JK-B2A8S20P inconsistent voltage measurements

[aurijinboat]

Hello,

I’m new to DIY LiFePo4 but I’ve been lurking for a few months and received 8 EVE LF280K from Docan following your advices. I also ordered a JK-B2A8S20P with Docan (see picture of the label on it) that I‘m having issues with.

The config now is 1P8S to precharge the cells for the first time before top balance, see attached picture. They are then to be used as 2P4S. I follow this guide.

I have access to a multimeter which is said to be 0,09% accurate for DC tensions, ie +/- 0,003V at 3,5V (keysight U1241B), however I don’t have a calibration report on it.

Under no load nor charger, the multimeter reads for all cells 3,326V +/- 0,001V vs 3,348V +/- 0,001V on the BMS. Seems OK.

To precharge the cells, I use a dual power supply, both outputs in parallel, at 28,8V 4 amps each.

Under a 8 amps DC charge at 28,8 V target voltage, the BMS reads cell voltage difference of around 0,09V and starts to active balance at 2 amps, while the multimeter reads 3,335V +/- 0,001V on all the cells. Same voltage difference on the BMS only with top balance off on BMS settings (see attached picture).

I sanded a little bit the contacts and cleaned with acetone with the same results.

Do you have any idea of what could induce the BMS to read unbalanced cells while they are if I trust the multimeter? Is the BMS faulty?

Thanks for your help

 

[aurijinboat]

Hello,

Same problem today : voltmeter sees a cell voltage difference of +/- 0,001V between all cells while JK BMS reads a difference of 0,057V maximum and triggers active balance at 2 amps.
What is weird is that when charging is off BMS sees a normal cell voltage difference of 0,001V again.

Can you please confirm that BMS is faulty?

Thanks

 

[Raikoh]

I would try to swap the cell to see if it is the balance wire that is causing the issue.

Just reading the BMS screen, it looks to me the pack isn't top balanced but i could be wrong.

 

[fdom]

My BMS is within 1-2mV to my dig voltmeter. 20mV difference sounds a lot. Check whether the overall pack voltage is calibrated correctly and that the BMS match your DVM reading of the the pack voltage

 

[RCinFLA]

Check BMS voltage reading vs actual cell readings with near zero charging current. This should negate any connection resistance effects.

Since you do not really know which voltmeter reading is accurate, pay attention to relative voltages.

With no current through cells, if relative voltage readings are different on BMS compared to DVM then BMS is giving incorrect readings.

Since there is a bus bar in series with all but one of the cells for each BMS sense wire pair, any poor connection voltage drop on the bus bar due to I*R voltage drop will add or subtract from BMS cell voltage reading.