Aliexpress 24V 100AH battery horrible internal resistance.

[MattiFin]

OK, let's say this is indeed the case. How do I resolve this without access to the pack and no active balancer? I imagine by keeping the battery at float voltage of 3.4 per cell or 27.2V but for how long?

Would you say, if indeed there is imbalance I should see more current consumption at float voltage as the only cell that is below tries to get charged.

How does charging of the cell that is at much lower voltage work? Is it even possible to equalise such a battery?

There is no easy way to fix the imbalance. Per youtube teardown video the BMS is jbd-hp28sa with passive balancing.
I couldn't find the balancing current anywhere but assume that it is not more than 100mA even at best.
Fixing 10Ah imbalance with 100mA balancing current would take 100 hours or about 5 days and we don't know if the balancing current is even 100mA.
If you have adjustable lab power supply set it to something like 100mA and 30 volts and let the battery charge until it hits OVP.

 

[Luk88]

There is no easy way to fix the imbalance. Per youtube teardown video the BMS is jbd-hp28sa with passive balancing.
I couldn't find the balancing current anywhere but assume that it is not more than 100mA even at best.
Fixing 10Ah imbalance with 100mA balancing current would take 100 hours or about 5 days and we don't know if the balancing current is even 100mA.
If you have adjustable lab power supply set it to something like 100mA and 30 volts and let the battery charge until it hits OVP.

Let's hypothetically assume I had access to the inside, but it was a professionally made puch pack encased in a hardest polyurethane foam known to man... Nicely done too, with conformal coating on all connections. Silicone on plugs etc. So there is no, changing it to parallel and top balancing.

Then, let's say I found voltages were

3.38
3.66
3.51
3.64
3.44
3.69 (yes, no mistake!)
3.48
3.49


And the BMS type was GS-8S-JH100A-TG2, most likely 60ma balancing current. Could one realistically do something without messing it up too much to balance it?

 

[MattiFin]

Let's hypothetically assume I had access to the inside, but it was a professionally made puch pack encased in a hardest polyurethane foam known to man... Nicely done too, with conformal coating on all connections. Silicone on plugs etc. So there is no, changing it to parallel and top balancing.

Then, let's say I found voltages were

3.38
3.66
3.51
3.64
3.44
3.69 (yes, no mistake!)
3.48
3.49


And the BMS type was GS-8S-JH100A-TG2, most likely 60ma balancing current. Could one realistically do something without messing it up too much to balance it?

No need to switch to parallel.
You could hypotetically drain the highest reading cells with your electronic load or 12V 55W car light bulb for couple of Ah.
Or alternatively charge the lowest reading cells with adjustable power supply. Or combine both.
3.38 to 3.69 cell to cell variation is enough imbalance to explain your results but it hard to say for sure with voltages alone on lifepo4 chemistry.

It is quite hassle to manually balance the highest and lowest offenders but couple of hours of babysitting with electronic load could bring you to point where the 60mA internal BMS can handle the rest of it.

Kurwa-balancer: Connect the battery to charger, measure cell voltages and apply light bulb load to cell that reads highest for 10 minutes. Wait for 1 minute while charging, re-measure every cell and apply light bulb load to highest reading cell. Rinse and repeat until all cells are above 3.5v while charging.

 

[Brucey]

Nope, computer, for data logging. Pretty graphs etc.


How is "long time"? A day, a couple of days? I have about 10 days left to decide if it's worth keeping.

I've read reports of people with eg4 batteries requiring 20-40 cycles to get sorted out

 

[Luk88]

No need to switch to parallel.
You could hypotetically drain the highest reading cells with your electronic load or 12V 55W car light bulb for couple of Ah.
Or alternatively charge the lowest reading cells with adjustable power supply. Or combine both.
3.38 to 3.69 cell to cell variation is enough imbalance to explain your results but it hard to say for sure with voltages alone on lifepo4 chemistry.

It is quite hassle to manually balance the highest and lowest offenders but couple of hours of babysitting with electronic load could bring you to point where the 60mA internal BMS can handle the rest of it.

