When the taps get that bad, you have to bypass the pack.

Also, there are transient issues that can disable DC-DC. Sometimes a 12V reset will restore function.

 

The stupid thing still starts off the battery pack...

as for pack B, I'm not sure it i should be concerned.
charge 4 ended at 1105am on the 14th at 178
the following day it was at 165 which I understand is normal, but 2 days later it has lost another 4 volts.

2volts per day? I suppose up high that may be ok as it will lose less when it has less to lose... got the load tester yesterday... Trying to gather disassembly data.

best tap at 16.37, worst at 16.06 and most are pretty evenly spread between.

 

If you look back a page or two, I gave you some guidelines on voltages vs. time. You still haven't accepted the fluid nature of NiMH chemistry and the loosely correlated voltage to capacity/SoC relationship.


It is not at all linear. At higher SoC, voltage/capacity loss is higher than at lower SoC.


Trying to draw relationships where connections are tenuous at best is frustrating. You need to look at everything on a comparative basis.


16.06 to 16.37 is not good. At this stage, anything outside 0.1V from max is suspect unless the max is an outlier.


Disassembly is very easy. 4 bolts securing the corners of the junction board, 4 bolts under the grey rubber plugs, disconnect the thermistor/PTC plug(s). Once junction board is removed, it's very obvious. 10mm hex head bolts security the sticks, cross head screws securing the PTC strips if present and small bolts or screws at the corners of the orange plates securing them to the case.


Lastly, after you have removed the junction board, check for voltage between the PTC strips (on a screw head) to a stick from each half of the pack. Leaks have become more common lately. If you don't detect voltage between the PTC circuit and the sticks, you are likely fine. If you do, exercise caution during disassembly until one of the orange end plates is off, and you'll need to check for voltage between each PTC strip and its associated stick.

 

to be fair, I DID say when its at higher charge it has more to lose (losing .1% at 180V is alot more in absolute terms than at 80V)

I think both the high and the low are outside the normal field but the next closest on both sides still aren't within .1V

Had some other issues this weekend... (https://www.civicforums.com/forums/2...ml#post4729433)
hoping to disassemble this weekend or early next week. Probably in next few days will put pack B in the car (on bypass) and begin a recondition cycle on pack A

Car now always starts with the backup starter which actually makes me feel better because its not killing the good sticks without recharging them anymore.

 

https://www.homepower.com/articles/v...e-battery-pack

this is where I got the idea that even flipped off the pack still had volts at places (if the switch only disconnects 40% you still have 60%!

Pretty sure I see you on there interacting with a guy using his dogs name as his own, so you've read this already too.

90 bolts, fasteners etc? 90!?

 

When you cut a circuit, you cut a circuit. There is no voltage between the end points. The only place you will see voltage is under the rubber plugs - the two "halves" There will be no externally accessible voltage.

20 sticks 2X bolts each = 40
20 sticks 2X screws each PTC = 40
4 bolts on the junction board.
8 bolts or screws holding the end plates to the case.
4 bolts or screws holding the end cover on.

96?

 

the math adds up. I think my fear was that there would be more different types of fasteners to remember what type and where.

I have pack b partially disassembled but then remembered that I would need to charge pack a for load testing.
then the thought occurred that unless I let pack A drop to the same voltage I'm testing B at, I will have imperfect data to determine the top 20 sticks.

even following the plan to run b for a week while A was grid charging again suffered this flaw. Suggestions?

as it is now, I'm optimistic I could get A out of the car and grid it for 24 hours (starting in about an hour). But would need to have reassembly completed about an hour later. Mathematically not possible to test them all in that time.

thinking I may have jumped the gun here lol

 

Not sure what feedback you're looking for, but the comment "let pack A drop to the same voltage I'm testing B at" is certainly not what I recommended, nor is it useful for testing. The purpose is to test the packs for comparison purposes.

If Pack B sits for 7 days following a grid charge, Pack A needs to sit for 7 days following a grid charge. Voltage be ****ed.

To reiterate, resting voltage of NiMH is an extremely unreliable criterion for stick health, EXCEPT to confirm failed cells.

 

ok, so it's absolute sitting time rather than level of self discharge (as measured by volts)? In my little pea brain for some reason I was thinking that since I was measuring loaded voltage that the un-loaded voltage had to start at the same value for it to be valid test

Problem remains I appear to lack a way to compare one set of 20 to the other.

does lightly loaded voltage sag *tend* to indicate behavior (relative to others) at higher load?

Its too bad, I was excited to record the teardown video but I'm thinking I really shouldn't do that now due to time constraints.

seems like this thing comes apart pretty easily I got the side without all the fancy stuff on it off and most of the bolts out in about 10 minutes and the side with all the hardware you got the six bolts mounting that panel and then the four underneath the rubber stoppers before that whole module comes off, right? That's really not near as complicated as I was afraid it would be

 

I have re-quoted the detailed process and explanation from the first page for reference.

Time is what you deal with in a car. It sits for periods of time (20-22 hours/day). You need to know how stick performance varies with time - hence a 7 day wait for all 40 sticks.

Lack a way to compare? The above test procedure gives you EXACTLY that - a way to compare 40 sticks to one another in an apples to apples fashion. You're subjecting them to exactly the same tests over the same time period.

The pack breaks down fairly easily. It's the reassembly that stinks, particularly the alignment of the PTC strip end. Make sure you take pictures of the thermistor wire routing at the junction board end (after board is removed).

I've mentioned it before. The above procedure requires minimal down-time for the car. Basically the time it takes to swap a back between A removal and B install. Then how long it takes to you remove B, break it down and build the best of two packs and install. Touch time is pretty minimal too. The longest time demand will be the load testing at about 1-2 hours.