7 hours ago, Sidewinder said:

@splunker

Great. Congrats.. you've done great so far!!!.

Just off the cuff, your 2 pack delta voltage difference looks excellent. couple of mV. The [problem is one needs to monitor it during a full daily cycle of charge and discharge. I've noticed on my 2 packs great differences, but it always recovers to a few <10 mV. 

Regarding the cycle count: Mine show like 460 after nearly 4 years of sparse use (SOC% down to 60% mostly per day). So I deduce that the cycles are measured from full (100% minus the lowest point e.g. 60%) = 40% = 0.4 of a cycle, and it add all these dips together for the cycle count.

I have heard of some BMS that regards that as 1 cycle, so it is over estimating the cycle count by a large percentage. So if your charge/discharge cycle is very high per day, then the cycle count will also be high. Maybe someone can explain it more accurately.

Thanks for the link re BMS software, I'll get a copy and keep it in my arsenal!

You may want to take a look at "System Config" tab. That will show how the battery is setup. Quite a number of parameters, but you will see things like OVP (Over Voltage protection,  & e.g. when the balancer (if BMS has the functionality) kicks in. (+-3.65V, if i remember correctly). In there you should also see what Protocol the BMS support/talks. Pylontech seems to be the most prevalent.

So plugging the Serial cable into your HA server won't show anything (yet). the only way to get the info into HA (for me, there may be other/newer/easier methods now) is via Mqtt. Once the correct integration is added, it all comes to life.... let us know if you succeed.

lol thanks @Sidewinder

ja absolutely, monitoring change over time will be important to undersand whats happening. As of 1 day, only 1 battery has iterated a charge cycle... cant help wondering...I'll watch for a while and see what happens.

460 over 4 years is kind of amazing, your batteries will last forever!

I poked around, and maybe the parameter tab shows what you are talking about. You're right there are a ton of battery configs in there, need to be super careful not to accidentally change something lol. yip the balancing stuff is in here too, 3.45V on this one.

WRT the protocol, I'm not very clear on the RS485 vs RS232 distinctions, but RS485 is set to Darfon. Pylontech is in the dropdown but i'm way too nervous to make any changes :). CAN is on Pylontech though, and as I understand, that's the protocol that the inverter wants...?

 

In terms of HA, theres no way to get a cable between the batteries and the server unfortunately. Way I'm seeing it at the moment, in order of increasing effort and decreasing cost:

1. solar-assistant all-in-1 talking to HA over wifi R3k - looks easy
2. solar assistant+Pi Zero ~R1500 - little more effort
3. RS232 serial -> ESP board -. Wifi to HA via MQTT - Lots of effort, may be impossible, but no cost (unless I blow up the batteries)

Will need to do some reading around the topic, i dont know much about serial protocols at all. Hopefully some clever person has already done this...

Anyway, thanks again

On 2024/07/23 at 9:07 AM, Sidewinder said:

@splunker

Great. Congrats.. you've done great so far!!!.

Just off the cuff, your 2 pack delta voltage difference looks excellent. couple of mV. The [problem is one needs to monitor it during a full daily cycle of charge and discharge. I've noticed on my 2 packs great differences, but it always recovers to a few <10 mV. 

Regarding the cycle count: Mine show like 460 after nearly 4 years of sparse use (SOC% down to 60% mostly per day). So I deduce that the cycles are measured from full (100% minus the lowest point e.g. 60%) = 40% = 0.4 of a cycle, and it add all these dips together for the cycle count.

I have heard of some BMS that regards that as 1 cycle, so it is over estimating the cycle count by a large percentage. So if your charge/discharge cycle is very high per day, then the cycle count will also be high. Maybe someone can explain it more accurately.

Thanks for the link re BMS software, I'll get a copy and keep it in my arsenal!

You may want to take a look at "System Config" tab. That will show how the battery is setup. Quite a number of parameters, but you will see things like OVP (Over Voltage protection,  & e.g. when the balancer (if BMS has the functionality) kicks in. (+-3.65V, if i remember correctly). In there you should also see what Protocol the BMS support/talks. Pylontech seems to be the most prevalent.

So plugging the Serial cable into your HA server won't show anything (yet). the only way to get the info into HA (for me, there may be other/newer/easier methods now) is via Mqtt. Once the correct integration is added, it all comes to life.... let us know if you succeed.

Thanks @Sidewinder

Unfortunately I can't plug the battery directly into the ha server as it's in a different part of the house.. will have to come up with a solution.

I noticed something strange today again I think, maybe you can comment. For the first time in a few weeks, and the first time since the battery fw update, they actually charged to full. They got to 99% and stopped increasing, but the batteries appear to still be drawing current. Previously, they would stop charging, and the house would just run off solar. It's a little concerning that they seem to keep drawing but never increase. Typically the orange solar power curve drops.off to match the house requirements once the batteries stop charging.

Thanks in advance