What is the recommended charge/discharge cycle to test an individual cell out of a pack?

If you're interested in finding the weakest modules, why not start with a discharge? especially if you have the ability to measure the total mAh discharged from each one.
Remove the battery pack from the car. Let it sit for a day. Discharge each module down to 5.8-6.0 volts and document the mAh number for it. You'll know right away which ones are weak.

When I rebuild a battery the first thing I do is a discharge of each module. Granted, I have enough chargers/dischargers that I can do all 28 modules at the same time, but it's usually very obvious which are the weak modules. Keeping in mind that the Prius maintains a SOC between 40-60%, some modules will show <1000, 2000, or 3000 / 4000+ mAh on that first discharge. When most modules are in the 3000+ range and you have a couple that are in the 1000 range, it's easy to see which ones are causing the problem. Although they can sometimes be recovered, it's typically much less time consuming to just replace them.

 

Mine are setup for 1.5 amp discharge and 2 amp charge, but my cycles are established to coincide with my work schedule so when I'm doing cycles, I get two full cycles a day. One from 5 am to 5 pm and one from 5 pm to 5 am. more or less.

I don't think the rates are critical if all you're doing is comparing modules to each other. More important would be ensuring each module is tested using the same methodology.

 

I was rushing mine, but I tested at 5amp charge, 2 amp discharge, but my charger only ever did about 1.4 amps max discharge. I was also only looking for discharge capacity. Some did swell some but when you discharge them, it fixes it unless they swelled a lot and then they might still be puffy. My charger was doing whatever voltage it wanted to. I could only discharge to 6.4v and then it would charge to around 8.7v iirc. Also they usually only started to puff up around the 8v+ range

Doing that voltage range, that got me to where most of my batteries were just over 2000mah and my best were 4100mah and some of the modules I bought where about 4500mah from a 2014 and 3000mah from a 2006. The batteries I bought were advertised at 7.9v, not capacity or year. (had to decode) Car is a 2010 with 193k on it now and this still didn't fix my p0a80 but I think this is the best performance I've gotten out of the car since I bought it in 2017 with 163k on it. (I did all this in December/Jan) Passed my Texas inspection with the code.

I just keep resetting the code and it's been maybe 300miles since it came on, which is unusual. It's probably going to come on tomorrow now that I've said that. This brought my Dr Prius results from 25% to 55%. I'm waiting until summer to see if I want to dump the car or invest in project lithium. I'm curious if helping the battery, helped the AC, which is in good working condition apparently, but absolutely sucks.

 

This brings up some questions I was meaning to ask regarding the CQ3 settings and choice of equally viable charging options.

1. TMR-JWAP, you have published you charge at 2A, Higgens, you have been charging at 5a. According to the NIMH safety doc, charging at up to 1C (6.5a) is fine if you monitor deltaV, set capacity and temp limits. Is there a reason for 2a vs. 5a charging?

2. TMR-JWAP, you also mention for the CQ3 you set deltaV to 20 as it is unreliable (I believe). The same NIMH safety doc says deltaV is always unreliable below 1.6a so it sounds like a potential weakness of the CQ3? I do see, when charging at 5a that I get deltaV stops that are not terribly "early" to an 8000mah capacity limit (catches at 7500, where 2a always hits CAPA and stops the cycle even with default deltaV)

3. The CQ3 says it can measure IR, but no details as to how, ever tried it or how it is supposed to work?

4. The balloon test seems to be a very good one. What I have seen is a very small initial pop-up (not really inflated), that then leads to an orange sized rapid inflate, then if I keep charging the balloon again rapidly inflates. (this on weak cells I was going to discard) Interesting thing is that the balloons deflate in the same stepwise manner - deflate all at once, not as in slow escape. It is as if the balloon is maintaining pressure (makes sense) and when the pressure drops within the module, the balloon deflates, returning the vaporized contents to the battery. It's a striking pattern. I have taken to stop charge at first, minor erection so to speak, before any real inflation happens as a stand in for a temperature limit. All supposition on my parr but interested in any comments from more experienced folks.

5. TMR-JWAP, you mention using 4 CQ3's and doing all 28 modules at once. Many folks are hesitant to condition adjoining modules - do you think this is from your 2a Charge or just another reconditioning myth? (need space between modules)

Notes, I have been testing different charge rates and found 1C charging is too high for me (plumping, gas discharge via balloons before deltaV kicks in), 5a seems to impact charging cycles better than 2a without the issues of 1C but I have to monitor the 2 hour 5a charge for impact (balloon erection). What I mean is 2a charge/5.9v disc/8000 capacity settings fail deltaV (stop a C/D cycle on capacity) and improvement seems to take more cycles, whereas 5a charge/5.9v disc/8000 give me deltaV protection and require fewer cycles to get to 6000+ capacity but sometimes I get balloon erection and if I miss it, inflation.

So I am assuming:

1. 2a charging means perhaps more cycles to 6000+ capacity, loss of deltaV but you don't have to monitor as closely - set and forget? But I need to change to D/C cycles if I don't want to face interruption in charge from CAPA (loss of deltaV)

2. 5a charging should work as well, but requires initial 2 hour monitoring but seem to get good deltaV (on CQ3) at default in my testing (infrequent balloon impact). Improvement to 6000+ capacity comes with fewer cycles (average 3)

Is there a factor in 5a/2a that I'm missing?

Thanks for all the sharing.

- Jeff

 

My charge setup is based on a couple things. The main limiting factor on these cycles is the discharge, since most chargers are limited to 5w or 10w heat dissipation. I use the max discharge possible for the chargers I have (10W). The charge side of it is based a bit on my work schedule and limiting the heat put into the modules. The two amp charge rate is relatively gentle, coincides great with accomplishing 2 cycles per day, is well within the capacity of my cables and allows my cooling system to keep modules at a very reasonable temperature. Yet, it's high enough that a high IR/bad module will swell slightly, making itself clearly known on the very first cycle. It also allows me to hit the start button and walk away with practically no fear of a catastrophic event occurring. It also does not cause the pressure relief to lift, even on a bad module. The schedule is most important for me. I record data and start a new cycle at 5am, get home from work and record data and start a new cycle between 5-7pm. It works extremely well for my schedule. I look at the battery pack for about only 10 minutes a day.

The delta shutoff will still occur on low capacity modules. It's fun to watch capacities increase each cycle and typically, every module starts reaching the charge mah limit and none charges stop early.

I can cycle all 28 simultaneously because I have a good cooling system. It's essentially exactly the same as the system installed in the car, but with the fan running at full speed 100% of the time. I have the battery case bottom and lower ductwork connected to an OEM fan, which is powered by a prolong charger unit and harness. Fan blows full force into the area under the modules and is forced upward between the modules. Heat is not a problem. The air coming out gets noticeably warmer when the charge cycle gets near the end, as excess charge is converted to heat, but nothing warm enough to be of concern.

I've used this exact "recipe" on a set of brand new modules, and probably (100 or so) 5k-10k mile modules to get data. That data gave me something to compare other modules to. If a new module tests out at ~7000mah using my 'recipe', I know a used 2007 module that tests out at 6600mah is in pretty darn good shape, and so on and so on. My numbers don't have to be super accurate, they just need something to be compared to.

 

That's a fantastic summary of the why behind your setup. Very much appreciated, especially since I am using the same charger.

Thank you!

- Jeff