Using the Universal grid charger

[bluesight]

I guess something unexpected did happen with the maintenance charge. It ended with a code 7, max milliAmp hours. This was way higher than previous charges at 7004mah, versus 4200mah for the last maint charge after a soak/discharge cycle. But the previous attempts failed because of the SampleTime being set to 30 vs 45. Maybe setting to SampleTime to 60 is too much of a stretch for detecting the flat top of the curve? Otherwise, MaxChg_mAH is going to have to be set higher to reach a flatter part of the curve? There appears to still be 300min left in topping time that wasn't yet used, so that should not need to immediately be increased.

Not sure, but something seems odd about all this...

B

[Mike Dabrowski 2000]

That sounds normal.
The MAXMAh is set to 7000, the charge stopped for R 7, @ 7004, as it should.
The sample time resets to zero if the pack keeps rising more that .4V during the sample, then 60 minutes more without a rise of .4V could run almost as long as a soak charge.

What goes back into the pack in MAH is meaningful for cycle to cycle comparison only if the sample time is the same for each of the cycle.
Some well used packs continue to rise and do not plateau.
A healthier pack will plateau, and then actually start dropping, as the temperature starts to rise, so looking at the Temp difference right at the end of the charge will give you more confirmation that the pack is thoroughly charged. Can you post what the curves look like?

[bluesight]

I'm not sure any of these will tell you what you are asking for. As I say, I messed up by hitting the close (X) on the data logger when I meant to just hide it (-). But I restarted and got an end-of-charge portion for the last 15 minutes. I can't get back a graph of temp data in the compare tab (I don't think), so I did what I could with an excel graph merging temps from the csv files. The temps are all in what looks like a low range (mid 50s F). It was cold outside by the end. There were some higher temps in the mid 60s when the outside temps were higher and at the end of soak and during discharge (I think). Temp diffs did go up a bit near the end of the topping charge, but it didn't seem like much.

So the 7000mah charge limit is good for stopping overcharging if a plateau isn't reached? Is that as important, or more so, than the topping charge time limit? I was thinking of trying to push the mah limit up to get closer to the time limit, but not if it is a bad idea. The topping charge did take a long time, getting to be a good fraction of the soaking charge time, as you mention. But the shape of the graph is a lot different when coming from my 138V discharged pack.

It's all interesting data, but hard for me to interpret for good or bad behavior. Ideally, I would like to see the discharge get to 135V (I think) so I don't get code 6. And I would like to see the topping charge curve flatten out to know it is really topped out.

B

[Mike Dabrowski 2000]

It would be useful to see the whole charge as the slope change at the end, and the slope and temp graphs will give us the whole picture.
I would suggest that you raise the
MaxChg_mAH* 8000
MaxChgTime* 1500
TopTimeMN* 1200
These values should allow the 60 minute sample time plateau to stop the charge, so we can see if your pack actually has a plateau.
DO you have a Maint charger or GCO1?
I don't have peoples IC screen names, so I cant cross ref to my owners list.
The charger SN will also get me to your information.
Mike

[bluesight]

The attachment in post in #580 actually shows the full topping charge except for the last 15 minutes, which continues pretty much the same. The scale on the temp graphs there may not be that useful. But the temp attachment above shows most of the temp graph over the same period. The little glitches at the end are probably due to leaving the hatch open for a more extended time while working on the computer.

I will try the new settings at my next one-cycle conditioning probably in a month. The MaxChg_maH should make the most difference since that is what stopped the topping charge this time. I only used 623 of the 900 TopTimeMN setting, and 809 of the 1400 MaxChgTime setting for this cycle, so upping those might help, but probably won't be the limiting factors.

I have the Grid Charger with the Discharger. More info in a PM.