Kurwa-balancer: Connect the battery to charger, measure cell voltages and apply light bulb load to cell that reads highest for 10 minutes. Wait for 1 minute while charging, re-measure every cell and apply light bulb load to highest reading cell. Rinse and repeat until all cells are above 3.5v while charging.

So I have used your "kura-balancer" method, but I doubt it'll translate to 10 extra Ah. Why, because it was too easy... All the cells are around 3.51 3.6 all I had to do was to pull two that were highest down by connecting a 0.18r resistor which pulled about 15~13 amps for maybe 2 minutes each. Then I charged the lowest cell at ~4amps for maybe 10 minutes.

Just this allowed the pack to go 0.3V higher. I charged it until it pulled no more current at 28.3 and I run the load test again. Constant power at 160W. Well see if my efforts gave me 2ah at least. I seriously doubt it'll be 10. But who knows, perhaps that one cell was cutting off when the rest of the pack still had 9ah left of juice.

I'm very interested how the voltages will look at the bottom.

I've read reports of people with eg4 batteries requiring 20-40 cycles to get sorted out

Oh, my. That is a lot... But it gives some credibility to all these sellers that say "give it few more cycles" when you report under capacity.

 

[Luk88]

Not much if an improvement
I haven't had a chance to measure the cell voltages at the bottom. I'll do later. I wonder, how much voltage above 2.5 can ~2% SOC be? Because each cell, but one, would have to be ~2% too high for this result.

 

[Luk88]

IDK, to me this looks pretty well balanced. After standing for a number of hours:
2.7
2.809
2.776
2.792
2.769
2.737
2.801
2.849

Why are they above 2.5v? Because the BMS cut off when one got to 2.5 (probably no 1) and then they recovered a bit. Does this look to anyone like all cells other than 1 are ~2% higher?

 

[Luk88]

I've recharged the battery slowly with a supplied charger (5A). Then after standing for few hours all cells were pretty close in voltage. I run another test and it's 93Ah :-(

Kepworth will be going back too if I get a better one I'd really like to have these 7% more. However, I have to give it points for good (low) internal DC resistance. It is the only ready made cheap battery so far that can be used at 1C.

 

[Brucey]

I've recharged the battery slowly with a supplied charger (5A). Then after standing for few hours all cells were pretty close in voltage. I run another test and it's 93Ah :-(

Kepworth will be going back too if I get a better one I'd really like to have these 7% more. However, I have to give it points for good (low) internal DC resistance. It is the only ready made cheap battery so far that can be used at 1C.


View attachment 211292
View attachment 211293

How long did you leave it on the 5A charger? Did it come with the battery?

 

[Luk88]

How long did you leave it on the 5A charger? Did it come with the battery?

Yes, the charger was included. It uses a front side port.

24h or so, but I believe the charger cuts out once it reaches certain volt/amp combination. When I came 24h later the charger was off(~17h in it was still charging) . I measured the cells and I started the discharge test as they all were in the ~3.48V region (I have it written down in the workshop).

Also regarding this charger and BMS. I have no idea if this is a faulty bms or what, but before I top balanced the cells it got one of the cells to 3.71V and it was still charging! I just pulled the plug. This was before cells were balanced.


Additionally I wonder if one can detect cell imbalance by the shape of the discharge curve. See my previous discharge curve with like a little voltage plateau after it dropped, then it finally dropped all the way. After balancing the end of discharge is one solid drop.

 

[Luk88]

Ok, I did get my money back for the cheapest battery on AliExpress's. The battery brand is Ekpaowe and the seller is EKPAOWE Global Brand Store. Also in the end the seller tried to pull the old "ship it to this alternative address" scam (then Ali doesn't see you using their shipping label and youiat the sellers mercy). And once I shipped it back to Aliexpress he tried to dispute my return... I'm not sure on what basis. In the end the return switched from Return/Refund to just Refund. Perhaps the objective of his dispute in the final stage was so they don't return the item back to China, but dispose of it.

Either way. This EKPAOWE battery story shows the kind of surprises one needs to be ready for when buying stuff from China.