aquarat

54.2 means 54.2/15 = 3.613 VPC average. This isn't great for the cells, but not bad either. The small charges and discharges are inevitable, and would not be hurting, in my opinion.
However, it would be better for the cells to stay around 3.5 VPC once fully charged, or around 52.5 V total. It would be great if one of you Pylontech guys figured out the commands to send to the Pylontech BMS (via the console port, an RS-232 port different to the main RS-485 data port) to change the float voltage to that 52.5 V. I think that then the inverter-cable-to-BMS system would work well, with good cell longevity, and few if any panic disconnects from the BMS due to the Axpert overshoots.

Hi there.
I came across this problem last week when trying to install new US3000C batteries.
I know you asked this a little while ago, but hopefully this helps you...
I asked around and tried to get a definitive answer to this. No one could help.
I kept getting given the incorrect cable pin outs.
Eventually I used a bit of logic with the manual and what I could find on the net, and figured it out.
I made up my own cable and it worked perfectly.
here's the pinouts for the new US3000C model.
Battery Pin 8 to Inverter Pin 3 ; Battery Pin 7 to Inverter Pin 5
Good luck
Let me know if this helps you?

Hi,
Here is a quick zip file of the program I use to poll the batteries over RS485 and push the read numbers out to OpenEnergyMonitor's MQTT server.
Maybe it will be helpful.
I'm not the original author - I just modified to publish over MQTT using a JSON structure, and to scan for batteries.
 
 
pylon485-elbow.zip

I haven't been to this site in a while haha. My Must inverter burnt out/failed after a while and I replaced it with a Voltronic Axpert King unit. I think the Must unit failed due to me turning on a dodgy setting (setting 09 I think? It's "auto restart on overload" I think). I think the failure is related to the TVS diodes. I've still got the unit, need to get it repaired.
Both the Must unit I had and the Axpert that replaced it have some oddities to them. The King offers a lot more control via it's comms system than the Must unit, so I have a NodeRed automation flow that manages the state of the unit based on weather data. It's pretty cool. The King protocol is fairly well documented.
It is possible to forcefully manage the Must units in a similar way by having a computer-controlled upstream contactor/big relay, so that one can shut off power to the inverter from the utility. I have this for my King unit, even though it doesn't really need it.
My house runs on Home Assistant via a message broker (MQTT). I had and have a bridge between the serial interface on the inverters and my home MQTT server via a tiny Raspberry Pi  Zero SBC, which allows me to log telemetry and also control the inverter. All the software is DIY. The graph above is from Home Assistant based on data collected through the bridge I created (the data itself is in a relational database). Someone else has since extended this software to include support for the Axpert Voltronic King units.
The Must units have their own serial protocol and the protocol is not documented, however, the software the inverter ships with is very easily decompiled. This is what I did, I decompiled the shipped monitoring app, stuck some boilerplate code on it so it could be run from a command line and would dump all available telemetry to the output terminal in JSON format. I take the output and ship it off to the message broker where Home Assistant picks it up, parses it and stores it in the database.
Because the code is the product of decompilation, putting it on something like Github would probably constitute a copyright infringement - that's ironic though given that the product it supports is itself a copyright infringement. I could probably upload it regardless and if they get unhappy they'll just ask it to be removed.

I experimented a bit more today: it seems to start disconnecting/connecting the PV array when the float voltage is set to a value lower than 57.1 VDC. I'm guessing that it can't bring the voltage from the array close enough to the normal float voltage, so it just drops the array (in simplistic terms).
I originally based the wiring of the panels on wanting reduced current and the Must Power clone unit, which indicated that up to 130 VDC was acceptable (the panels don't seem to get above 121 VDC open circuit) - whoops. I'm really not in the mood to rewire them haha 😂.
Thanks for the insights/verification of my unit's config^