Gen II Prius Individual Battery Module Replacement

No, on the Gen II, it is the opposite. Module 1 is at the opposite end from the ECU counting back towards the ECU, making the module closest to the ECU module 28.

 

Thanks so much for clarifying that for me!

 

I've seen other graphs depicting NiMH cells like this. Just looking at an OBDII app. I shouldn't be worried if I see the SOC change increase when charging past 70% or discharging below 45%?
From what I'm seeing normally between 3 and 6 bars, limited the number of pids per second it shows the SOC change every 0.5% if I were drawing a lot of amps or regen-ing from high speed.

Where as if I were at 7 bars and the SOC was at 70% or more, the SOC can increase quicker by 1.5% per refresh or second... Visa versa at the bottom end, in heavy traffic with demisters on, etc.

 

Check out this thread. It is a kit of new cells less expensive than the Toyota pack. There are now over a dozen in use.

Prius Battery Replacement Kit (GenII/GenIII) with NEW custom cells | PriusChat

 

Interesting thread. Something I have used as one of my reference. Have an 05 that now has over 235,000 miles. Such a simple car to fix <except for the MGs>. Recently, I noticed the SOC dropping 1-2 bars after first starting and moving in the morning, an early sign of a module <or more> depleting too quickly relative to the rest. After driving everything is normal. This is after 15,000 miles since I last pulled the pack. Since I had some free time, I pulled the pack after doing a full 18 hour charge with Prolong and put it on the bench and load tested all 28 and 2 modules were outliers. By giving a full charge just before pulling, it reveals which are stressed under load. It also matched what Tech Stream was showing too. Always keep spares so these can go in at anytime. The pulled ones goes through 3-4 charge/discharge cyces and they come back to about 5500mah and low resistance. Only 2 never came back in the past. This is done after the pack is back in the car and become spares. Reshuffled the modules so the pairs were compatible in terms of voltage drop so the computer sees the blocks about all the same voltage, capacity, and resistance. Balanced voltage by hooking all positive together and all negative together for 12 hours. Its gotten to the point that to physically disassemble, test, put the pack together, reassmble back into the car in about 6 hours. Another 12 hours spent waiting for the pack to balance and 18 hours waiting for the Prolong to charge. Probably the last time I do it as I got a hold of 2012 with only 30,000 miles, garaged, and ding free with a CA emission <live in one of those states that is not CA> aka good for 10/150,000. My goal was to go past 300K with the 05, but probably won't see that now or maybe someone else will. Thanks for sharing your experiences, helped me refine my own.

 

So, I recently purchased a Turnigy Reaktor Quadkore. (the 4 in 1 unit) I’ve been attempting to use it to rebuild a Gen 2 battery pack. I chose the Reaktor because of it regenerative discharge option, which lets you discharge at up to 20a, (I’ve been doing 5a) greatly speeding up the process. So far, I’ve only used it on 4 modules, as I’m trying to get all of the settings right.

First, I can’t find a way to see a record of charge/discharge capacities after running the 3 discharge>charge cycles. I can only see the capacity from the most recent operation, which is the last charge cycle. In the product description on the website it says one of the features is “Log file storage: 16 Mb (36 hours)”, which I assumed meant it would record discharge/charge capacities for later viewing. I talked to Hobbyking technical support tonight, and when asked about the log file storage feature, they replied “Really sorry i think that feature is not present in the charger” Am I missing something obvious? Does this charger keep a log?

From what I’ve read here on Priuschat, people recommend charging by capacity, rather than relying on deltaV. On the Reaktor, there is a setting to limit charge capacity. (which I set to 7500mah) The problem is, if it reaches that capacity it stops the whole cycling process, rather than just stopping the charge and moving on to discharging. Is there a setting I can change to remedy this? Of course I could just do individual charge and discharge cycles, but I’d really prefer something that didn’t require constant attention.

I think if I could get these two issues sorted out, I’d be happy with the charger.

Also, just a heads up to anyone else considering buying a Reaktor charger, it looks like they recently updated the firmware in the charger. They sent me one with the new firmware, but the manual for the old firmware. I was really confused for a while. After checking Hobbyking, I found a PDF of the new manual.

If these questions have already been answered in previous posts, please just let me know and I’ll go digging. I’m slowly working my way through the whole thread, but it’s going to take a while!

 

I also use the Reaktor and have the same frustration with using the capacity limiting. Effectively, the Reaktor treats reaching the capacity limit as an error condition and aborts any programmed sequence. I find this to be a major deficiency in their firmware as compared to other chargers such as, e.g., the old B6 chargers. I have succumbed to just using deltaV and setting a capacity limit that is unlikely to be hit. (I think I have them set for 9000 mAHr.) I am unaware of any setting to work around this, but I would certainly be interested if this were added in a firmware upgrade.

As for the data from multiple cycles, I have not tried to use the recording function. On mine, if you press the arrow buttons the display cycles through showing the charge and discharge figures for each of the cycles that was run.

 

Hm, when I press the arrows, I can only see capacity from one operation. Do you have the Quadkore, or the individual unit? I'm wondering if the single unit might have different functions than the quad.

