Gen II Prius Individual Battery Module Replacement

New to the prius battery conditioning, but here it goes. I bought a 2004 prius knowing the battery needed work. Happened to get a full pack on craigslist from another 2004 (for $300), since I knew if I had more then 5 dead cells and issues with ebayreplacements modules, it would be more costly then that battery. Happened to install it in my 04' and it works fine after it recharged itself on the road (took a while) but it seems to even get better mpg then my 2008. At least this lets me buy time while I mess with the original battery. And if my original battery doesn't have too many bad cells, I may buy more cells and just resell the spare battery pack (or even recondition my 08's too).

From the advice of ozmatt, I ended up getting the 1000w reaktor. But I have a chunk of questions (after reading pages 50-70 of this forum) that I would like to ask before diving into this.

1) How should I power the reaktor: I have 2 small 12v 10Ah batteries but am not sure on the life of them (I assume I should use a battery to benefit from the increased discharge rate from the Reaktor). Then I have several computer PSUs but the max spare one is 350W. I also a few spare 200w xbox power supplies which output 12v, but I assume the PSU is a better choice. But I also have access to a 2/10/50 amp car battery "trickle" charger .. I am thinking of using this car charger along with the batteries to charge the reaktor, or? As I am mostly unsure if the reaktor can operate under 1000w..

2) Cooling when charging: Should I remove the packs from the whole battery harness when charging/discharging? As I have seen a lot of setups, anything from using a 2x4 to attempt to prevent against bulging to users using the original vent cover to push air current to cool the battery. Also I noticed the reaktor has a temp gauge, should I place that near the cell when charging for it to properly charge and prevent bulging?

3) Can I only charge packs of cells at a time? As I have seen people discharge the cells individually. I guess the easiest way to ask, is there 28 cells in a gen two or is it 28 packs?

4) I haven't preformed any tests yet. Should I first check the voltages (even though the pack was only at 2 bars when pulled from the car). Or should I charge them all to a certain point and wait say 24h and check them again?

5) How does on test internal resistance? I think the method I described in #4 is that process, but I always thought of an Ohm reading as resistance.

6) Do I need any type of diode(s)? As I assume the reaktor would already be able to direct the current to only one way, but I have seen some people talk about using at least one.

7) What's the point of the delta V cutoff? As I see people recommending 3mV but I am not sure what the purpose of it is for.

8) With my reaktor in mind, how should I run the main tests? Here is what I am thinking:

NiMH
3 cycles of Charge --> Discharge
Discharge at 20A to 5.4V
Charge at 30A (if I can) to 7500mAh
timer limit xxx? minutes, cool down 5 minutes

9) How long should I expect to have to wait per cycle/cell with my charger?

10) Are people just using 12v lights to drain the cells because their charger is slow at discharging? Or would it be more useful to use a shop light (that uses a fixture bulb but runs off 12v source) so I can monitor and plot the change in current (V) of each cell.. Or does my charger already plot this for me?

11) What is the purpose of grid charging? Does that help increase the life of the battery or improve MPG, or should I not even worry about it?

12) When putting the battery back in, is there more tests I should perform? As I have seen the method to max charge it with the car, but not sure how to gather data from that test (like if I need to get a better OBDII scanner, compared to my $15 one).

Thanks for any advice in advance . I am just trying to base info off of my previous electronics knowledge and schooling (chem/physics) lol. Plus I will try to put up pictures or a post of my setup when I have it figured out.

 

Keith, in that TEchnical Memo you've mentioned the warranty is void under any combination of the folloowing conditions: if "nuts" are shifted, if the dates are out of 3 month period or if modules are NOT from the same Batch.
In the first example I put that module from another pack for demonstration purpose and specifically chosen one donor module with only 3 month difference to stress the point - the Batch is different, hence that module is not from that pack.

 

If you power with a rechargeable source, you can use up to 30A to discharge "regeneratively" into the battery. 20Ah of sink is insufficient. I use a 122Ah Deep cycle Marine battery from Wally World. I would still limit the current to 20A based on my own experience.

If you power with a power supply, you aren't going to be charging at more than 6.5A. 6.5V*9.2V = 59.8W + inefficiencies. Note that your power supplies are rated on their total power output. You need to be concerned about the power output of the 12V rail.

Without the regenerative power source, you can still discharge @10-13A. I would set it for 13A and leave it. It will be less at the beginning due to the 80W limit, but it will finish at 13A. I find this to be reliable.

leave in the clamp. Stick the temp probe in the hole on top. Don't worry about swelling. Set formation charge to 6.5A, 6 cells, NiMH sensitivity to 3, capacity cut off to 8450mAh, 30 minute wait between cycles.

28 modules containing 6 cells forming a 168 cell battery in the Gen2 battery pack.

The Reaktor 1000W can handle more than two modules in series, but I recommend working with 1 MODULE at a time.

Initial resting voltage is critical. Record for all modules at the same time.

The Reaktor has an internal check, but it's good only for reference purposes and are sensitive to the quality of the clamped connection and cleanliness of the terminals. IMHO, testing at 13+Amps is sufficiently high to identify outliers, i.e., you can skip the IR test because high IR modules will have notably lower capacity than the others, and the quickie IR test will confirm.

