Yes, full cells get "abused" during a grid charge, but it's minimal. The discharges reverse any deterioration.


Oh well.. at least you know what's up with those weak sticks from pack B - clearly in the unacceptable range.


As additional confirmation of the good pack B sticks - when it comes time to pull that pack again, give it an overnight rest and record all the stick voltages again. You will want to note any outliers as they may be deteriorating in ways the testing didn't detect.

 

Stupid question:
It looks like this technology's longevity and durability got worse with each generation. Any guess why Honda stuck with inferior engineering?

 

I can only speculate that they took the path of least resistance and attempted to fix the existing rather than redesign and re-tool.

They started with cylindrical cells and by the 03-05 civic had them mounted at an angle behind the rear seat to minimize the impact on the cargo area. The Gen1 Prius has a huge "hump" where the battery is, and it does impact the trunk space significantly.

Prismatic (prius) modules need to be used in as flat a position as possible. Their construction requires it. Cylindricals can be mounted however you like. The redesign and re-tool to convert the civic to prismatic would have been incredibly expensive.

When faced with the issue of improving the existing or completely changing and re-tooling, they went with the seemingly sensible choice. Unfortunately, in retrospect, the approach they chose likely cost them more.

They went from the Panasonic cells (03-05) to the lower capacity/lower resistance Sanyo cells (less heating)(06-11). Then they "improved" the flow characteristics in the pack for 06-08, which didn't help much and improved it even further in the 09-11, which got even worse.

IMHO, the Sanyo cells are much more sensitive to heat. They've performed remarkably well in the Ford Escape Hybrid (FEH) due to it's air conditioned battery cooling, but when something goes wrong in the FEH with the cooling system, they fail just like the HCH2... consistent and nearly uniform aggressive self-discharge, i.e., they won't hold a charge. I personally own one of these, and there is another user on this site that recently replaced his FEH battery due to the same type of failure.

Honda tried to solve problems incrementally rather than start over from scratch. It ended up being the wrong choice. They didn't realize how big the problem was in 03-05, and they just kept making it worse with each iteration.

They got sued for every model year except the 09-11. They had a 30% failure rate in the first 3-4 years and before most of the 09s ran out of warranty, they gave the CA 10yr/150K warranty to all states in 2015.

I just took a call from a lady with an 09/105K... dealer didn't know about the extended warranty.

 

Wow pack alpha definitely saw some degradation. Coming up on completing the first discharge and if it continues the way it is is probably going to be mid four point something 4 calculated capacity pretty sure it was over 5 just 3 months ago.
odder still the weakest volt taps aren't really sticking out yet ( at 146v now)

14.77
14.66
14.5
14.51
14.53
14.97
14.52
14.61
14.73
14.47

I am sure all will be known once I hit the sticks with the load tester

 

4.35 computed capacity, that's disappointing.

ah there they are.
Down at 115v

12.91.
11.56
11.84
14.12
7.11
11.85
14.14
12.41
8.13
12.65.

 

I think I've mentioned it, but tap voltages at nominal, 144V, are far more valuable at identifying outliers. At 115, even good taps may show issues. Clearly the X.XX taps are the worst and likely have either bad sticks or sticks with massive voltage depression.

Obviously, the 14.XXV taps are likely your strongest. With a little luck, the 4 sticks on those taps are suitable for replacing the bad 3 in bravo.

Keep at it!

 

oddly both charges had to be cut around the 20 hour mark due to large increase in exhaust air temp. I don't have a thermal reading but I put my hand on the exhaust air and it felt noticeably warm.
this didn't happen on the initial charge of pack A 3 months ago. The voltage was only 173 both times.
it's pretty good airflow, the pack would need to be putting out a little bit of heat to make it warm to the "touch"

I'll upload a video attempting to show the airflow tomorrow

the weak tap next to the 14, do you think that could be illustrating a reversed cell?

 

Time, peakV and heat are all signs to shut it down. I prefer to use an IR thermometer and measure the temp. I shut it down at 95°F (+20°F above the inlet). I would try to get to 23 hours on the next charge if the exit temps stay below 95°F.

No "could be" about it. Almost certain. On severely deteriorated packs with imbalanced sticks, you can get reversals starting in the vicinity of 135V. At 115V, you likely had a 5+ reversed.

 

Discharge 2 begin
940. 165v .59amps
1340. 153v .57. End

1030 156. .38. Begin
1340 end. 149.5. .37. End

1120. 152. .37
116. 145. .36. Reading, not end.
155pm 144.0v .36amps. End

By my math the capacity worked to 4.5amp hours.
But this was with the tail end at .36amps drain. And not .54amps like the first one.

Pack A voltage taps at 144v

14.55
14.51
14.30
14.32
14.41
14.58
14.43
14.33
14.35
14.28

No obvious buttheads but this is very low load.
third charge cycle begins tonight after work.

hope to perform high load test on sat.

 

Looks like a 0.3V deviation.

Truly bad sticks show up at this point. When you go below nominal at low current, things get pretty crazy pretty fast.

Your previous cycle at the HIGHER voltage of 146V shows that you had a 0.5V deviation, which suggests there has been an improvement overall. Of note on the negative side is the low tap in the first discharge is the low tap in the second, so those sticks may be on the weaker side. Then again, we aren't talking huge deviations.

When you do the load test, you will likely be able to identify trends in the prior data. Then again, you may find things contradicting themselves - this is why this crap is so frustrating sometimes, and these things preclude simple predictive measures.

There will never be any simple predictive measures until one gets down to the cell level. It gets pretty simple there. Trying to deal with 120, 12 or 6 cells in a group will always mask the outliers to some degree.

That's why I've encouraged you to conduct cell voltage measurements on any sticks you decide are good. It's the last place to check for outliers. Yeah, it's tedious, but it adds one more layer of confidence to the pack's longevity.