• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Problem with JK-BMS B2A24S20P, both showing identical problem with cell 9 on 2 separate batteries

So you actually knew that having a slightly longer 300A busbar was going to cause this your #9 cell to be harder to balance? C'mon you gotta be joking. The vast majority of battery builds on this site are not taking that into account.

I don't think he's joking. I think he's wrong.

Given that a good threshold for imbalance is 20mV, arguing that a difference of 4mV is going to make a significant difference is neither logical nor perhaps even sane.
 
@sunshine_eggo
You feel the longer busbar on cell 9 is not the reason for this issue?
How do you account for this cell 9 issue?
Personally I just don't get why it's happening.
 
@sunshine_eggo
You feel the longer busbar on cell 9 is not the reason for this issue?
How do you account for this cell 9 issue?
Personally I just don't get why it's happening.

Has the cell been moved to see if the problem moves with the cell?

Has the BMS lead been moved to see if the voltage delta moves to the other side of the bus bar?
 
So you actually knew that having a slightly longer 300A busbar was going to cause this your #9 cell to be harder to balance? C'mon you gotta be joking. The vast majority of battery builds on this site are not taking that into account.
I considered the possibility as well as the simple fact was there are no ready made buss bars for the Winston’s cells. I did lengthen all of my BMS sense leads the exact same amount as there is no offset like with the jk/jbd offerings
 
Has the cell been moved to see if the problem moves with the cell?

No for 5 reasons.
1. both batteries are having the identical problem, the probability that it's just a coincidence that cell #9 is bad on both is 1 chance in 256 (i.e. - the probably of rolling snake eyes on two 16-sided dice)
2. because disassembling the compression setup would be difficult as a tiny bit of cell expansion has cumulatively created high tension on the steel rods, the nuts will fly off like bullets if I try to remove them, and I might not even be able to reassemble it because the threaded rods are just barely long enough as they are. (I'm using thick aluminum sheet for compression ends, as recommended by EVE).
3. Others are reporting the same problem with cell 9 lagging on the JK-BMSs, indicating that it's not the cells themselves
4. because @ogremustcrush and especially @rogerheflin both had the same problem and seem to have figured out how to smush this bug
5. because thanks to the folks here I discovered my BMS never had a chance to balance the cells because my charge voltage was stupidly set too low for that to ever happen

Has the BMS lead been moved to see if the voltage delta moves to the other side of the bus bar?

Unless I misunderstood him, mod said that's not a good idea. https://diysolarforum.com/threads/p...9-on-2-separate-batteries.100491/post-1366204
 
Last edited:
No for 4 reasons.
1. both batteries are having the identical problem, the probability that it's just a coincidence that cell #9 is bad on both is 1 chance in 256 (i.e. - the probably of rolling snake eyes on two 16-sided dice)
2. because disassembling the compression setup would be difficult as a tiny bit of cell expansion has cumulatively created high tension on the steel rods, the bolts will fly off like bullets and I might not even be able to reassemble it because the threaded rods are just barely long enough as they are. (I'm using thick aluminum sheet for compression ends, as recommended by EVE).
3. Others are reporting the same problem with cell 9 lagging on the JK-BMSs, indicating that it's not the cells themselves
4. because @ogremustcrush and especially @rogerheflin both had the same problem and seem to have figured out how to smush this bug



Unless I misunderstood him, mod said that's not a good idea. https://diysolarforum.com/threads/p...9-on-2-separate-batteries.100491/post-1366204

Because what you said didn't sound right to me then or now. Cells are typically numbered in order, and talking about #4 and #9 seems insane. Let's use pictures:


1740979185084.png
 
Because what you said didn't sound right to me then or now.
What question are you responding to here? "What you said" ... I said a lot of things in this thread, no idea what this refers to.

If you're referring to the longer busbar theory, that's not my theory. If you disagree with it, why not take it up with @rogerheflin who proposed the theory? I've already said I cannot understand how a slightly longer busbar explains the phenomenon. But I also don't have an alternate explanation. Rather than saying @rogerheflin and @Daddy Tanuki here are insane, if you have an alternative explanation, maybe propose it? 🧐😊 You're the most knowledgeable person in this thread so I'd readily accept what you have to say. But I ain't swapping cells until I try what @rogerheflin proposed.
 
Last edited:
I've already said I cannot understand how a slightly longer busbar explains the phenomenon,

Yes, but do you understand why it's okay to move a BMS lead from one end of the bus bar to the other and how talking about swapping #4 and #9 sounds a little insane?

Do you see how in one configuration the "trouble" cell would be #8, and in the other configuration, the "trouble" cell would be #9

but I also don't have an alternate explanation. Do you have an explanation?