B

Originally Posted by Mike Dabrowski 2000

It would be useful to see the whole charge as the slope change at the end, and the slope and temp graphs will give us the whole picture.
I would suggest that you raise the
MaxChg_mAH* 8000
MaxChgTime* 1500
TopTimeMN* 1200
These values should allow the 60 minute sample time plateau to stop the charge, so we can see if your pack actually has a plateau.
DO you have a Maint charger or GCO1?
I don't have peoples IC screen names, so I cant cross ref to my owners list.
The charger SN will also get me to your information.
Mike

[bluesight]

The SampleTime setting I used at 60 appears to be lucky to have worked for the maintenance/topping charge. I have a hard time evaluating the scales on the graph, (in the #580 post), but it appears that the high-current charge ends at around 16200 on the scale, and the voltage does not rise to the same level again (under the low-current charge) until about 17500. I am kind of assuming each of the intervening 1300 points represents the 5 second sample interval. 12 samples per minute gives 1300/12 or about 108 minutes, a lot longer than 60 minutes. This is also what I seem to get when looking at the csv file, though it isn't quite so clear cut due to what appears to be noisy data.

At this point, I'm not sure why SampleTime 60 minutes even worked. And it would appear 45 minutes should definitely not work. And it is no surprise that 30 didn't work on my first 3 cycle attempt shown at post #533. That is unless I really don't understand reading the files/graphs or how SampleTime actually works. I expect the luck was that the initial sample wasn't at the peak of the high-current graph (165.8v according to the csv file), but back a little ways.

Getting SampleTime right may be tricky for deep-cycle charging. I didn't see the problem for maintenance charging my already charged battery (per post #553) where the voltage recovered within 30 minutes (SampleTime set to 45 min).

As I say...Tricky...Maybe...

B

[Mike Dabrowski 2000]

Ok, I forgot about that graph of the full charge. What iseems to be happening with your pack, is that some of the cells in the pack have the same issue as this subpack,
http://99mpg.com/blog/whatactuallygoeswr/subpack19/

where we can clearly see the difference between cell 1, and 2 with a clear plateau, and the rest of the cells, especially cell 5, where the voltage keeps rising, and it seems to never plateau.
Here on stick 10:
http://99mpg.com/blog/whatactuallygoeswr/subpack10/
we see cell 4 has very high IR compared to the rest of the stick. Cell 4 did eventually plateau, kind of softly and at an elevated voltage.
On subpack 14,
http://99mpg.com/blog/whatactuallygoeswr/subpack14/
we see a better matched but still low capacity group of cells, that as a group have about the same IR
Here on Subpack 1,
http://99mpg.com/blog/whatactuallygoeswr/subpack1/
we see a subpack that was unbalanced
See initial Volts, where cell 2(red) and 4 (blue) were quite a bit lower in SOC than the rest.(Upper left trace)
After charging, we see the voltage on those two cells never plateau, but they do get to the same voltage as the rest of the stick, long after the rest has plateaued.(Upper right trace)
Then on the bottom left, we see that both cell 2 and cell 4 had a different discharge curve than the rest. (Lower left trace)
This was a single cycle right after they were pulled from the car.
Several more cycles may have brought cells 2 and 4 back in line with the rest.
This was my original pack that ran 158K 80K of which was with MIMA.
Because your pack runs so high in voltage, and seems to never plateau, there may be several sticks with weak/high Internal Resistance cells, You can see from my old pack sticks that there are several subpacks that were balanced and low capacity, but many others had cells with high IR, and that is the killer once the condition gets too far.
Reconditioning can lower IR to some extent.

My pack was pulled when it set the 1449 code, and the code would come back right after starting the car, so I expect that the car does a test of the battery during the engine crank?
I am nearly setup to start the MAXIMA and KingKong subpack and cell test series.
I am also starting to test my Mode 8 High current subpack IR test device, which may be our best way to test the subpacks for that important aspect of a subpacks performance.

[bluesight]

So, when I do a soak charge, as the first portion of the graph on post #580, should I be expecting the curve to go up to a plateau/peak and then trend down to a slightly lower voltage where it should eventually level out (as your graphs show for individual cells, and especially for subpack/stick 14)? This would have to happen within the soak time limit, I assume.

I am not set up to look at the individual sticks or cells, so I will need to do what I can with the entire battery pack as a "black box". It sounds like doing more cycles with some adjustments in settings, and maybe manual observation (to stop the thing if it looks like it is getting out of hand) might be in order. I will likely wait until next week to avoid fighting really cold overnight temps. It is supposed to be in the 30s with lows in the mid to upper 20s next week. If temperature is a contributing factor in getting individual cells to behave better, maybe expecting ideal results in the cold is a bit much to expect.

