EVE matched cells, some cells drops very fast

[rgmaninit]

Hello everyone.

I bought matching EVE cells from Dongguan lightning, performed top balancing, and arranged them in 16S configuration (48v 280ah).

When the charging/discharging current is above 20amps, I observe 0.20v difference between the maximum volt and minimum volt, which triggers my low/high cutoff voltages. I think this 0.02v difference is huge (Imagine some cells reaching 3.00v while the rest are at 3.20v).

My busbars and bolts are tightened. Internal resistances of cells are close.

Only one solution worked, which is to re-arrange some cells (in my case, I just swapped 1 cell that drops fast, with its adjacent cells), and viola, I only had 0.02v difference between max and min during 20-30amps charge/discharge. (lowest cell 3.01v, highest cell 3.03v)

But after 2 days, the issue happened again. I posted this issue in a different forum and learned that 3 other people experience this very same problem.

Anyone here who also experience this issue?

 

[snoobler]

How much did you pay for your "matched" cells? You may have been told you were sold matched cells, but unless you paid a noteworthy premium, it's unlikely have have "matched" cells.

Given that you top balanced them, you shouldn't be experiencing any significant imbalance. I note that you didn't indicate you tested them for capacity. Without that information, it's difficult to state anything conclusively.

Additionally, moving a cell should cause the issue to move, not resolve it. This suggests a problem with the BMS.

1. Did you test the capacity? If so, what was it?
2. What is the actual internal resistance, and how was it tested?
3. These 0.2V differences during 20A charges/discharges - are these at the extremes of SoC, i.e., nearly empty and nearly full?
4. Is your BMS set to only balance during charging and above a certain voltage?

 

[rgmaninit]

Around 90dollars per cell. I was told by the supplier that these are "matched" cells. But I've just read what you guys posted and I think these are not really "matched".

I tried to capacity test, but BMS hit low cut-off voltage too early because of 2 "runner" cells. Then I tried to charge-discharge-charge while figuring out what could have been the issue.

#2 - 0.10 to 0.20 mΩ ( see attached picture for the tool they use to show me the "internal resistance" of each cells)

#3 0.2V difference occurs at any SoC as long as we have more than 20amps. At 10-15 amps, the difference is lower and is tolerable.

#4 I have SMART DALY bms but I haven't changed any settings related to balancing. I just notice that it is enabled during charging. There is a setting called "balanced open start volt" which is set to 3.20V by default, and "balanced open diff volt" which is set to 0.05V my default.

Thank you!

 

[snoobler]

Are you using bolts to secure bus bars? If so, are you certain they're not bottoming out in the terminal? Add washers.

Were those measurements YOU took, or were they from a video sent to you?

If YOU took those readings, cells with that low of an IR should not have issues at all with 20A discharges. You have a connection or a BMS sensing problem.

Confirm any BMS readings with a voltmeter.

 

[rgmaninit]

I am using the busbars and bolts that came with the batteries. They are well tightened. What do you mean by "not bottoming out"?


BMS reading is consistent with the reading of my voltmeter. Thought this might be the issue previously, but the reading from BMS is the same with the voltmeter. There must be runners during charge/discharge.

The thing is, at rest, the cells go back to almost the same voltage.

 

[snoobler]

The bolts may be stopping at the bottom of the terminal inside the cells instead of fully clamping the bus bar to the terminals. Add a washer under the head of each bolt to be sure.

They're going back to the same voltage at rest because Your balancing voltage is too low. You've possibly undone your top balance. Set balancing to only during charge, at or above 3.4V and 30mV deviation.

 

[rgmaninit]

I see. That is not the case. the busbars are fully clamped and washers are already added.

But this is are the default settings. Why would it be possible to undo the top balancing?

Actually, I have purchased an active balancer and just waiting for it to arrive. Based on Will's video, it might be unnecessary because we have brand new cells, but I received multiple feedbacks that it is really good to have - so I bought it anyway.

so based on your statement, active balancers will also undo the top balance?

 

[snoobler]

AGAIN, DID YOU TAKE THE INTERNAL RESISTANCE READINGS YOURSELF, OR ARE THEY FROM A VIDEO THEY SENT YOU?

