2 pack faling behind with charging and get unbalanced (update 9/07/2020)


MASSIVE restock of EV Batteries at www.batteryhookup.com 5% coupon "POWERWALLS"

barry

Member
Joined
Oct 19, 2019
Messages
66
Well unbalanced pack (other pack oke now) still gets unbalanced..and falls behind (over time, two 2 days) with the rest it is not mutch but it falls behind..
I charged the pack separate of the rest i had a 20 milli head start turned ballancing off now the pack is 5 millivolt behind so a differants off 25 millivolt
The differents off the highest pack and the lowest is max 2ah so that can't be the problem

What i did to check the pack

Charge pack up to 4.1 volt disconnect all the cells and wait for a few days to check for SD, No SD"s
Check for dead/heaters cells, all good
Did IR check all 25 milli ohm
Check for broken fuses, All good
Check for bad connection buspars, All good

running out off idea's
 

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
2,125
Hi @barry....

I have 84 packs of 100p so far - and have had 'balance' problems. I see that you found a bad cell on one pack /fixed that but still have a 'off base' pack. Assume the bad pack is (as you say) not self-dicharging, and is similar cells to others etc etc etc... then here's something I did to 'tweak' a few if my packs on the theory that they just don't have the sameah as the other packs - for some reason.

I bought some 4 cell 18650 holders and popped in some cells. Then used alligator clips to temporarily attach the extra cells to the pack that was falling behind. Out of 84 packs - I had 6 that had this issue. For 5 of the packs, it just took 4 or 8 extra cells to smooth things out. That's 8ah - 16ah'ish additional ah on a 260ah pack - e.g. a 3-6% boost. After a few months / verification that things were 'fixed'- I simply added the cells permanently. This alligator approach can be done live.

Things are just not perfect - especially with a mix of different cells. Iobserve that its diminishing returnsto try to keep all my 84 packs 'exactly' within <0.03v of each other. They vary each day between 0.03v and 0.06v max difference as they charge/discharge between 3.5v and 4.0v/cell. Even though they vary thru the day a little bit, they return each cycle to the same relative state over time. Here's what they look like right now after a few months of no balance. Last balance was May 2020:

image_veclwr.jpg

Currently at 0.04v max difference. This will expand up to 0.05v or 0.06v as it reaches 4.0v/cell and then back to 0.03v max difference as it goes back to 3.5v.

It is reasonable to expect your packs to stay in balance if they are healthy... as a last gasp effort, maybe rebuild the bad pack with other cells? Its still a bit of a mystery as to why things tend toward stability rather than constant divergence- but I have personal proof that healthy packs will stay in balance and others have reported this as well - so don't give up :)
 

barry

Member
Joined
Oct 19, 2019
Messages
66
OffGridInTheCity said:
Hi @barry....

I have 84 packs of 100p so far - and have had 'balance' problems. I see that you found a bad cell on one pack /fixed that but still have a 'off base' pack. Assume the bad pack is (as you say) not self-dicharging, and is similar cells to others etc etc etc... then here's something I did to 'tweak' a few if my packs on the theory that they just don't have the sameah as the other packs - for some reason.

I bought some 4 cell 18650 holders and popped in some cells. Then used alligator clips to temporarily attach the extra cells to the pack that was falling behind. Out of 84 packs - I had 6 that had this issue. For 5 of the packs, it just took 4 or 8 extra cells to smooth things out. That's 8ah - 16ah'ish additional ah on a 260ah pack - e.g. a 3-6% boost. After a few months / verification that things were 'fixed'- I simply added the cells permanently. This alligator approach can be done live.

Things are just not perfect - especially with a mix of different cells. Iobserve that its diminishing returnsto try to keep all my 84 packs 'exactly' within <0.03v of each other. They vary each day between 0.03v and 0.06v max difference as they charge/discharge between 3.5v and 4.0v/cell. Even though they vary thru the day a little bit, they return each cycle to the same relative state over time. Here's what they look like right now after a few months of no balance. Last balance was May 2020:

image_veclwr.jpg

Currently at 0.04v max difference. This will expand up to 0.05v or 0.06v as it reaches 4.0v/cell and then back to 0.03v max difference as it goes back to 3.5v.