Diode is only needed for whole pack charging. You will only be working with individual modules. The Reaktor flows current in both directions. A diode would be counterproductive.

Please go to battery university and read up on NiMH cells. You need an education in the fundamentals. Since I trust you to comply, I'll answer, when NiMH cells get full at 1C or higher charge currents, they experience a voltage drop. This is "deltaV". 3mV means for a 6 cell battery, it's watching for an 18mV voltage drop to identify it's full.

Absolutely, positively do not discharge to 5.4V at 20A. I have done this in the past, and I don't endorse it anymore. You will not be able to hit 20A without a regenerative sink. You will be limited to 13A.

I do not like charging above 20A because of inefficiencies and heat.

If the Reaktor hits a safety limit, it terminates at that point. It does not go on to the next program. You want your safeties to provide a means of preventing catastrophe, not to regulate your process.

cool down to 30min per above.

1.25 hours per charge
30 minute wait
0.5 hours per discharge
30 minute wait
----
2.75 hours per cycle.

Yes. No. Yes. The Reaktor internal discharge power is 80W, so it's basically equivalent to a 6V, 80W bulb. You get complete charts including voltage, current, capacity and internal (charger)/external (battery) temperature plots. There's more, but you don't need it.

Grid charging is a means of charging the whole pack at once at low current for about 24 hours. When coupled with low current deep discharges, it accomplishes basically the same things as reconditioning individual modules. Read up on the "Quantitative" link in my signature for more info.

IMHO, yes. Code reading options in my signature. MiniVCI is the best option if you have a Win32 based laptop (can use others, but install is more challenging).

Again, go to battery university and read every page on NiMH. Shouldn't take you more than an hour.

 

I sit corrected. I was going off memory. Thanks.

 

Re: 2) Safety warning: I saw the photo and had a conversation with the guy who tried to air cool modules by using compressed air in their garage. That resulted in few modules being blown and plastic particles bursted into their face. They now use face masks and never attempt to use compressed air for that matter. Different sources may suggest different thresholds: 45C but never over 60C (you may wish to convert to F yourself). Temperature cut-off durign tests is must have. I wrote earlier that you can use LM35 temperature sensors with your DIY RC chargers.
Re: 11) Many educated people are still asking that question.....Opinions differ. I am not in favour...

 

FYI: Things for you to be aware off when using DIY tools: It would be difficult to capture this:

I am seeing regularly on discharge graphs in my HV Battery Analysers the steep fall (vs slow decrease) of the individual voltages in modules where the capacity measured @6A down to 6V is about 1.5-2AH. That steep fall happens when one of 6 cells inside the module quickly drops to zero while other 5 cells are still around 1.2V. Those modules are dead and not usable even if capacity was measured at about 2AH. I think I have published few videos where that has been captured.

 

I just came across a customer who drove on a failing battery and literally blew a couple of modules apart.
Bad stuff. Be careful out there people.

 

I recommend 50°C on the Reaktor (it's in °C), and the included probe is a LM35.

Concerning 11), educated people are providing DATA, not opinions. You would not be in favor as informed/empowered consumers will make less visits to the shop thus reducing demand for your system. I'm really not trying to be a jerk. That's the fact. I actually respect that motivation, but when you rely on opinion based on part on financial motivations, it becomes less than honest.

I would LOVE to see you present DATA rather than opinion. I would welcome the debate. I have now presented two sets of data where grid charging was demonstrated to give MASSIVE improvements in capacity. I will be re-testing both of those cars towards the end of this month.

HA has literally hundreds of favorable reviews on eBay. The Honda IMA community will line up and say, "it worked great for me!" Yes, that is all anecdotal, but when ALL the anecdotal data is consistent, it carries weight. You want data? Visit 99mpg.com and check out Mike D's data on the results of grid charging/discharging.




MERGED

Actually, with a Reaktor and it's logging ability, you can capture this with 100% certainty. Furthermore, you are incorrect in your characterization. The rapid drop is typical of the weakest cell in the module exhausting its capacity. A single cell drop-out is characterized by a rapid drop FOLLOWED by a leveling of the voltage and a more shallow discharge slope. This is indicative of a cell reversal and the remaining cells supporting the voltage. This sticks out like a sore thumb on a REaktor log.

Furthermore, testing at higher currents further reveals this condition and triggers cut-off sooner at lower capacities. At 13-20A discharge, it's likely you would see barely 1000mAh out of that same module. Clearly garbage.

Lastly... if you're working with modules that are testing at only 2Ah, they belong in the junk heap. They're done. I won't personally use anything that tests under 4375mAh@20A.

 

Keith, I am aware of solution for Hondas, aware of Mike's design, saw photos of installations, had provided recommendations in relation to additional must-have safety features to be implemented with that grid charger. I have no wested interest in that. That is your business.
Grid chargers installed in Toyota Hybrids is out of scope for me. Not interested.

Chargers in EVs and Quick Chargers are in scope.

