CANbus Data Error SOC - Pylontech CAN

[ircoopr]

Hello! First up thanks for the fantastic fork of this really useful tool, it's allowed near-perfect operation of my DIY battery system connected to my SunSynk 5kW, amazing.

I've been having a few errors recently that I'd love to discuss, primarily around SOC reporting from JK-BMS to the SunSynk via CAN. Occasionally I have the system report 80-100% SOC (as expected), then suddenly just drop to ~25%. The example today is shown:

That data is from the SunSynk Modbus, but it's also reflected on the SunSynk API and on the Inverter screen.

Looking at the logs from ESPHOME I have the following.
[17:05:49][D][sensor:127]: 'jk-bms-can capacity remaining': Sending state 85.00000 % with 0 decimals of accuracy
---some lines that are not relevant----
[17:05:52][I][main:298]: send can id: 0x355 hex: 55 0 63 0
[17:05:52][D][canbus:033]: send extended id=0x355 rtr=FALSE size=4

55 = 85% which is correct, and what is reported by the JK-BMS RS485 connection.

Before I start snooping the CANbus is this a known issue at all? I'm usually able to fix this by power cycling my ESP32. It feels like it could be a CAN buffer/peripheral issue? It looks to me like different data is being sent in to the actual frame, based on the value it drops to and the fact it moves about a bit (20-30) I suspect it's a piece of temperature data....

Any support is appreciated. I need to dig out some equipment to be able to log the raw CANbus which I will do to confirm what's going out on the actual bus to ensure it's a problem on the ESP end, not the SunSynk end.

[Uksa007]

Hi,

It not something that I have seen, I suspect you have either:

• Signal integrity issues from the ESP32 to the inverter.
  • Consider using my Interface kit, it removes all the hacked together modules and is bulletproof.
• Inverter doing something weird and using its own SOC.

My SOC from the inverter:

Please consider supporting my work so I can continue to develop new feature and provide help and assistance.
https://www.patreon.com/Uksa007Codedevelopment

[ircoopr]

CAN data looking OK on the bus.

Getting 0x355 on to the bus every 1second (though there are some 1.5second gaps):

All packets seem valid. And inverter seems to be responding roughly every second

Reset ESP32 to get back to working state. Interesting to note that packets are more consistently 1second apart:

Almost feels like the 2 different SW screens on the inverter have different packet debounce allowances? This is so far the only difference I have noticed between the failed state (that appears to happen eventually) and the 'good' working state. I should note at this point I have modified the code to add in an additional bus measurement (SunSynk RS485), but I believe I had the same issue as this before this. Unfortunately it's not super easy to turn the error off / on, for example, since I last posted this evening is the first time I've seen the error occur again.

[Uksa007]

Looks to me like you are running old code there is a new version with better compatibility with Deye, your inverters seems very similar to Deye.

Try running the latest version:
#2

[ircoopr]

I've definitely built from much more recent code than that (for example I have the "V1.11 Major update: Charging logic now has absorption, sliders for Charge Volts and Amps." available, and I'm pretty sure my build is from the last month or so

Appreciate the suggestion though :)

[Uksa007]

You need V1.13.3 14/5/23 Improve compatibility with Deye and other inverters

I can tell you are not running this version, as your charging voltage is 0.

[ircoopr]

Thanks, I'll try a rebuild, apologies, I thought I had a more recent build, seems I am running "# V1.13 fix array too small".

[Uksa007]

Closed due to inactivity.

Please consider supporting my work so I can continue to develop new feature and provide help and assistance.
https://www.patreon.com/Uksa007Codedevelopment