Is it normal for LiFePO4 batteries to warm up while charging?

[Hedges]

Yes, correct, but you don't have enough digits difference between 3.157V and 3.155V to get much precision.
Maybe 3.157 is really closer to 3.158, maybe closer to 3.156
Maybe 3.155 is closer to 3.154, maybe 3.156

So that could be (3.156 - 3.156)V / 0.62A = zero ohms,
or (3.157 - 3.154)V = 0.0064 ohms

I'm looking for a figure in the range 0.00025 ohms, typical for LiFePO4 cells.

If you used the following DMM, could get more digits resolution of voltage.

3458A Digital Multimeter, 8.5 Digit

Keysight's 3458A is a fast, flexible, and accurate 8.5-digit multimeter. With a system or on the bench, the 3458A digital multimeter offers low cost of ownership.

www.keysight.com

Otherwise, load your inverter with about 2500W so you draw 100A from battery. Use system measurement of current draw if available, unless you have a clamp ammeter.

Higher current will give more digits resolution on your volt meter.

CM1000A 1000A T-RMS AC/DC Clamp Meter

Amazing deals on this 1000A T-Rms Ac/Dc Clamp Meter at Harbor Freight. Quality tools & low prices.

www.harborfreight.com

Accuracy of a meter is usually quoted as a percentage of reading plus a number of counts in the last digit.
Taking relative measurements like your two voltages helps reduce the error.
We like to have 10 counts better resolution that what we're trying to quote, at least 4 counts.

 

[lawful-nervous]

Yes, correct, but you don't have enough digits difference between 3.157V and 3.155V to get much precision.
Maybe 3.157 is really closer to 3.158, maybe closer to 3.156
Maybe 3.155 is closer to 3.154, maybe 3.156

So that could be (3.156 - 3.156)V / 0.62A = zero ohms,
or (3.157 - 3.154)V = 0.0064 ohms

I'm looking for a figure in the range 0.00025 ohms, typical for LiFePO4 cells.

If you used the following DMM, could get more digits resolution of voltage.

3458A Digital Multimeter, 8.5 Digit

Keysight's 3458A is a fast, flexible, and accurate 8.5-digit multimeter. With a system or on the bench, the 3458A digital multimeter offers low cost of ownership.

www.keysight.com

Otherwise, load your inverter with about 2500W so you draw 100A from battery. Use system measurement of current draw if available, unless you have a clamp ammeter.

Higher current will give more digits resolution on your volt meter.

CM1000A 1000A T-RMS AC/DC Clamp Meter

Amazing deals on this 1000A T-Rms Ac/Dc Clamp Meter at Harbor Freight. Quality tools & low prices.

www.harborfreight.com

Accuracy of a meter is usually quoted as a percentage of reading plus a number of counts in the last digit.
Taking relative measurements like your two voltages helps reduce the error.
We like to have 10 counts better resolution that what we're trying to quote, at least 4 counts.

Makes sense! I'll order a nicer multimeter and try again.

Also, I've been charging the battery that was at 0% to match it to the other battery. It got to about 10% then suddenly dropped back down to 0%, even though there was no load on it. With my rudimentary tools, it seemed to have the highest resistance and different voltage between cells. Any idea why it would behave like that?

Thanks for all the advice! You have really been super helpful. Do you happen to have a buy me a coffee page or something similar?

 

[Hedges]

Instead of trying to read voltage to 10 microvolt resolution, I suggest briefly applying a 100A load. That would also give two more decimal places.
Do you have an inverter to connect? I used a couple electric radiator heaters 600/900/1500W to set the load I wanted.
But you do need a DC current reading. Is that available from BMS?

I bought the HF Ames clamp meter because it has DC clamp ammeter (not just AC), 1000A high enough current for my system (or automotive starter), 10 mA resolution, and at $100 much less than top names with those specs.

I haven't worked with the batteries, so don't know what's going on with the one. Try charging or discharging at modest current, see what voltage is present. That should show if it is in the middle SoC, near charged, or near empty. There may be something needed to reset it. Search the topic, and if no luck maybe post a question with battery make and that issue in title.

Coffee is always good, but I don't have a tip jar. I guess we can all just benefit from swapping experiences. If I ever do go lithium I'll have a head start.
Good luck, keep us updated!

 

[brum]

There may be a single defective cell. Can you measure the cell voltages when it gets to 0% SOC and at 10% SOC? If the BMS provide cell voltages you can also get them from there.

 

[lawful-nervous]

Instead of trying to read voltage to 10 microvolt resolution, I suggest briefly applying a 100A load. That would also give two more decimal places.
Do you have an inverter to connect? I used a couple electric radiator heaters 600/900/1500W to set the load I wanted.
But you do need a DC current reading. Is that available from BMS?

I do have an inverter, but it's only for 24v. The SoC of the batteries is wildly different right now, so I'm nervous about putting them in series to use the inverter.