It is reasonable to expect your packs to stay in balance if they are healthy... as a last gasp effort, maybe rebuild the bad pack with other cells? Its still a bit of a mystery as to why things tend toward stability rather than constant divergence- but I have personal proof that healthy packs will stay in balance and others have reported this as well - so don't give up :)

Thank for your explanation!! i will try that.
i know that my packsare wellballanced and maybe i look to mutch at this :D when i see your numbers at 0.03 diff mine are 0.006 diff then i have nothing to worry about :Dbut still the one pack is drifting away an maybe over time the diff is to mutch


Thank for your explanation!! i will try that.

i know that my packsare wellballanced and maybe i look to mutch at this
biggrin.png
when i see your numbers at 0.03 diff mine are 0.006 diff then i have nothing to worry about
biggrin.png
but still the one pack is drifting away an maybe over time the diff is to mutch
 

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
2,125
barry said:
i know that my packsare wellballanced and maybe i look to mutch at this
biggrin.png
when i see your numbers at 0.03 diff mine are 0.006 diff then i have nothing to worry about
biggrin.png
but still the one pack is drifting away an maybe over time the diff is to mutch
0.006 diff - WOW. I don't even try for that - when I balance, as soon as it gets to 0.04v max difference I turn it off and go with it. Since I only charge up to 4.0v (with an average of 3.9) I don't even think about it till I get to a100mv max difference.... since that would be 4.1v in the worst case - e.g. perfectly safe.

Would love to hear how it goes when you find your solution.

As @Daromer points out from time to time - its the operation that matters, not the 100th decimal point. My battery bank is 'operating' perfectly - charge up, discharge down, day after day - and the packs all stay sane for month after month and doing the job I need with no maintenance. I would argue that's the very definition of success :)
 

gauss163

Member
Joined
Jun 28, 2020
Messages
278
OffGridInTheCity said:
It is reasonable to expect your packs to stay in balance if they are healthy [...] Its still a bit of a mystery as to why things tend toward stability rather than constant divergence- but I have personal proof that healthy packs will stay in balance and others have reported this as well

It's not clear why you still think it's a mystery when the reasons were explained in this recent thread, viz. the processes that cause imbalance generally occur very slowly in healthy packsand, further, self-discharging tends to decrease imbalances (see below) if the seriesed cells (or packs) are reasonably well-matched.

Generally, for common Li-ion chemistries the self-discharge rates (exponentially) monotonically increase vs. voltage/SOC. So if each cell (or seriesed pack) has roughly the same self-discharge curve, then imbalanced cells will not diverge further, since the higher voltage cells self-discharge at higher rates (and will eventually catch up to the lower voltage cells given enough time - just like a faster car chasing a slower car always gains distance on it, but that may fail if itis not always faster).

However, when your seriesed "cells" areactually "Frankenstein 100p packs" with a motley mix ofchemistries etc this might possiblyfail, i.e.there is no guarantee that a higher voltage series packalways has a higher self-discharge rate (even if that is true for each cell in the pack). But for large packs averaging effects likely increase the probability of maintaining this, which is probably part of what is behind your observations.

If you wish to continue discussion of thisit is probably best to do so in the original thread. I can probably dig up links to studies if that would be of interest.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,760
barry said:
What i did to check the pack

Charge pack up to 4.1 volt disconnect all the cells and wait for a few days to check for SD, No SD"s
Check for dead/heaters cells, all good
Did IR check all 25 milli ohm
Check for broken fuses, All good
Check for bad connection buspars, All good
@barry,

All 25m??
Are these all the same cells from the same manufacturer? I don't mean to be questioning your measurements but I have yet to find 2 cells that have the same IR readings let alone 56 of them from different manufacturers (I'mgoing by the picture of the packfrom here)

Of what I can tell you have a mix ofLG (LGDBMG11865), (Samsung)INR18650-29E,(Panasonic) NCR or CGR18650XXand a Sanyo of some kind in the mix.
It does appear that you are mixing high drain and low drain chemistries in a rather small 56p pack.
Again I do not want to question your IR testing results but I do find it peculiar that all these cells have an IR of 25m?.
Now if you meant a max difference of 25m? between the lowest and highest I can go with that.

Question:Is the cell ratio pretty much the same in all your packs?
The only thing I can think of is that the pack in question has a higher percentage of some cell that is causing it to drop more than the others.

Also a few days to check for SD. Takes a bit longer than that for a SD to show up esp if it is a slow leaker. My minimum rest period is 25 days and I like to stretch it to 30 to 40 days. I then wont use a cell in my powerwall with a drop of more than 0.05V.



