Difference in discharge amperage

[Frank P.]

I built 2 LiFePo batteries (8S, 24 V, 176 Ah) with a 100 A BMS each (the one sold by OverKill solar). They are connected in parallel. They are hooked up to a 3000 W interter. When I put a load of 1500 W on the inverter I remarked on the BMS xiaoxiang app that one battery was delivering 38.2 A and the other 26.3 A so that after about an hour, their was 132 Ah left on the first battery and 146 Ah on the second battery. I would like to understant why each battery is not delivering the same current.

 

[Bob B]

The current can be calibrated .... I had to calibrate mine to make it accurate.

You need to have a way to accurately measure and control the current in and out in order to do that.

 

[Substrate]

Are you connected to your inverter in a ladder-type configuration, that is tapping off just one of the two parallel batteries and not connecting the negative to 1st one, and the positive to the 2nd one to make sure you share the load as evenly as possible?

We're assuming that your parallel connections are tight and proper too..

 

[Frank P.]

Thak you for your coments. Yes, the negative is connected to one battery and the positive to the other and the connections are tight and proper. I also know that the measures are accurate because I double checked with an ampmeter that gave me the same readings as the app. Concerning calibrating I am not sure to understand what you mean.

 

[Frank P.]

The current can be calibrated .... I had to calibrate mine to make it accurate.

You need to have a way to accurately measure and control the current in and out in order to do that.

Concerning calibrating I am not sure to understand what you mean.

 

[Frank P.]

Are you connected to your inverter in a ladder-type configuration, that is tapping off just one of the two parallel batteries and not connecting the negative to 1st one, and the positive to the 2nd one to make sure you share the load as evenly as possible?

We're assuming that your parallel connections are tight and proper too..

Thank you for your coments. Yes, the negative is connected to one battery and the positive to the other and the connections are tight and proper. I also know that the measures are accurate because I double checked with an ampmeter that gave me the same readings as the app. Maybe another precision: the voltage of the two batteries always remained the same for both batteries. When I started at the same voltage and where both at 26.5 V when I disconnected the load.

 

[Bob B]

Concerning calibrating I am not sure to understand what you mean.

Do you have the Android app or the IOS?

I can tell you how to calibrate with the IOS version.... but don't know the Android version.

Calibration is simply making sure the current readings are correct.

 

[Hedges]

I built 2 LiFePo batteries (8S, 24 V, 176 Ah) with a 100 A BMS each (the one sold by OverKill solar). They are connected in parallel. They are hooked up to a 3000 W interter. When I put a load of 1500 W on the inverter I remarked on the BMS xiaoxiang app that one battery was delivering 38.2 A and the other 26.3 A so that after about an hour, their was 132 Ah left on the first battery and 146 Ah on the second battery. I would like to understant why each battery is not delivering the same current.

That would suggest different resistance somewhere, whether bars, cells, contact to a terminal.
With a heavy current being drawn, use a DMM to measure each segment, alternating between the two parallel strings. Across each busbar, across each cell, across the cables paralleling them at the ends.
You might find one bad contact, in which case it is likely hot too.

I had an imbalance issue with paralleled inverters, turned out to be one brand of breaker, where two of them weren't well matched.
Replacing that, went from 3:1 current imbalance to closely matched.

 

[Gazoo]

Do you have the Android app or the IOS?

I can tell you how to calibrate with the IOS version.... but don't know the Android version.

Calibration is simply making sure the current readings are correct.

If one needs to calibrate and all they have is Android, then the BMS must be connected to a computer to calibrate it. There is no way to calibrate the BMS using the Android app as far as I know.

 

[Frank P.]

If one needs to calibrate and all they have is Android, then the BMS must be connected to a computer to calibrate it. There is no way to calibrate the BMS using the Android app as far as I know.

I do have the Android app. But I double checked the amp readings with an ampmeter and it gave the same readings as the app. So I am assuming that the app is well calibrated.

 

[Frank P.]

That would suggest different resistance somewhere, whether bars, cells, contact to a terminal.
With a heavy current being drawn, use a DMM to measure each segment, alternating between the two parallel strings. Across each busbar, across each cell, across the cables paralleling them at the ends.
You might find one bad contact, in which case it is likely hot too.

I had an imbalance issue with paralleled inverters, turned out to be one brand of breaker, where two of them weren't well matched.
Replacing that, went from 3:1 current imbalance to closely matched.

Thank you for this suggestion. I will surely do it but I would need more explanation on how to do it because I am not familiar to measuring resistance other than to use the multimeter to see if current flows.

 

[Hedges]

Thank you for this suggestion. I will surely do it but I would need more explanation on how to do it because I am not familiar to measuring resistance other than to use the multimeter to see if current flows.

The resistance scale of a DMM is good for larger values, like 1 ohm or 1k ohm.
The values in a battery pack are more like 0.001 ohm, which can't be read that way (and you can't use the ohms scale on a circuit with power applied.)