My interest and new toy is in front of me - 3.6KW water cooled charger made in Japan.
For going mainstream in battery rebuild business (hundreds of packs) one may consider 5A-10A chargers.
For very small rebuild business operating from the garage - 1.5-2A charger may be a cost effective solution.
For Lexus 40 module pack (or Highlander ) I have portable 1A chargers designed by Toyota/Denso operated from 12V. Those are resque chargers - purpose of using those is when the pack has been depleted too much after stored for too long.

I am not offering and not going to offer on-board grid chargers for Toyota Hybrids. All yours...

 

If I get say a 90Ah battery, would that be sufficient? And to verify, your advising to charge at 20Ah and discharge at 13Ah then (assuming with the higher Ah marine battery)?

I knew there were more cells within each pack (since it's hard to get 1.5v out of one cell alone). But wasn't sure what people were referring to.

Will do.

I figured that, especially with the regenerative.

Makes sense as you explained but I will read the site, forgot about it actually.

What voltage to should I go down to then? And just at 13A, right?

So would you recommend doing a grid charge before reinstallation?

Will look into.

 

Thanks for your methodical questions Michael H. I am having the same questions since its hard to filter thru 67 pages on this thread while also sorting out the mistakes/lessons learned. I am a few weeks behind you; I just ordered a charger and next up will be looking for a reliable ebay module seller to buy 3 modules from in the hope that 1-2 are good.

 

90Ah will probably do it, but the specs matter. It needs to be able to handle the currents. The regenerative process is very inefficient (about 50%). You will be chasing the 12V State of charge. It's important that it not drop too low. Cycling consumes capacity even with regenerative discharges.

No, charge at 6.5A. If you use a regenerative power source, discharge at 20A. If you use the Reaktor's 80W internal capacity, set to 13A.

If you go regenerative, set limit to 15V for the 12V.

6.0V. I also recommend activating decreasing discharge function. If you're doing 13A discharges, set to 10%. If you're doing 20A discharges, set to 7%.

Not necessarily; however, a grid charger is "easy mode" when it comes to balancing. You know you've driven all cells to 100% SoC.

 

One more thing before I reassemble. I've read that the modules should be placed adjacent to other modules with comparable capacity. How important is this?

I've also read that the higher capacity modules should be placed in the center of the pack (if memory serves me). Is this also an important consideration?

 

No. For the most people you are going to do the best you can with the limited equipment you have.
You can take those suggestions and implement them if you want. But if you don't it wont make a huge impact. If it's easy for you to do. Go for it. But if you didn't, mostly likely you wouldn't notice much of a difference. You'll get a chance to do it again in the future and you can try different ideas then as well.

 

I agree with Eric.

It's one of those optimizations that may not make much of a difference. If you have a wide variation in capacities, then it can become more important.

You can simply sort all your modules by their capacities, pair on that basis and put the strongest pairs in the center.

If your capacities are relatively close to one another, say 5-10% variation, it probably doesn't matter much.

As a final note, if you're already breaking a whole pack down to replace multiple modules, then it's a worthwhile exercise since it just becomes a matter of sorting during reassembly.

I had 25% capacity variation in the modules I used for my Gen1 rebuild with Gen2 modules. I sorted them optimally for this reason.

 

I'll stick with my softer charge and discharge rates, until I see a module come back to life, then I will test it under higher current rates. So far, my settings have worked very well, and I am on my 5th pack, with three Prii now parked in the driveway.

Tj

 

Those modules you're tickling just came out of a car getting 80A slammed into them and 120A yanked from them. Furthermore, your < 1C charge rate makes deltaV unreliable, so you need time or capacity provisions for cut-off. Furthermore, you have a much higher chance of overcharging individual cells at the lower rate if the module is significantly imbalanced. What I mean by this is that lets say 5 cells are in sync, but one cell is lagging by 20%. The voltage drop of the 5 cells when they are full will be too low trigger cut-off. This will result in an extended overcharge while the last cell catches up. If you're charging them at 1C, the voltage drop of the 5 will be enough to trip deltaV cut-off and prevent them from being over-charged.

There are basically 3 charge rates:

0.1C for 16 hours
0.3C for 5 hours (or deltaV cut-off, but not 100% reliable)
1C or higher to deltaV cut-off.

You are playing in an area of uncertainty with NiMH. I'm not at all saying they won't work, I'm saying they're unnecessarily conservative and pose as much risk as any you might perceive from higher currents. The only low-risk option is the 0.1C charge for 16 hours. With adequate cooling, there is no downside.

The Reaktor formation charge is also an excellent tool for balancing individual modules. Unfortunately, it takes about 3.5 hours to charge a module; 1 hour at 1C, 2.5 hours at 0.1C as a topping charge.

 

I replaced one bad module in a 2006 with 180K in May of 2014 (original battery). It came back last week needing another module, so it lasted 23 months and over 40K miles.

 

Very nice. Did you make any effort to balance?

 

My balancing methods back then were to replace the bad module with one that matched (the best we could) the rest of the pack. We did a little discharging/charging on the ones we were using but it was nothing like what people are doing now. I think each case is different.