 

Hi everyone,
My Prius has 205K miles and i am rebuilding my battery now after one of the modules failed.
A have a couple of questions:

1. I didn't do a load test, but started cycling modules 3 times. If i a module shows high capacity is it safe to assume that it's in good condition? Or do i need a load test still?

2. I am using settings - 1A discharge, 5A charge, 7500mah cutoff, 5 mins cooldown.
How do i know which modules are bad? Is that a minimum capacity that i should look for? Some of them are only 5000mah after 3 cycles, and some are up to 7500.

Also i had two error codes - blocks 9 and 11, but only module in block 9 is showing lower voltage. Should i still replace module 11?

Thanks!

 

First, a load test will definitively detect a weak module and is a fast east step in the rebuilding process. Second, all of yor modules should have well matched capacities or you will risk having to go through the process shortly after finishing.

The module that was paired with a failed module just needs to be treated the same as the others. It is not more suspect than the others.

Good luck with your rebuild.

JeffD

Posted via the PriusChat mobile app.

 

You can do as little or as much testing as you want. Once you have opened the pack you will likely be faster the next time you do it.

Each additional test helps incrementally in giving you data points as to how good/bad your module is.

You are limited to the modules you have on hand. (ones from your original pack plus any additional you purchased/acquired)

Basically pick the best of whatever you have.

Obviously a 6500 mah battery cannot have 7500mah capacity. But you can compare it in relation to the others you have on hand.

Voltage readings without putting the module under load can be misleading.

For example a 12v car battery may look shows 12v when at rest, but as soon as you try to start the car (putting the battery under load) you may see the voltage immediately drop to 10v or less.

Again, you can choose to do as little or as much data collection as you want. After a while there may be some diminishing returns though.

Some guys get lucky. They test by looking only at resting voltage, take out the low one, and slap in a random module and all is well.

Other times it doesn't go so well. But since their labor is free and they have plenty of free time, and they don't need a reliable car, it doesn't really matter. It saves them money and they are having fun working on their Prius.

 

Thanks!

Since i already started recycling them, would i have to match the voltages before doing the load test? So basically something i could do at the end.

So the charger is wrong then? Or it's a bad module? I have quite a few of them.
Why is that recommended to have a 7500 mah cutoff then, and not 6500?
I also noticed that one of these modules got swollen when it was on third cycle.

 

Not necessarily bad or wrong.
May be that your charger isn't able to tell when the module is really full. So it just keeps dumping power into the module until it reaches the arbitrary mah maximum that was set. (in this case 7500)

You may want to charge the modules while they are in the pack clamped under pressure. Swelling up can irreparably damage a module.

A module has 6 cells in it. They don't all fill up at the same rate. Once a cell is full, the full ones burn off the overcharging as heat (hopefully) while waiting for the others to catch up. Your charger still counts this overfilling, even though the energy spills out as heat and does not remain in the cell.

7500 is a safety maximum mainly so you have some upper threshold to keep the charger from running indefinitely. Especially in the case where a module does not fill properly, or trip the other triggers your charger uses to tell a module us full. (For example, voltage depression)

 

Yes, it is better to start the load test with all modules at the same voltage for comparison, but the relevant result is how far the voltage drops under load and how ell does it recover when the load is removed. Any module that is out of line from the others should not be reused.

You should also equalize the module voltages before reassembly as well. This can be best done by putting all of your good modules in parallel (assuming all of them are close in voltage to start with or large currents will flow).

Eric is quite knowledgeable so listen to him.

JeffD

 

Is it more precise to check the discharge capacity then?
It seems to be way lower. Like around 4500 or so.

Btw i am using Hobbyking Quattro 4x6s

 

The discharge capacity readings are the real capacity of your modules. Note that since you are discharging at a low rate (probably 0.7A which is about 1/10 the rated capacity - 6.5 amps for one hour), you are overstating the real capacity of your modules. It is the discharge capacity of the last cycle that is relevant (hopefully the highest one).

JeffD

ps: Your Prius will be quite happy with modules at 4 amp hours as long as they are a matched set. You will not see a noticeable decrease in performance until the modules drop to about 2.5 amp hours capacity.

 

As long as you have a *consistent* way of assessing your modules then the important thing is how they are relative to each other.

You may be overstating capacity, but at least you are *consistently* overstating it.

And with that, you can still tell which are better and which are worse.

 

The problem is there are losses when you put electricity into a module then try to see how much comes back out.

Load testing is more important than capacity because i can have all test well for capacity and then fail load testing and give you a red triangle while the reverse will likely work just fine for a while anyway. Obviously the more they are matched the better your result is likely to be long-term.

 

Got some results from capacity test. 11 modules still to go, but so far got this:
Most of the modules are 4300-4700mah.
4 Modules are over 5000 (5000-5500)
And 3 modules are low:
1) 2084->2829->3107->3317->3708
2) 2443->3559->4069->4091
3) 2834->3676->3846

So two questions :
1) should i keep cycling the low ones until they get to 4500 or should i replace them?
2) is that ok to have such a deviation in the pack - 4500 - 5500?