OffGridInTheCity said:
I bought some 4 cell 18650 holders and popped in some cells. Then used alligator clips to temporarily attach the extra cells to the pack that was falling behind.

This will expand up to 0.05v or 0.06v as it reaches 4.0v/cell and then back to 0.03v max difference as it goes back to 3.5v.
@OffGridInTheCity
I think the 4 cell holder with alligator clips is a brilliant idea to chase down an underperforming pack.

As you know I built my packs with a bunch of specific criteria in mind IR being one of the paramount measurements along with SOH.
It is interesting to see that your voltage per pack deviation is exactly the reverse of mine as my cells reach ?3.4V the max difference between cells is 0.06V and when my pack reaches ?4.0V the deviation is negligible maybe occasionally 0.01V but mostly straight across the board.

This is my batrium system today slightly cloudy and the AC is running.
As you can see past midnight my bank was depleted to ? 3.45V and the variation is obvious but as the system was being charged the packs started to come closer together and at ?3.8V there is only a 0.02V difference which disappears as I get closer to 4V. Peculiar that our packs run in exact opposites.

image_hoivts.jpg


Wolf
 

OffGridInTheCity

Active member
Joined
Dec 15, 2018
Messages
2,125
>It is interesting to see that your voltage per pack deviation is exactly the reverse of mine as my cells reach 3.4V
I use Batrium Auto-Level to do my balancing - which rums all day and night butconcentrates at night- e.g stable, batterylow voltage. When I'm happy I turn it off 1st thing in the morning - again a bias at bottom balancing.So it seemsnatural for mymax difference to expand as things charge up to 4.0v Did you balance at the top? - maybe that explains things?

I love your close balance... likely due to your diligence in pack construction! :)
 

gauss163

Member
Joined
Jun 28, 2020
Messages
278
Wolf said:
I think the 4 cell holder with alligator clips is a brilliant idea to chase down an underperforming pack.

Pack "patching" (temp. or permanent)is quite common by Li-ion hackers. I've used it to fix powerbanks, laptop packs, etc. that had one very bad cell. Often it's easier than tearing down the pack and doing it right (esp. if you don't have matching cells). It's reasonably safe as long asyou know what you are doing.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,760
OffGridInTheCity said:
>It is interesting to see that your voltage per pack deviation is exactly the reverse of mine as my cells reach 3.4V
Maybe part of the answer is in balancing. I use Batrium Auto-Level to do my balancing - which tends to concentrate thebalance at night- e.g 3.6'ish/cell- and then when I'm happy I turn it off 1st thing in the morning. So its natural for mymax difference to expand things charge up to 4.0v Did you balance at the top? - maybe that explains things?

I do have the batrium set to balance at the top but it hardly ever has to as my controllers are set to 56.7V and once there all cells are even with maybe a 0.01V difference if any at all.

Wolf
 

barry

Member
Joined
Oct 19, 2019
Messages
66
Wolf said:
barry said:
What i did to check the pack

Charge pack up to 4.1 volt disconnect all the cells and wait for a few days to check for SD, No SD"s
Check for dead/heaters cells, all good
Did IR check all 25 milli ohm
Check for broken fuses, All good
Check for bad connection buspars, All good
@barry,

All 25m??
Are these all the same cells from the same manufacturer? I don't mean to be questioning your measurements but I have yet to find 2 cells that have the same IR readings let alone 56 of them from different manufacturers (I'mgoing by the picture of the packfrom here)

Of what I can tell you have a mix ofLG (LGDBMG11865), (Samsung)INR18650-29E,(Panasonic) NCR or CGR18650XXand a Sanyo of some kind in the mix.
It does appear that you are mixing high drain and low drain chemistries in a rather small 56p pack.
Again I do not want to question your IR testing results but I do find it peculiar that all these cells have an IR of 25m?.
Now if you meant a max difference of 25m? between the lowest and highest I can go with that.

Question:Is the cell ratio pretty much the same in all your packs?
The only thing I can think of is that the pack in question has a higher percentage of some cell that is causing it to drop more than the others.

Also a few days to check for SD. Takes a bit longer than that for a SD to show up esp if it is a slow leaker. My minimum rest period is 25 days and I like to stretch it to 30 to 40 days. I then wont use a cell in my powerwall with a drop of more than 0.05V.



OffGridInTheCity said:
I bought some 4 cell 18650 holders and popped in some cells. Then used alligator clips to temporarily attach the extra cells to the pack that was falling behind.