A single washer may not be enough. I recommend you remove at least one terminal to confirm that the bolt isn't bottoming out. Most prefer to install a stud into the terminal and use a nut to clamp onto the stud, so there is no chance a bolt could bottom. I'm not speculating about a possible issue. I'm discussing a fairly common issue. I have 40Ah CALB cells, and the supplied bolts with a THICK bus bar, a flat washer and a lock washer are just enough to keep from bottoming. If I take a single washer out, it will bottom.

Also inspect terminals for flatness and cleanliness.

Top balancing puts the cells in balance at the top of the charge profile. If you engage in balancing at any other point on the profile, you undo that top balancing because you're not at the top. You minimize this by only balancing when you're on the way up and above a certain voltage.

Yes, active balancers will undo a top balance, but they may be powerful enough to maintain voltage balance at every point on the curve.

 

[AussieSim]

Lots of factors can affect the Top Balance, 16 cells is 80kg, your bench might bend slightly causing the busbars to make non-flat contact with the cell terminals.

I found this myself when I used a YR1035+ IR meter in the parallel balance assembly busbars and some cells had a much higher IR than the naked cell terminals.

I also found with the top balance that it is not enough to charge to 3.6+ and immediately disassemble. I top balance to 3.63, turn off the charger, and let them rest in parallel for 48 hours.

 

[Gazoo]

Hello everyone.

I bought matching EVE cells from Dongguan lightning, performed top balancing, and arranged them in 16S configuration (48v 280ah).

When the charging/discharging current is above 20amps, I observe 0.20v difference between the maximum volt and minimum volt, which triggers my low/high cutoff voltages. I think this 0.02v difference is huge (Imagine some cells reaching 3.00v while the rest are at 3.20v).

My busbars and bolts are tightened. Internal resistances of cells are close.

Only one solution worked, which is to re-arrange some cells (in my case, I just swapped 1 cell that drops fast, with its adjacent cells), and viola, I only had 0.02v difference between max and min during 20-30amps charge/discharge. (lowest cell 3.01v, highest cell 3.03v)

But after 2 days, the issue happened again. I posted this issue in a different forum and learned that 3 other people experience this very same problem.

Anyone here who also experience this issue?

Suggest you read this thread:

Active Balancer with Daly BMS, 51.2v 280ah pack? | DIY Solar Power Forum (diysolarforum.com)

You probably received the same screws everyone does. Doubtful they will bottom out. IMO they are not very good. I used 6X20mm grub screws.

Also as suggested make sure your cell terminals are clean. You will probably have to sand them. Another thing is the holes in the busbars are stamped out with a machine which leaves ridges. The ridges need to be filed off. I am attaching a photo of one of my cells before I sanded it.

 

[rgmaninit]

Lots of factors can affect the Top Balance, 16 cells is 80kg, your bench might bend slightly causing the busbars to make non-flat contact with the cell terminals.

I found this myself when I used a YR1035+ IR meter in the parallel balance assembly busbars and some cells had a much higher IR than the naked cell terminals.

I also found with the top balance that it is not enough to charge to 3.6+ and immediately disassemble. I top balance to 3.63, turn off the charger, and let them rest in parallel for 48 hours.

Is this okay? I thought it's bad to keep them at high SOC?

 

[rgmaninit]

AGAIN, DID YOU TAKE THE INTERNAL RESISTANCE READINGS YOURSELF, OR ARE THEY FROM A VIDEO THEY SENT YOU?

A single washer may not be enough. I recommend you remove at least one terminal to confirm that the bolt isn't bottoming out. Most prefer to install a stud into the terminal and use a nut to clamp onto the stud, so there is no chance a bolt could bottom. I'm not speculating about a possible issue. I'm discussing a fairly common issue. I have 40Ah CALB cells, and the supplied bolts with a THICK bus bar, a flat washer and a lock washer are just enough to keep from bottoming. If I take a single washer out, it will bottom.

Also inspect terminals for flatness and cleanliness.

