I'm scared. Post-reconditioning the HVB overcharged and warbled.

Thank you, haha.

So folks, two things have happened:

1) I heard back from Jack, and he didn't tell me a mAh value for 80% to 40% SoC discharge, but he did say that the test starts at 70% instead of 80%. It's because getting to 80% is difficult, which yeah I definitely am having a hard time getting modules to hold 8.4V

2) Discharger arrived and I've tested 3 modules so far.

Module 9 in block 5: dead as dead, won't even hold a charge it just drops immediately to 6.7V

Module 10 in block 5: 562mAh @ 6.5A from 8.4V (self dis to 8.19V) to 7.2V



--(changed starting discharge voltage to 8.2V)--

Module 21 in block 11: 450mAh @ 6.5A from 8.2V (self dis to 8.1V) to 7.2V

 

So if we look at the amount of time this takes for those two modules, between 255 and 310 seconds, that appears to line up pretty close to the DrPrius test result of 33%. 300/900 = 33%

The mAh measured is a bit strange, but expected to be lower than what people are measuring from 100% to 0% SoC.

More testing required. The process is very fast though and I didn't let any smoke out of the magical box or create a new balloon friend.

 

Ok. I've tested a little over half of the pack so far. It's looking very interesting! Just a reminder for any future readers, I still don't know the actual capacities so what's more important here isn't the Usable mAh values but the spread among the modules. The ebay modules from previous swaps are in the middle and the strongest. I'll have a better idea of true capacity after pairing the blocks together better using this data, then running DrPrius test when the re-assembled pack is in the car. I'd also like to say that this process is very efficient and at least for me, so far, quite safe. No heat or swelling charging up to 8.4V @ 4.5A or discharging to 7.2V. It takes around 4 minutes to charge a module (from 7.6-7.8V nominal) and about 4 minutes to discharge. I cut off charging if the module gives weird readings like going up in voltage too fast or too high. Another early sign of a bad module is that it'll discharge too fast from 8.2V on it's own. Example: Module 1 self discharged to 7.88 before I could switch over to discharging it. Good/Best modules self discharged to or held ~8.1-8.2V.

 

It could be that my best modules in the pack are only 3000mAh. Too soon to know. But after a pack test with DrPrius I can calculate a multiplier for the BD250 capacity measurements. I'm still happy to get this data. It's very clear which modules are bad.

 

Pack tested.



So I'm going to pull out 1, 5, and 9.

Then I went and paired the modules better, at least in a computer, to visualize the difference I should expect to see.



Note, the projection is on the right. It's missing 2 blocks. Up next I get to test the spare modules.

But yeah, I'm satisfied with the results. Jack confirmed DrPrius (and the car) use the weakest block to measure capacity of the pack. So the capacity should go from 33% to ~53% (funny math here using mAh per % using lowest block as reference point, won't know for sure until testing with pack in car). I think it'll be closer to 60%, still not great but better than continuing forward without pairing.

 

This ended up taking quite a while, but more so because I was busy doing other things much of the time.

The verdict:

The pack is back in the car, I didn't have to buy any new modules because I had enough spares. Estimated 77% capacity w/ resting voltage diff of 0.08, compared to 33%. Lowest voltage diff I've ever seen in this car! Wowzers.

What ultimately worked well for me:

1. Pull pack from car
2. Prolong top balance
3. Discharge with BD250 each module individually between 8.2V and 7.2V @ 6.5A (odds first, then evens, or vice versa, cooling fan activated with prolong you should feel the cool air coming through the top)
4. Log pre-reconditioning mAh for operating range in spreadsheet, segment into tiers of weak, ok, and best (this is relative to each other, not a brand new module)
5. If you have spares you suspect might be better than your weakest after reconditioning, swap them into the pack now in place of your best modules (for me I removed the 4 highest operating range capacity modules) OR continue to step 6.
6. Prolong top balance
7. Prolong discharge to 134v / charge / discharge to 101v / charge / discharge to 84v / top balance final charge
8. Now it's time for the second logging of operating range capacities by repeating step 3 and 4
9. Make a new spreadsheet listing 28 modules for your new pack, swapping the weakest N modules (for me it was 4) with the stronger modules you pulled out in step 5
10. In an ideal world, you'd repeat step 7 to boost those stronger modules that didn't go through reconditioning, but I didn't because I needed to get my car out of the garage, and it would require another round of capacity testing for optimal pairing. Not worth it for me.
11. It's pairing time. To do this, I sorted this new spreadsheet by mAh, highest to lowest. Then in a new empty column, top to bottom, I typed "14, 13, 12, 11...1, 1, 2, 3, 4...12, 13, 14". Then I sorted that column from highest to lowest. These are now paired modules, but I did one more step to put the pairs with the lowest capacity difference in the middle of the pack. In a new empty column, top to bottom, I typed "8, 8, 7, 7, 9, 9, 6, 6, 10, 10, 5, 5, 11, 11, 4, 4, 12, 12, 3, 3, 13, 13, 2, 2, 14, 14, 1, 1". Then sort by that column highest to lowest.
12. We now have intelligently paired blocks, with blocks in the optimal order.
13. Prolong top balance & put pack back in car. Discharge or wait before driving so the car doesn't force charge a 100% SoC battery.

Notes:

• Thanks to the capacity logs I am fairly confident that this process more than doubled the usable capacity of my pack, but still need to confirm this with a Dr Prius test
• The point of this process was to be more efficient than measuring capacity of individual modules with discharge rate limited hobby chargers. I had no problems whatsoever with 6.5A discharging, provided good airflow and cool ambient temps.
• It took no more than 6-8 minutes to capacity test each module, closer to 3 hours for my pack but it could be longer if your modules are stronger. So for me, 6 hours total. Most of the time I was watching YouTube videos, pausing when I needed to log a measurement, switch the clips to the next module, and start the next discharge (it auto terminates at 7.2V and beeps)
• Altogether, including all of the top balances, cycles, and waiting 12 hours before unpacking to swap modules, this process took about 1 week. Most of that time was doing things you'd be doing anyway if you have prolong charger & discharger, and most of that time was unattended or with mild attention.
• If you think any of this was unsafe, feel free to speak your mind. I am not suggesting anyone do what I did, but if anyone decides to make that choice of their own will, you have a guide.
• One extra note, letting the pack settle for 4 days did bring to my attention 2 modules with higher than average self discharge but measured with OK capacities. This is worth doing if you have the time. I did this after step 7 and excluded them from the new pack.