SOK SK48V100 communication issue RS485C/B

[ernie314]

Hi there,

This post is for users of the SK48V100 battery (latest edition 2023). Hopefully someone can give me some insights.

First my setup:
5x SK48V100 ("the 2023 edition" as supplied to Europe)
1x Victron Energy Cerbo GX
The batteries are connected to each other using the RS485B/C ports.
The #1 battery is connected to the Cerbo via the CAN port.

Issue:
The Cerbo shows that the BMS reports the voltage drops to zero volt every now and then (every few hours and sometimes every few minutes). At the same time other values also show weird values such as the Discharge Current Limit going down by 95A or more. This happens for like one or two seconds, and then everything is normal again.
First fix: After a lot of trial and error I removed two specific batteries from the RS485B/C bus. That (seemed) to stop the problem.
I connected the problem batteries via RS-232 to my laptop to check them, they seem fine.
First conclusion: A hardware problem in two BMS's... ok, I could be unlucky, it happens. SOK told me they are going to send me two new BMS's. So, the issue seemed resolved...

I decided two closely monitor the remaining 3 batteries...

Issue continues:
Now, every few days I still see the Discharge Current Limit going down by 95A for like a few seconds. That is like one complete battery going "off-line".

So, I decided to attach an oscilloscope to the RS485B/C bus to see what is going on. I can see the #1 battery pinging 14 slaves. And every slave that reports back, receives a 2nd query and I can see the slave respond with the requested data. But every now and then it looks like two devices are talking at the same time (the signal is not High or Low, but also halfway of the signal level), and sometimes a slave does not respond. When I add the first two problem batteries, these problems happen continuously. If I only use those last 3 batteries, the problem is a lot less, but still there. I made screenshots of the scope and will attach them to this post. The first screenshot shows a BMS not responding (3 responses on the left, should be 4 like on the right). The 2nd screenshot shows a problem visible as a bright spot. The rest of the screenshots zoom into this spot, to show that it's like two BMS's trying to speak at once?

I changed the cables btw. But that doesn't seem to be the problem, the signal integrity look fine.

Questions:
- Anyone else have this setup and experiencing this?
- Can I be just unlucky and have 3 (or more) faulty BMS's?
- Could it be a firmware problem in this new version of the BMS?

Thoughts?

Cheers,
Ernie

 

[HighTechLab]

I've never had issues like this reported here, so I'll watch this closely.

I think technically the DIP swithches don't really need to be in order, perhaps alternating them, in the event that for somehow one of them is not quite registering the address right?

Not really sure on this one. When you get the 2x new BMSs, perhaps it will give us a chance to dig into it deeper and see if that fully resolved the issue.

 

[ernie314]

Hi, thanks for helping out. I will wait for the new BMS's. This will probably take while (China to Europe shipment...). In the meantime I will keep digging.

I've had already tried changing the DIP switches, that didn't help unfortunately. I agree that the DIP switches only need to be unique. I can see the master #1 battery pinging all the other numbers (2 thru 15) and then sending a query to each battery number that responded to a ping.

Do you know if there is there a way to change the baud rate? Maybe the system will be more tolerant at a lower rate.

 

[HighTechLab]

Do you know if there is there a way to change the baud rate? Maybe the system will be more tolerant at a lower rate.

I don't believe there is.