I don't think I will need another soak charge, so it will likely be separate 1 cycle discharge/charge operations and try up to 6 of them with parameters that get me as close as possible to a reasonable plateau on the top-off maintenance charge (increased maxchg_mah, for example). I am not sure if there is a way to push the discharge harder by further increasing the SlopeDifMax beyond 30.

B

[Mike Dabrowski 2000]

The soak charge in the post 580 graph was still climbing, so it is possible that the plateau will not happen, until the voltage gets even higher?
I looked through my pack cycle graphs, and don't see a nice example of the whole pack plateau, but all of the graphs are from weak old packs, and we are only talking about 10 or so packs.
We need to remember that there is a very wide range of pack conditions, and that some will respond to reconditioning with the charger/discharger and some will require stick replacement.
I think it is safe to say that if your pack plateaus t a nice low number like 168-170, that it will be a much better pack than one that exhibits the non plateau and rises to well over 175V.

Cold temperatures could have some effect on the graphs, and once I get set back up, I will run some test at the stick level graphs on a stick that is cycled outdoors (6 F this morning) ior in the garage (25-35F) and then do the same stick in doors to see what the difference is.
I just wish I had a couple of un interrupted weeks to get deep into this testing and data gathering and try to get some hard data to help us better identify the signature of the different deterioration effects that a pack can exhibit.

I will be running some graphs on my MAXIMA rebuilt pack as soon as I get it together, so we can finally have a baseline of what a pack of new sticks looks like.

Eli, have you ran some cycles with the charger/discharger datalogger on one of the rebuilt packs?

Yes the pack is a black box with 120 cells that can have widely different conditions, as we can so easily see in the cell level traces, so even if we were tracing the pack and graphing each stick, it would still be a fuzzy picture, so our task is to learn how to interpret the information we can get from the charger/discharger/datalogger, to determine when we have to pull the pack and start working at the stick level.
The stick level IR tester charger( mode prototype is ready to test, which I hope will give us a way to quickly determine which sticks are the bad actors from the important IR standpoint.
This test will hit the sticks with an adjustable 60A pulse load, and measure the voltage drop.
Today is test day for that prototype.

[bluesight]

It would be useful to be sure about what the absolute worst case conditions are to avoid damage for discharging and charging. At this point, my understanding is that you would NEVER want to go above 180V (on the Insight Gen1 pack - 1.5V/cell - that is default value) for charge, and you would NEVER want to go below 120V on discharge (1V/cell, though the default is 129V). There must be an outlet/pack temperature max that you would NEVER want to get above (I see diff temp and PTC temp, but not sure how to interpret PTC value...seems like indirect resistor reading representing internal pack temp - default is 200). I maybe need to worry about total current charge to NEVER exceed (max mah - default of 7000), though that seems like it should be limited more by voltages and temps. The Max Charge/Soak/Discharge times seem like the variables that can be adjusted to attempt to get to the limits of each of the other important worst-case conditions and see which causes the most trouble.

I am coming to understand many of the macro limits need to be watched carefully to avoid individual stick and cell damage at the "micro" level. This would be particularly true for temperature where one cell might really get hot but not be reflected in the overall pack temperature. And for low voltage where one or more cells might be "dead" at 0, and going below 120V for the pack could bring remaining cells to below 1V per cell and damaging them. Not sure how it works for high volts, though some of the graphs you show indicate some cells can go beyond 1.5v where others are less, thus potentially permanently killing the cells that get beyond 1.5V (probably by overheating, though that does not seem to be reflected in your graphs).

I think the safest bet for me is to push the soak/topping charge times and max mah to hit more against the max total voltage (180V) while watching or further limiting the PTC temp (maybe below 200 if I can get an interpretation of what that is). With the cold weather charging, the PTC value may not be very helpful (it reached only 72 by the end of the topping charge, for example). That doesn't mean individual cells might not be getting too hot.

I can worry about discharge issues later.

B