Coincidence.

In-common defect/firmware issue.

It's not like Chinese manufacturers won't hesitate to swap out components for cheaper ones...
 
Yes, but do you understand why it's okay to move a BMS lead from one end of the bus bar to the other and how talking about swapping #4 and #9 sounds a little insane?
Sorry bro, I am not following you. You seem to be saying contradictory things but since that's unlikely, I guess I just don't understand. Maybe you're being sarcastic or something. But if I strike you as insane, fair enough. I've been called far worse.

>Coincidence / firmware etc.
Okay maybe, but I don't understand why you are completely discounting @rogerheflin's solution. Anyway, I'm going to try what he's suggesting because he literally had exactly the same problem I've having. If that doesn't work, then I'll chase the Coincidence theory down. And the possibility that it's a JK-BMS problem is what I originally gravitated to. Thanks for your input btw, I don't discount anything you have to say, I always respect your expertise.
 
Sorry bro, I am not following you. You seem to be saying contradictory things but since that's unlikely, I guess I just don't understand. Maybe you're being sarcastic or something. But if I strike you as insane, fair enough. I've been called far worse.

>Coincidence / firmware etc.
Okay maybe, but I don't understand why you are completely discounting @rogerheflin's solution. Anyway, I'm going to try what he's suggesting because he literally had exactly the same problem I've having. If that doesn't work, then I'll chase the Coincidence theory down. And the possibility that it's a JK-BMS problem is what I originally gravitated to. Thanks for your input btw, I don't discount anything you have to say, I always respect your expertise.
I think you are partially reading the data wrong.

Cell 9 is always being balanced out.

The cell it is being balanced out to moves around. In your first screenshot cell 4 is 3.459 but there are also other cells that are 3.459 and there are other cells that are 3.458 (there are a lot of really close cells that could be picked to balance to/from). So I think if you watch it, the cell that charge (to or from cell 9) is being balanced to is moving around. It may take minutes to change cells, but it should be changing the cell that is not #9 often (at least over the minutes timeframe).
 
Just to summarize (bus bar between 8 and 9 is longer even if it doesn't look like it - for the sake of argument):

1741009518290.png

If a busbar is influencing, moving the BMS lead will move the issue to the adjacent cell. At the + on cell 8, cell 9 will have the issue. If on the - on cell 9, cell 8 will have the issue.

The claim that the bus bar is influencing is testable. Just move the lead, and the issue will move to the adjacent cell.
 
Just to summarize (bus bar between 8 and 9 is longer even if it doesn't look like it - for the sake of argument):

View attachment 282175

If a busbar is influencing, moving the BMS lead will move the issue to the adjacent cell. At the + on cell 8, cell 9 will have the issue. If on the - on cell 9, cell 8 will have the issue.

The claim that the bus bar is influencing is testable. Just move the lead, and the issue will move to the adjacent cell.
i thought he was discussing swapping the lead form like 9 to 4 or something like that. my bust.
 
Really appreciate your post. So are you saying the cell 9 busbar has more resistance because it's a couple centimeters longer than the other busbars, and that's behind these cell 9 issues? It's kinda hard to understand that an extra 2 centimeters of 300A solid copper busbar could make an issue.

But it's amazing that your chart is showing exactly the same issue as I'm facing, that cell 9 goes out of balance as the current drops off.

Is your cell 9 eventually coming into balance during the float with the other cells at higher voltage? From that chart, it looks like the BMS is not managing to balance it. What is the actual voltage gap you're seeing?
Yeah, a fraction of a percent more resistance which results in a fraction of a percent more voltage drop than other cells, which is all we are really talking about when we are talking about single digit mV differences during bulk. Not a problem on its own, but if any "balancing" occurs while that extra voltage drop is present it will compound issues.

Mind this extra resistance from that busbar should affect cell 8 as well (depending on where the balance leads are located on busbar), seems like there may be some quirk of the JK-BMS using 2 leads for that connection (cell 8 pos and cell 9 neg) that is contributing. In my case both of those leads are connected in the same spot in the middle of the long busbar between 8 and 9 though so you would think they would be equally affected though. 8 is usually a little low in my case, but never as bad as 9.

I come into balance fine during absorption, takes maybe 30 min for pack to balance out to <10mv on the worse pack. Other pack also has cell 9 typically the lowest voltage at absorption stage, but it is usually balanced in under 5 min.

