Runner cell "trigger" voltage

[Don B. Cilly]

I recently got 8 280Ah cells from Shenzhen Luyuan (Amy Wan). Grade A-.
They looked good (very well packed too). No dents, no scratches, no swelling.
But, one cell goes over voltage. It's not the BMS, the cell actually does. Voltmeter confirms. Switching cells order doesn't change things. It's that particular cell.

Now, it starts going over voltage (does so very quickly, takes a long time to balance back).
It does it at roughly 27.2 volts - the cell jumps from 3.38V (like the others) to 3.7 and more.
The BMS cuts the charge, and then it takes some half an hour for it to balance back.

Thing is, before reaching that voltage, it behaves quite normally. Puzzled.
-

​

 

[sunshine_eggo]

By saying "doesn't change things," I'm interpreting the following example:

Cell # 7 hits peak.
Swap cell #7 with cell #8
Position #7 still hits peak even with a new cell in place.

If the above is an accurate example, you have a sensing lead issue. Ring terminal crimp, issue at the connector, etc.

 

[SparkyJJO]

If you've got a bum cell, maybe talk to the seller about a replacement?

I'm assuming you top balanced everything when first building the pack to 3.65 volts?

 

[sunshine_eggo]

If you've got a bum cell, maybe talk to the seller about a replacement?

The info presented does not indicate a "bum cell."

I'm assuming you top balanced everything when first building the pack to 3.65 volts?

Maybe you haven't read it, but prior to parallel charging, it indicates to build the battery with cells in series with the BMS attached and fully charge it. If it fully charges, and you get full capacity on a discharge, parallel top balancing is not needed. Even if it is needed, this method vastly reduces the charging time needed for parallel top balancing.

 

[Hedges]

You think it is balanced?
Keep bleeding off charge until rest of cells reach 3.65V without this one being higher.
Then see how it behaves.

 

[Ampster]

Have you parallel top balanced? As others have said you could try bleeding off that cell with a resistor if parallel top balancing is too much effort or you don't have a power supply to do that.
In the meantime reduce your charging voltage to 27 volts which is about 3.37 volts per cell and you can limp along to use your pack.

 

[Pim57]

It seems that one cell has 2% (from your linked post I read the pack is at 98% SOC) less capacity then the best cell in your battery.

If you were to replace the cell with a good one, you would find another cell will start ‘jumping’.

For non grade A cells you will have to accept some differences in cell capacity. In your case the spread in cells is around that 2%.
Not bad at all I think?

 

[DThames]

I read this and chuckle about as I found 27.2v is my best boost voltage setting.....for the same reasons. It is what it is, as we like to say at work.

 

[Don B. Cilly]

Cell # 7 hits peak.
Swap cell #7 with cell #8
Position #7 still hits peak even with a new cell in place.

If the above is an accurate example, you have a sensing lead issue.

No, it's the opposite. Whatever position I put the cell in, it jumps to 3.7+ as soon as the battery hits 27.1-27.2.

I found 27.2v is my best boost voltage setting...

I charge the whole (I have two other batteries paralleled) to 27.7.
I'll try and charge the lot to 27.2 (and 27.1) and see what happens.
If that gets rid of the protection issue, even 27.1 should be good enough.
I mean, I don't really need 27.7, do I?
-

 

[Ampster]

I mean, I don't really need 27.7, do I?

No, you are in the nineties in terms of SOC at 27.1. The runner might balance on its own once it hits 3.4 volts since the balancing typically starts at 3.4 volts. It might take some time depending on the balancing current from your BMS.

 

[Don B. Cilly]

I'll try. Being almost 8PM here, it's a bit late, but I've set the SCCs to 27.2 and tomorrow will be the thing :·)
This (JK) BMS doesn't seem to have a voltage trigger for balancing, just ΔV (delta cell volts).
-

 

[sunshine_eggo]

No, it's the opposite. Whatever position I put the cell in, it jumps to 3.7+ as soon as the battery hits 27.1-27.2.



I charge the whole (I have two other batteries paralleled) to 27.7.
I'll try and charge the lot to 27.2 (and 27.1) and see what happens.
If that gets rid of the protection issue, even 27.1 should be good enough.
I mean, I don't really need 27.7, do I?
-

You are observing the typical behavior of gray market cells. THAT cell just happened to hit 100% before the others.

They are not matched or balanced based on SoC.

The cells likely shipped at a 30-50% SoC, and they all needed slightly different input to get to 100%. In series, they all get exactly the same amount of input.