Top balancing puts the cells in balance at the top of the charge profile. If you engage in balancing at any other point on the profile, you undo that top balancing because you're not at the top. You minimize this by only balancing when you're on the way up and above a certain voltage.

Yes, active balancers will undo a top balance, but they may be powerful enough to maintain voltage balance at every point on the curve.

Sorry I didn't get your question. No, I did not take the IR reading. It was only from the video they sent me.

Could you recommend an IR tester? (I noticed that my UNI-T has ohm symbol, can I use that instead?)

Okay understood. New idea to me, I guess if I have an active balancer, no need to perform initial top balancing?

 

[Gazoo]

Is this okay? I thought it's bad to keep them at high SOC?

Yes it's ok for a short time. Over a period of months cells will begin to lose capacity if left at a high SOC.

 

[snoobler]

The unit in the video is a YR-1035 or a variation thereof. About $50 on ebay. Almost always from China.

I am now extremely dubious that the cells have the claimed IR, that they are matched at all, or that they have similar capacities. 20A load shouldn't cause a 0.2V variation on low IR, matched and top-balanced cells.

 

[AussieSim]

Is this okay? I thought it's bad to keep them at high SOC?

They're not staying at 3.6v, as a parallel pack they'll settle down to 3.5v and below together.

When I immediately disassembled at 3.61v their voltages drifted apart as they settled. Keeping them in parallel to settle for 2 days they remained identical voltage on disassembly.

 

[Steve_S]

Another case of miscommunications:

EVE or Lishen or ANY of these for $80 a piece are Bulk Commodity Cells.
The vendors "Match" by looking at the Storage Voltage and verifying the IR is the same or very close AT STORAGE VOLTAGE.
The storage voltage (2.99-3.100 +/-) is how they are shipped from the factory post Factory Production run tests.
Matching & Batching adds a MIN of 50% to the cost of the actual cells.

Please see: STATUS of ESS/Storage Cells being purchased. (commodity cells) (diysolarforum.com)

In January some vendors will be selling Manufacturer FACTORY Matched & Batched Binned cells in lots of 4,8 & 16 which will be tested & guaranteed as matched. They WILL cost a minimum of 50% more per piece. So a 280 at $80 as a Bulk Commodity cell and 280 Certified as Matched will be $140 or more.

EVE, CATL and two other Manufacturers (which I know of) will be providing that to the open Vendors.
Vendors will have the option to provide one or the other.

These commodity cells will generally behave well between 3.100 to 3.400 (their primary power curve). Above & Below these numbers some cells will deviate. but the AH beyond thos points is actually not that significant, roughly 10%, which most reserve off anyway to increase lifespan and cycles.

See here: General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V (diysolarforum.com)

 

[Dzl]

In January some vendors will be selling Manufacturer FACTORY Matched & Batched Binned cells in lots of 4,8 & 16 which will be tested & guaranteed as matched. They WILL cost a minimum of 50% more per piece. So a 280 at $80 as a Bulk Commodity cell and 280 Certified as Matched will be $140 or more.

If that ends up coming to fruition and it is verifiable and more transparent than the current situation that will be a very nice option to have available (if it comes in below the cost of buying direct from the manufacturers)

I suspect many folks will continue buying the currently popular unmatched non-grade-A cells, but choice is good, I know I would strongly consider paying the premium for known quality matched cells, particularly if it includes a test report and documentation like CALB cells come with.

 

[Steve_S]

Well, if budget allows the Matched & Batched is certainly the way to go.
The Bulk cells still do pass the Grade-A batching for "the spec" but that is not the same for EV grades and so on... In the past month or so, I've learned a heap more about all of these shenanigans.

 

[ArthurEld]

So, they will offer matched cells for 3x price.

But we all know that the regular ones work just as well for our purposes.

You could probably get an extra 5% Ah out of the matched cells for 300% cost

 

[Steve_S]

So a 280 at $80 as a Bulk Commodity cell and 280 Certified as Matched will be $140 or more.

How is that 3X the cost ?
How is that Drama class coming, seems to be working.
Sit & have a nice big coffee & ponder a bit.