CAN communication error

[Kockums85-35]

Hello

I'm experiencing instability in the CAN communication between the Inverter and the battery. I'm not sure whether this has started recently or if the problem has been persisted from the start.
This instability makes the inverter throw an error code, B06, PcsBatCommunicationF

Just before the Inverter gives the error code, the battery discharge is paused for a few minutes before it starts to discharge again, and this way it goes in a loop.

What has been done to try to rectify the problem:

• Replaced the shielded network cable between the Inverter and the Stark CMR and between Stark and battery
• Measured the termination resistance between Can_H and CAN_L at the Inverter-> 139Ω / 68Ω (depends on the position of the DIP switches)
• Measured the termination resistance between Can_H and CAN_L at the battery-> 124Ω / 62Ω (depends on the position of the DIP switches)
• Restarted the system several times during the last days.
• Tried both separate CAN channels and common CAN channel for the Inverter and Battery

However, it looks like there's a few strange characters after the choosen inverter protocol the inverter:

Does anyone know if this is a sign that something is wrong with the communication or if this is just an iPhone issue? It was not like this previously.

Current setup:
Inverter: Afore AF3K-17K-THA 230V
Battery: Leaf 62 kWh
Hardware: Stark CMR Module 

Thanks
Vidar

[dalathegreat]

Seems normal for Afore. We discussed this on the Discord

Afore is not adhering to the standard way of using BYD CAN. It might be better to switch to Afore CAN. But if you do, note that:

Lots of Afore discussion on the Github, but it should be working great.

[Kockums85-35]

Hi Daniel
First of all, I'm so impressed over the work you have put down in this project :)
Ahh, so this is an known issue. Apologise for not searchhing Github..
I guess I need to change "Battery brand" in the Inverter setup according to the Inverter protocol to "Afore Battery" as well?

One quick question regarding termination recistance, with inverter and battery on separate CAN channels, should I measure approximately 60Ω or 120Ω at each cable when everything is povered down? I'm getting a little bit confused here :)

Thanks, and keep up the great work!

[fastbike]

One quick question regarding termination recistance, with inverter and battery on separate CAN channels, should I measure approximately 60Ω or 120Ω at each cable when everything is povered down? I'm getting a little bit confused here :)

The CAN bus requires 120Ω at each end to dampen signal reflections, because these are in parallel the measured resistance will be 60Ω

[Kockums85-35]

Thanks, I think I got it 😊
The goal is to achieve 60Ω in both the Inverter and the battery when connected to separate CAN channels in the Stark CMR👍

[Kockums85-35]

Hi guys
I'm experiencing a strange issue with my setup:
Inverter: Afore AF3K-17K-THA 230V
Battery: Leaf 62 kWh
Hardware: Stark CMR Module

I was adviced to change the Inverter protocol from BYD to Afore to fix the problem I had with random inverter restart and error messages, see the start of this thread.
I changed the Inverter protocol to "Afore battery over CAN" and changed the battery type to Afore in the battery parameter setup in the inverter.
So far so good, but, now that the solar panels finally has started to generate some power, I noticed that the charge effect is somehow limited to approx. 1 kW. All additional power gets exported to grid.

Switched back to BYD protocol and BYD battery and the battery was able to consume all the excess production. Being March, the sun is quite low in southern Norway but I reached 4.5 kW charge at max today.

My question is, have anyone with similar setup experienced similar issues with the Afore battery settings and can point me in the right direction?
I have been through all the inverter settings many times without being able to detect why the charge current is restricted.

Vidar

[Kockums85-35]

I'm struggling with the time difference to China, but the initial response from Afore was that:
"Battery charging limits are related to cell voltage.
When cell voltage exceeds 3.5V, the inverter will limit the charging current."

I believe that the Leaf cells operate between 3.5V - 4.1V

My Supplier in Norway shared his solar plant setup with me, and I can see that his Afore batteries runs at at a lover cell voltage than my Leaf battery, see attached pictures.

Afore battery:

Leaf battery:

Whether this is something Afore can include in the setup as a configurable parameter or not hasn't been answerd yet.

Is it possible to scale the cell voltage in the BE SW sent in the CAN messages to the inverter? Or will this cause other unwanted problems?

[dalathegreat]

It is very possible that this is the solution! Try this fix! We use the same for Ferroamp, which also requires LFP to function, so we fake-rescale the voltages. #2094

[Kockums85-35]

Interesting :)
I'm not sure if I followed all the steps to the point, but I somehow managed to update the Stark CMR to version 10.2.dev 😊

Now the cell voltage has dropped with approx. 0.5V:

The weather forcast for tomorrow looks promising, so it will be interesting to see if this fix the issue 😊

Many thanks 😊

[Kockums85-35]

It looks like we have found the culprit here 😊

While I'm having some sort of dialouge with the right kind of technical support at Afore, should we address the problem we have with the BYD protocol as well?
Dala mentioned this earlier inthis thread:

"Afore is not adhering to the standard way of using BYD CAN."

Is it possible to elaborate on this, or point at any specific error in Afores BYD CAN code?

[dalathegreat]

Wohoo 🥳 I will merge this to main, will be included in next release (10.2.0)

If you have their tech team on the line, you can tell them;
The ASCII string for Inverter brand in CAN message 0x151 is not sent correctly. It is being reported as "3???" instead of something more sane like "AFORE". Other manufacturers like GoodWe are handling BYD CAN properly (sending "GOODWE" in message 0x151)

case 0x151:  //Message originating from BYD HVS compatible inverter. Reply with CAN identifier!
  if (rx_frame.data.u8[0] & 0x01) {  //Battery requests identification
    send_initial_data();
  } else {  // We can identify what inverter type we are connected to
    for (uint8_t i = 0; i < 7; i++) {
        datalayer.system.info.inverter_brand[i] = rx_frame.data.u8[i + 1];
    }
    datalayer.system.info.inverter_brand[7] = '\0';
  }
  break;

[Kockums85-35]

Message sent, will keep you posted.

Thanks

[Kockums85-35]

I didn't get any feedback from Afore, but I switched over to BYD Inverter protocol and haven't seen any restart loop of the inverter 😊
Could you check the CAN messages whenever you have the time? Let me know if you need any CAN log from me.

[dalathegreat]

All is good from my side now! I added filtering to avoid corruption from Afore sending garbage when in BYD mode.

So now we fixed:

• Afore protocol can charge faster (since we emulate LFP voltage range)
• BYD Afore setting filters out invalid chars to avoid corruption.

I think we can close this discussion for now, very nice development!