The BMS does show current draw.

I haven't worked with the batteries, so don't know what's going on with the one. Try charging or discharging at modest current, see what voltage is present. That should show if it is in the middle SoC, near charged, or near empty. There may be something needed to reset it. Search the topic, and if no luck maybe post a question with battery make and that issue in title.

I've been charging it at around 10 amps. It got to 12% and stopped accepting charge altogether. I'll post some screenshots of the BMS data. The cells seems to be pretty out of sync with each other.

Coffee is always good, but I don't have a tip jar. I guess we can all just benefit from swapping experiences. If I ever do go lithium I'll have a head start.
Good luck, keep us updated!

? Thank you!

 

[lawful-nervous]

There may be a single defective cell. Can you measure the cell voltages when it gets to 0% SOC and at 10% SOC? If the BMS provide cell voltages you can also get them from there.

Here the screen shots of the data at 12%. It stopped accepting charge at this point. I'll post more pictures when it discharges to zero.

 

[Hedges]

I don't know why the BMS is unhappy and disconnected charging.
But I guess these are settled voltages. While charging, might have been higher.
What voltage is your charger set to? I would try to apply reduced voltage such that highest cell is within max limit, and see if BMS balances others up. (maybe meaning bleed current off high one while others charge.)

But I would have though 3.389V was much higher than 12% SoC.

"The batteries were recently deeply discharged from being in storage too long. I managed to revive them but am a little concerned about their safety."

Maybe cycle them? Discharge and recharge? Just guessing, I would have thought they'd reset to 100% at some voltage.

 

[lawful-nervous]

I don't know why the BMS is unhappy and disconnected charging.

Well!! I heard back from SOK customer service. They said that the battery is fully charged and the app is displaying the wrong percentage. Which would explain why the BMS would cut out. So I guess I'll have to tinker with that.

 

[Hedges]

OK, now you can try to determine IR with inverter and load.

Your BMS cutting out was just with a 12V charger, correct?
Your 24V system should have its charging voltage set so BMS doesn't disconnect, rather charger stops. And hold "absorption" or whatever the lithium guys call the charge cycle long enough to get some balancing.
You may have to start with a lower voltage while some cells are 100mV apart, then can raise higher once they are better balanced.

 

[brum]

Well!! I heard back from SOK customer service. They said that the battery is fully charged and the app is displaying the wrong percentage. Which would explain why the BMS would cut out. So I guess I'll have to tinker with that.

Yep, looks like that. The SOC should get correct after a full discharge and charge cycle.

The BMS should also recalculate the capacity it shows and this can be used as an indicator for cell healthy. The other indicator has been explained - the battery internal resistance.

The batteries were recently deeply discharged from being in storage too long

How many months, what storage condition, what was the charge before storing them and what do you mean by deeply discharged? I suspect that they may not have been deeply discharged in reality. The BMS is there to protect them and it is hard to damage a cell if the BMS works properly.

 

[windtrader]

We do though have the surge pump batteries strapped / banded with metal bands to keep any abnormal bulging from occurring - but again -- after 2 years - everything still seems fine on all of these systems ...

Did you design in any flex or give to let the cells expand some or did you band and secure 100% no slack? It appears LFP are designed to have a slight amount of expansion and there are some folks using springs to control the pressure in the pack.

The pack is constrained in one dimension only, the wide side. The shorter sides are open so can expand in that direction but still looking for the proper way to keep the cells compacted at the optimal pressure.

 

[lawful-nervous]

OK, now you can try to determine IR with inverter and load.

Well. I charged both the batteries to 100% and hooked them back into the system. And the MultiPlus is somehow causing the batteries to drop to 0volts on the BMS. I have to unhook the batteries and put a small load on them individually to get them to read a normal load again.

I checked with a multimeter at the battery terminals, before and after the bolt on fuse and shunt. I disconnected the MultiPlus and checked before and after the master switch and it was fine. As soon as I hook the MultiPlus back in though they drop back to zero volts.

Also discovered that the MultiPlus holds on to charge for awhile. It sparked when I tried to reconnect it, even though it wasn't connected to any power source. My multimeter said it was holding on to 10v of charge. It didn't drain until a used my led light strip to put a small load on it. Is that normal?

Your BMS cutting out was just with a 12V charger, correct?

I had the charger set to 14.5 volts, which is the recommend Charing voltage on the SOK data sheet.

Your 24V system should have its charging voltage set so BMS doesn't disconnect, rather charger stops.

Yeah, the MultiPlus should cut out before the BMS. But I was just using a little bench power source to try to balance them before hooking them together in series. And since I thought it was only at 12%, I accidentally let it charge to the point that the BMS cut out.

You may have to start with a lower voltage while some cells are 100mV apart, then can raise higher once they are better balanced.

Makes sense!