 8200mAh capacity limit) and got 5100-5600mAh - a much more believable number. In all, the Turnigy discharge numbers averaged 85% of that reported by the Prophet. I have no way of verifying the Turnigy numbers.

If you run a D->C cycle with the Turnigy and it reaches the charging capacity limit, it does not show you the results of the discharge... unless you start and cancel another cycle, then the discharge value shows up as the second D->C cycle in the results. The Prophet charger always showed the discharge value, but those values appear to be wrong.

For those searching threads - CAPA - means you hit the charging capacity limit set in the "user settings" of your charger.
DRY means it is in the cooldown period between charging and discharging.

I found that my load test results improved as the modules were balanced. I had some modules listed as questionable becuase they did worse than a majority of the other modules. After cycling their results became consistent with the others. The modules exhibiting the lowest of acceptable discharge mAh were consistent with the other modules in both the initial and post-balancing load tests.

I have one really bad module. It shows the worst results in balancing, the load tests, and showed a 0.6V lower voltage than the other packs when I pulled them from the vehicle. My point here is - the modules showing 0.15V lower than the others turned out fine after balancing and their load test results are consistent with the rest of the pack. If I didn't keep track of my load test and charging results, I could not associate the modules' results between initial and final testing - except for that one bad module.

I can't identify any other modules to replace... unless I wanted to get two 5100mAh modules swapped out to match the others that get around 5300mAh. I might cycle those 5100mAh modules one more time. I'm either lucky or ignorant here. Thoughts?

 

So from what I understand (seems to be obvious), no one really charges at 5a. Obviously heat increases the higher the charge rate in amps. What is the fastest you guys discharge/have seen discharged at? My chargers (turnigy reaktor 250w/10a will discharge no faster than 2a and my prophet sport quad is only somewhere around 1.3-1.5 consistently. Any chargers out there that go significantly faster? Looking to upgrade.

 

I've been charging at 5A with a Reaktor QuadKore (the older version; the current version doesn't show the discharge and charge capacity numbers).

NOTE: Refer to my early posts about the Reaktors - they changed software a while ago and the newer models are no good for cycling Prius modules and the Reaktors now don't show the mAh numbers that we require. You either need to find a secondhand Reaktor, or buy a Charsoon Antimatter charger via AliExpress.com.

On the QuadKore I get 2.7-3.1A discharge (ie the discharge increases as the module voltage decreases).

I also have an iCharger 406 DUO - that one discharges at around 10A.

Without getting into regenerative discharging, the quickest discharging charger is the iCharger 4010 DUO, which will discharge at around 20A. That's an expensive charger, though.

Regenerative discharge is where it's at - that's where another battery is used as a power supply, and when discharging the charger feeds the energy from the Prius module back into that battery.

With regenerative discharge you'll get a consistent 20A discharge out of a suitably rated charger.

 

They don’t show during the cycle or don’t show at all even after? I bought 8 singles a few months back and they show during. Reaktor 250W that is.

 

For speed, can a device such as the SkyRC BD200 Battery Discharger & Analyzer be used to quickly discharge the modules to about 6V before charging them to 7250mAh with a standard charger/discharger?

 

Ok so I need some help. I am trying to follow along but I am having issues. So my battery had six bad cells. I bought "new" ones and put it all back together, but another error came up. So now I want to try reconditioning the batter but at a total loss. I want to find a way to do this with a Dc6 charger that plugs into the wall (does anyone know where I can get one or what brand that does this). Then I need help understanding what steps to do with that device. Any help would be great. Thank you.