Simply bleed off the high cell until the others catch up, which might take many hours/days, or simply conduct the parallel balance charge per the guide, which will likely tank hours or a day or two depending on the amount of imbalance.

Bleeding off the high cell may be similar to whack-a-mole. You bleed off one, then another hits peak. Now you have two cells to bleed off, etc. The parallel balance charge avoids chasing multiple high cells.

 

[Don B. Cilly]

It seems quite balanced now :·)


We'll see tomorrow if I got the "trigger" voltage right (at 27.2).
-

 

[sunshine_eggo]

It seems quite balanced now :·)
View attachment 103673

We'll see tomorrow if I got the "trigger" voltage right (at 27.2).
-

No. Don't fall for this.

The word "balance" has no meaning when cells are below 3.40-3.45V. Your cell voltages are equal. That doesn't mean anything when it comes to state of charge and behavior the next time you charge.

"Balance" refers to state of charge, not voltage.

LFP working range voltage curve is very flat, so cells at substantially different states of charge will reflect almost identical voltage in the working range.

 

[Steve_S]

Again, I will repeat for the Final Time.
One all cells are Top Balanced to 3.6xx IN PARALLEL allow to Settle to "natural Working Voltage.
Assemble pack in Series and connect BMS per Spec.
DO NOT Bother charging ANY Battery in the Bank with over 3.425Vpc / 27.4V per battery.
Float from there to finish the last 5-7% which is Constant Voltage/Variable Current.

SET Balancing to START @ 3.37Vpc which will actively transfer excess while "HOT" to lower voltage cells.
Optimal for LFP Balancing IS 0.010V Trigger.

Also refer to THIS New Post (as of this morning)

Steve_S-TECH: What is BMS SoC and LFP Battery Bank information.

Intro: BMS SoC and system SoC readings are always a source of great debate and even great argument(s). This Posting addresses MANY of the issues not considered, nor weighed and delves into the Important MInutia which can & often does create confusion & misunderstandings that result falling...

diysolarforum.com

New JKBMA Manual is DONE, some minor image edits are being done NOW and should be posted publicly within 24 Hours.

Hope it Helps, Good Luck.
Steve
.

 

[Don B. Cilly]

27.2 still triggers protection.
27.1 does not, but the 7th cell still runs quite high.
27.0 looks fine.

About replacements, Amy Wan has gone AWOL. We'll see next week.
Qishou say they can only ship 4 cells. When I asked why they offer "samples" on their site (1 piece) they did not reply. They seem rude in general.
Exilporc advertise EU stock. They did give me a quote for a 1-piece sample. Shipped form China.
eBay EU vendors (expensive), one (German) says "Sorry we can't ship one", the other (Dutch) did not reply.
Let's see what Amy says (if and) when she comes back to (business) life.
-

 

[Steve_S]

27.2 still triggers protection.
27.1 does not, but the 7th cell still runs quite high.
27.0 looks fine.

About replacements, Amy Wan has gone AWOL. We'll see next week.
Qishou say they can only ship 4 cells. When I asked why they offer "samples" on their site (1 piece) they did not reply. They seem rude in general.
Exilporc advertise EU stock. They did give me a quote for a 1-piece sample. Shipped form China.
eBay EU vendors (expensive), one (German) says "Sorry we can't ship one", the other (Dutch) did not reply.
Let's see what Amy says (if and) when she comes back to (business) life.
-

oivey.
1) Remember the Time Difference ! Luyuan is also a Family Business & not crewed by sales shmucks.
2) start a CHAT or MESSENGER session which works better for communications.
3) shipping cells is crazy as crap, the most effective for ONE Cell shipment is by Air Freight. (not as $ nutty as you would think.)
btw: assumptions to conclusions != smart.

 

[Don B. Cilly]

Funny you should mention time zones. At the moment... my son is in D.C. Six hours less. Amy W. is in Shenzhen. 6 hours more.
My daughter is in French Polynesia (working). 12 hours more (or less, irrelevant :·)

Exilporc gave me a quote for $250 for one cell, including air shipping form China.
-

 

[Hedges]

Still don't see reason to believe you have a bad cell.
Need to top balance to something like 3.65V (Which could mean bleeding off high cells manually), then operate system at a voltage lower than that.

 

[Don B. Cilly]

OK, I'll (re)try the balancing act. Starting tomorrow, it's almost 8PM and I have no grid.
It'll probably take a few days... still, why should that cell jump in voltage and the others not?
-