MultiPlus II 48/10000/140 - DC charging is unstable

Late to the party, but … ive seen SOC spikes on packs where the inter-battery comms was flakey.
In this case, the slave bms units were not reliable at replying to a request just after another unit replied, for example unit0 was master, and queried (using rs485) unit1 for data, unit1 replied, then unit0 asks unit2, but unit2 doesnt reply. The result is that out of 4 batteries, sometimes the average soc reported to the inverter or SCC was 280/4 = 70%, sometimes it was missing one value so it was 210/4 = 55%.
The final resolution was a firmware update that added a small delay to let the bus go quiet for a moment after a bms had replied, but the temporary solution was to space out the ID numbers, so master was 0, and slaves were 2, 4, 6 instead of 1,2,3.
We found the issue by sniffing the bus and i wrote a python script to display a dashboard of the messages, especially a counter of unacknowledged requests, pm me if you want it.

Woow, you brought it up to the next level. Great job!. :smiley:

FWIW, we also found another feature/bug depending on your pov: usually the master reports the average soc of the pack, so with soc of 80,82,78 and 70, the master would report 310/4=77 or 78 depending on rounding.

One particular firmware on the master would report the lowest soc during charging, and the highest during discharge (or the other way around, my memory is fading) . This is fine with a pack where the std dev of the soc is low, i.e. under 2%, but if your pack is as described above, when solar incoming is approximately equal to demand, so the pack is frequently changing from charging to discharging and back again, the soc appears to oscillate between 70 and 82 - verybl disconcerting for the owner

1 Like

Hi there :slight_smile:

Just want to inform you that newest firmware for MultiPlus II - v555 dramatically improved performance on first site, and on other two completely fixed a problem. On first one the worst seen power factor on genset output is 0.98 lagging and it rarely appears (the vast majority of the time, power factor is 1), but the genset is pretty much stable and there is no need to use weak AC input anymore.

1 Like