Update: actually checking my charge yesterday, I might have actually seen the behavior on cell 8. Cell 9 was elevated during bulk like normal, but cell 8 was the low cell during absorption, cell 9 was actually above average in this case.
Screenshot_20250303_101020_Home Assistant.jpg
 
Last edited:
there may be some quirk of the JK-BMS using 2 leads for that connection
Everything you're saying makes sense but I don't understand this statement. Every cell has two balance leads if you count the negative being connected to the positive of the adjacent battery via busbar. Cell 8 / Cell 9 is not unique in that regard, and I don't think JK-BMS wiring is unique, is it? Sorry if I'm misunderstanding.
 
Last edited:
Everything you're saying makes sense but I don't understand this statement. Every cell has two balance leads if you count the negative being connected to the positive of the adjacent battery via busbar. Cell 8 / Cell 9 is not unique in that regard, and I don't think JK-BMS wiring is unique, is it? Sorry if I'm misunderstanding.
Oh, that might only be a quirk on the JK PB BMS, it has 2 B8 terminals. Some theorize that is because the 16s balance circuit is built out of 2 8s balance circuits back to back, thus the second B8 is the equivalent of B0 on first 8s. Under that train of through the first B8 on connector 2 goes to cell 8 pos, and the second B8 goes to cell 9 neg, but there would be only the busbar voltage drop between these points and in practice they are usually connected directly together or the second B8 is omitted entirely (which implies they are connected together internally and you probably don't want any voltage difference between them.)
 
I did too. That's why it sounded insane to me, like shuffling the order of the BMS leads.
Bruh, first you say "The claim that the bus bar is influencing is testable. Just move the lead." Then 5 minutes later you say swapping leads is "insane." Generally speaking I've learned a lot from you but in this particular thread, between the emojis and the repeated use of the word "insane" with no explanation, I can't decipher most of what you're trying to say. But hey, feel free to chalk it up to me just being dense.
 
Last edited:
Bruh, first you say "The claim that the bus bar is influencing is testable. Just move the lead." Then 5 minutes later you say swapping leads is insane. Generally speaking I've learned a lot from you but in this particular thread, between the emojis and the repeated use of the word "insane" with no explanation, I can't decipher most of what you're trying to say.
Moving the B8+ lead to the B9- lead is testable.

I did measure (on my system 6" 2/0 cable) the voltage between B8+ and B9- and got 7.5mv under 60A charging. So there is a large enough voltage drop to mess with balancing if balancing runs under heavy charging load and the voltage drop pretty closely matches what the calculation says it should be.
 
I did measure (on my system 6" 2/0 cable) the voltage between B8+ and B9- and got 7.5mv under 60A charging. So there is a large enough voltage drop to mess with balancing if balancing runs under heavy charging load and the voltage drop pretty closely matches what the calculation says it should be.
Well that's helpful to know. I still don't understand why this would be an issue that's unique to the JK-BMS. Or maybe it isn't, but I haven't heard of it before. Anyway, hopefully I'll start seeing enough sun that I can test and see if cell 9 can come into balance when the JK-BMS has finally been given enough time and the right voltage to balance.
 
Ok, that I can actually understand. Presumably that's what @sunshine_eggo was saying and I just didn't get it.
What is your start balance voltage set to? I had to set mine up to 3.42 (54.72), otherwise it started balancing under heavy charging. Even today (battery did not hit 90%, and never really charged super heavily) I hit 3.4v a few times with the battery under 90%.
 
Bruh, first you say "The claim that the bus bar is influencing is testable. Just move the lead." Then 5 minutes later you say swapping leads is "insane." Generally speaking I've learned a lot from you but in this particular thread, between the emojis and the repeated use of the word "insane" with no explanation, I can't decipher most of what you're trying to say. But hey, feel free to chalk it up to me just being dense.

Not sure what you're missing. YOU said to swap #4 and #9. @upnorthandpersonal said "bad." #4 and #9 are implicitly not adjacent, and swapping them would wreak havoc on the BMS = insane.

I gave you a picture showing two potential locations for the sense lead, and which wire was in play. I was hoping for the "picture is worth 1,000 words here."

I'm simply talking above moving the sense wire from one end of the bus bar to the other to see if the problem moves. Test this "longer busbar is the problem" theory.
 
YOU said to swap #4 and #9. @upnorthandpersonal said "bad." #4 and #9 are implicitly not adjacent, and swapping them would wreak havoc on the BMS = insane.
Right, I said it.
Right, @upnorthandpersonal said it was a bad idea so I dropped the idea like 25 posts ago.
Why do you keep bringing it up?
And why do you keep repeating the word insane and applying it to various peoples' ideas here, each time with zero explanation?
Maybe start by revisiting the definition of "insane." You're misusing the word.
 

diy solar

diy solar
Back
Top