Gen II Prius Individual Battery Module Replacement

Yeah we like the big guns!

 

I've always wanted to power my HVAC with a Prius battery.

 

Here is a nice little tool for batteries, that I hope alot of the other "battery experts" in the other threads also have...

It charges at up to 6.5A, discharges at 1.25A, and can go between 0v and 30v. It will automatically discharge, charge, and characterize a battery cell to preset voltage values using a current limit mode at the end of the charge to taper off to keep within voltage.

For example in this case I discharge to 12v between 1.25A and 10mA automatically until the current draw is less than 10mA at 12v. Then I charge it back up t o17.74v at between 3.25A and 10mA. In the first picture it is just tapering off the current to keep the voltage at the upper limit and measuring the Ah rating using measurements every 33ms. Second picture is at the beginning of a cycle. It graphs the voltage, current, Ah capacity, and ESR of the battery.

Then it stores all this data and you can "play it back" if you want. It becomes a power supply with the exact voltage and current curves of the battery it tests. Very handy... A steal at $3k.





This was taken as it was still charging at 50W+, and it ended up around 8.5Ah total actual tested capacity. OE modules are 6.5Ah brand new. Woot.

 

anyone looking for some 06 cells? I have 56 from 2 different batteries ... ranging from 7.3v to 7.5v

 

Yup. If you measure the resistance through each module you can also calculate the heat created. P=I^2*R

 

As already discussed several times, a replacement module should ideally be as similar as possible to the rest of the battery pack. That applies especially to internal resistance (IR), discharging capacity, "behavior under load" etc.

A load test, e.g. with a load of approx. 5 amps using a 12 V automotive bulb (55W+65W) for 120 seconds, is a good way to identify bad modules, but is it also a way to say anything about "the state of health" of a module?

I have a number of modules from different battery packs. I checked capacity by discharging them with 2A from "full" down to 6V. All were similar in a range from 5.800 to 6.300 mAh.
I load tested the modules (as described above). Each test started at exactly 7,8V. The tested modules showed voltage differences (before / after the test) from 0.16V to 0.38V.

Is it possible to say now that the modules with the lower voltage difference are "healthier" than the others?

Is the voltage difference (next to IR and capacity) a reliable way to find matching modules for replacing modules in a pack (voltage differences should be similar)?

Thanks!
Thorsten

 

Yes, correct. Healthier modules would have a lower voltage drop than others. You should see most of them group together with a few outliers (typically on the higher side) which often point to bad modules (even if capacity tests out fine). Post up the list of voltage drops if you want help figuring out which are good/bad.

 

For modules with equal capacity, yes. I think he is testing these at higher capacity, so the voltage drop may not be as comparable.

 

Yes, my modules have similar capacity (within approx. 500 mAh).
Which range do you consider as "equal"?
What do you mean with "...testing ... at higher capacity"?

 

Thor,

Your Prius often will pull 100 amps from the HV battery which will result id a higher voltage difference than the one you see at a 5 amp test. If all of the module pairs exhibit the same voltage drop (within 0.3 volts), your ECU will be happy with the battery. A higher voltage difference will cause alarms. So your lower voltage drop modules are "healthier", but the most important thing is to have a matched set. I have told others in the past to pair up modules so that the sum of their voltage drops in a load test all match across the 14 pairs.

JeffD

 

This is a common misinterpretation of what's in the service manual. The 0.3V or more difference needs to be for all seven block pairs not module pairs. And when all seven block pairs meet this condition, it means that the battery ECU is faulty.

Here's proof from page HB-78 P112 Hybrid Battery Control.

 

J-man,

A "battery block" is a module pair. Note that the NO in the flow graph says to replace the HV supply battery assembly which is ALL of the 28 modules. The YES says that the battery ECU is bad (they are assuming that all of the module pairs do not go bad at the same time).

JeffD

 

Jeff,

Many people believe that if there is more than a 0.3V difference between block pairs (say V1 and V2), the ECU will set a code. This is not true. There can be for example a 0.7V difference between block pairs and this does not set a code. I know this because I saw this great a difference between block pairs in my HV pack after I repaired it and it ran that way for seven months and never set a code.

To repeat: A voltage difference of 0.31V between one block and another block does not set a code.

 

Jeff,
thanks. Yes, that makes perfect sense that at higher loads the voltage drop could be different - but I cannot test it ... ;-)
All the modules I am going to use for re-building the pack have a voltage drop of max. 0.02V. I think that's close enough.
So I don't have to worry too much about pairing up modules.

Does it make sense to place the "inner" modules to the outside positions (as they might have seen higher temperatures in their life)?

 

I think there's a place it also talks about if it is .3, or less, or something like that, you replace the hybrid pack. Can't remember though.

 

1. Its OK you place those modules outside. Just swap the middle ones outside.
2. The 0.3, or more readings was for force charging the hybrid pack, and observing the voltage block differences, as @Jeff has pointed out.

 

Testing the modules at higher loads will only amplify the difference between voltage drop results but the test I outlined above is more than adequate to find which modules will play nice together. Jeff is right that voltage drops are important for the ECU to flag a problem and it's best to try to have the blocks (pairs) of voltage drops add up to the same but the reality is this may not be possible depending on what your module test results show. Just be smart about how to put the pack together using best practices and you'll have the best chance at creating something that will go the distance. Typically the modules towards the center of the pack suffer from capacity loss due to heat so for that reason I suggest the lower capacity modules go to the outside and the others towards the middle where heat doesn't escape as easily.

 

No Dxta, that's not what @Jeff stated. He stated

Jeff did not say whether or not all seven block pairs needed to be above 0.3V to cause "alarms". And by the way, there are no "alarms" or DTCs when this condition is met. When all seven block pairs are above 0.3V the service manual simply instructs the user to replace the ECU - again, there is no code or alarm for this troubleshooting step!

To be fair, @Jeff might have meant the voltage difference between block pairs which sets the codes P3011 to P3024.

I did not include this earlier (and really regret it now) which precedes the troubleshooting step I introduced in post #2031.


This code or codes set for one or more of the 14 "module pairs" indicating the voltage for the block is below what Toyota allows. Toyota for its part does not state what this permitted voltage (or voltage difference between block pairs) is. As you can see in the image from the service manual above (page HB-76).

Those who've had a shorted out cell in a module are intimately familiar with the block voltage for a P3011 to P3024 when monitoring all 14 blocks with Techstream or the Torque Pro app. It's 1.2V lower than the others when there is a the shorted out cell in the bad block. Which is the reason most Prius owners opt to replace the faulty pack with a remanufactured or new one because they wouldn't (or couldn't) DIY a repair like many of the engineers who frequent Priuschat have done. Yes, I admit it. I'm an engineer. But only for five more years - woo hoo!

While it's theoretically possible to get a weak block code if the voltage of each module in a block is low enough but not shorted out, I never saw that happen because my HV pack never discharged to a low enough overall voltage to get into this "territory". And it undoubtedly is the voltage difference between block pairs that sets these codes anyway since all 14 block would most likely have lower than normal voltages when the pack voltage decreased well below its normal operating limits.

Which brings me full circle to the urban myth regarding the 0.3V limit difference between block pairs to prevent a DTC - NOT SO.

 

I know this was listed somewhere in these threads, but if I'm looking at block 8, it would be modules 15 and 16, but is module one closest to the ecru and module 28 furthest away? Just would like some clarification as I'm reading lower load voltage on obd reader in block 8.

 

Thanks for the clarifications.