This will expand up to 0.05v or 0.06v as it reaches 4.0v/cell and then back to 0.03v max difference as it goes back to 3.5v.
@OffGridInTheCity
I think the 4 cell holder with alligator clips is a brilliant idea to chase down an underperforming pack.

As you know I built my packs with a bunch of specific criteria in mind IR being one of the paramount measurements along with SOH.
It is interesting to see that your voltage per pack deviation is exactly the reverse of mine as my cells reach ?3.4V the max difference between cells is 0.06V and when my pack reaches ?4.0V the deviation is negligible maybe occasionally 0.01V but mostly straight across the board.

This is my batrium system today slightly cloudy and the AC is running.
As you can see past midnight my bank was depleted to ? 3.45V and the variation is obvious but as the system was being charged the packs started to come closer together and at ?3.8V there is only a 0.02V difference which disappears as I get closer to 4V. Peculiar that our packs run in exact opposites.

image_hoivts.jpg


Wolf

yes my cells are 25/26m? :Dit took me 1,5 year to harvest all the cells (14s60p) andall themeasurements were taken with a YR1030 andallfrom the same e-bike brand make and model (LG/SAMSUNG/PANASONIC 85% off the packs) it took me a long time to get them exept the sanyo


@wolf

Question:Is the cell ratio pretty much the same in all your packs? YES


OffGridInTheCity said:
barry said:
i know that my packsare wellballanced and maybe i look to mutch at this
biggrin.png
when i see your numbers at 0.03 diff mine are 0.006 diff then i have nothing to worry about
biggrin.png
but still the one pack is drifting away an maybe over time the diff is to mutch
0.006 diff - WOW. I don't even try for that - when I balance, as soon as it gets to 0.04v max difference I turn it off and go with it. Since I only charge up to 4.0v (with an average of 3.9) I don't even think about it till I get to a100mv max difference.... since that would be 4.1v in the worst case - e.g. perfectly safe.

Would love to hear how it goes when you find your solution.

As @Daromer points out from time to time - its the operation that matters, not the 100th decimal point. My battery bank is 'operating' perfectly - charge up, discharge down, day after day - and the packs all stay sane for month after month and doing the job I need with no maintenance. I would argue that's the very definition of success :)



image_hyfezc.jpg


@OffGridInTheCity :D :D
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,760
barry said:
yes my cells are 25/26m? :Dit took me 1,5 year to harvest all the cells (14s60p) andall themeasurements were taken with a YR1030 andallfrom the same e-bike brand make and model (LG/SAMSUNG/PANASONIC 85% off the packs) it took me a long time to get them exept the sanyo


@wolf
Question:Is the cell ratio pretty much the same in all your packs? YES



Then you did everything right in my mind @barry.
With 0.006V diff I think you are very much in the ballpark. I would even say you are in the infield.
Other than the 1 cell with the odd flame-out all is well. :) :)

Wolf
 

not2bme

Member
Joined
Oct 16, 2017
Messages
497
barry said:
Well unbalanced pack (other pack oke now) still gets unbalanced..and falls behind (over time, two 2 days) with the rest it is not mutch but it falls behind..
I charged the pack separate of the rest i had a 20 milli head start turned ballancing off now the pack is 5 millivolt behind so a differants off 25 millivolt

You say after 2 days it goes to about 25mV diff. But is this rate continuous? What is this after a week? two weeks? one month?
So if your packs are wildly different then you have some cells that are self discharging over a longer period of time. This is bad.

My packs differ anywhere from 5-25mV a day. I don't care about it. But it never 'drifts'. I can turn off my balancer for a month and my packs will stay the same.

Also what are your min/max balancing points? Are they too high/low that it's meeting the limit of the differences in your packs? Ie. you set your bms to balance at 4.1V and you also charge to 4.1V so one pack that has a lower amp pack reaching that max first then your balancer is discharging that pack thus sucking energy out of that pack. And when it drains it will now be 'unbalanced'. The only time it reaches 'balance' is at the top end (i.e. 4.1V). This the top end balancing. This is not necessarily a problem.
 

barry

Member
Joined
Oct 19, 2019
Messages
66
a little update.. :Dthe pack isin balance now and the diff is now 0.003 so thats pretty nice to see :cool:
i think i made a mistake in sorting i swaped 4 cells with 2000 mah for 2500+ mah and the pack is following the line of all the packs i think that pack whas low in cappacity and over time it whas falling behind with the rest of the packs.
 
Top