Resistance measurement is simply forcing a current and measuring the resultant votage.
Power a load with your battery. For instance, plug a 1500W heater into an inverter. That will draw about 130A at 12V, or half that at 24V. Divided between two batteries in parallel, so about 32A each except they are imbalanced.
Then use your DMM to measure voltage. Each cell might measure 3.2V; if one measures noticeably less than another, it could have poor contact.
Maybe you have a 1' long wire joining negatives of the two batteries, and another joining positives. Measure voltage across each. If same length, should show same voltage (like maybe 0.05V?)
Measure along each busbar, maybe from stud/bolt of one cell to stud/bolt of next. Something like 0.025V ~= 32A x 0.0008 ohms or so. If one busbar is different from the rest, it has poor contact.

Just compare all similar elements, look for outliers in terms of voltage drop.

Are these bolted aluminum cell terminals? How did you treat the surface? Some people saw cell voltages imbalanced (what does your BMS say?) and had a hot busbar. But scrubbing the terminals, applying corrosion inhibitor, torquing carefully, they got there cell voltages similar and nothing hot anymore.

 

[Gazoo]

I do have the Android app. But I double checked the amp readings with an ampmeter and it gave the same readings as the app. So I am assuming that the app is well calibrated.

I have done the same using another cheap DROK coulomb counter and they are within a few amps of each other. I have also checked the voltages with a DVM and they are all within a few mv's of each other. Close enough for me...

 

[Frank P.]

The resistance scale of a DMM is good for larger values, like 1 ohm or 1k ohm.
The values in a battery pack are more like 0.001 ohm, which can't be read that way (and you can't use the ohms scale on a circuit with power applied.)

Resistance measurement is simply forcing a current and measuring the resultant votage.
Power a load with your battery. For instance, plug a 1500W heater into an inverter. That will draw about 130A at 12V, or half that at 24V. Divided between two batteries in parallel, so about 32A each except they are imbalanced.
Then use your DMM to measure voltage. Each cell might measure 3.2V; if one measures noticeably less than another, it could have poor contact.
Maybe you have a 1' long wire joining negatives of the two batteries, and another joining positives. Measure voltage across each. If same length, should show same voltage (like maybe 0.05V?)
Measure along each busbar, maybe from stud/bolt of one cell to stud/bolt of next. Something like 0.025V ~= 32A x 0.0008 ohms or so. If one busbar is different from the rest, it has poor contact.

Just compare all similar elements, look for outliers in terms of voltage drop.

Are these bolted aluminum cell terminals? How did you treat the surface? Some people saw cell voltages imbalanced (what does your BMS say?) and had a hot busbar. But scrubbing the terminals, applying corrosion inhibitor, torquing carefully, they got there cell voltages similar and nothing hot anymore.

Thank you for your explainings and suggestions. I checked everithing and I found no difference in voltage between different segments (wires and bus bars) except on the two wires connecting the positives and negatives of the two batteries in parallel. But the same wires also delivered a different amperage (the same that the BMS app gave me). I changed the wire giving me the lower voltage but it did not chage anything. I also used the temperature sensor on my DMM to check each wire and each bus bar and did not find any place where the temperature was higher. These are brand new batteries that I just assembled so all the contacts are clean. Is their anything else I should loof for?

 

[Gazoo]

The BMS will not report the ah's correctly until each battery is fully cycled. I think the solution to your problem might be as simple as that unless I missed something.

 

[bgflyguy]

I would guess a bad connection as well. Whenever you are pulling the load touch the bus bars and terminals. If it's making that much of a difference I bet it is heating up.

 

[Frank P.]

Thank you for your explainings and suggestions. I checked everithing and I found no difference in voltage between different segments (wires and bus bars) except on the two wires connecting the positives and negatives of the two batteries in parallel. But the same wires also delivered a different amperage (the same that the BMS app gave me). I changed the wire giving me the lower voltage but it did not chage anything. I also used the temperature sensor on my DMM to check each wire and each bus bar and did not find any place where the temperature was higher. These are brand new batteries that I just assembled so all the contacts are clean. Is their anything else I should loof for?

I tested something else. While puting a load of 1500W on the battery pack, I measured the voltages between the wire connected to the C- and the wire connected B- of the BMS (C- is connected to the load and B- connected to the negative pole of the group of cells). This voltage read ~60V on one of the batteries and ~40V on the other. This proportion between these readings is the same as the one between the difference in amperage delivered by the two batteries. That being said, I don't know what deduction I should make of this observation if any.

 

[Hedges]

60V seems plausible for a well-charged battery with light load.
40V seems excessively low. In my lead-acid world, 42V might be low-voltage disconnect.

I think the model is that a cell is a perfect voltage source (with voltage dependent on that non-linear SoC curve) in series with its internal resistance.
And then wires and busbars are resistors.

I think the voltage drop compared to no-load is what should be proportional to current draw (not remaining voltage proportional.)
Seeing 40V is a 20V drop compared to the other battery at 60V. What was the voltage of each under no-load?
20V drop out of 60V would suggest that 1/3 of total power (coming from that battery) is heating it up, and 2/3 is delivered to load.

Is the battery which dropped to 40V the one delivering most current? or least?