Urgent - JK BMS went Offline in the middle of the night - SOC Stuck but still discharging ???

[Sky-HHI]

Now that the sun is higher in the sky, your system may be producing more power than before and may be causing one cell to go into over voltage protection( OVP)

 

[silverstone]

Lets start with the simple stuff first. Did you top balance the cells before building the batteries?

No, they were NOT top balanced. And cell #01 on Bat02 was from a previous batch ("spare") since the other was appeared to have leaked electrolyte. As such, on Bat02, IIRC Cell #01 is usually the lowest voltage while Cell #16 is usually the highest voltage.

Can you use the jk bms app to look at your settings?

Not really. Now it seems to be working again. I quickly tried to pull the JK-BMS config over MQTT but it seems it only causes it to store it in the logs ...

Those are my settings though

I find it hard to read your charts but I think one of your bms is turning off when one cell in over voltage, what do you have the OVP set at?

It ALMOST hit OVP (I quickly reduced the charger voltage/current when the battery was reading 3.58x, OVP is set at 3.60V with OVPR set at 3.45V).

What is the delta reading from the bms app?

Very low in the flat part of the curve. I'm a bit skeptical of going there and using the JK BMS app now that it seems to be somewhat working again (for how long :S).

 

[silverstone]

Now that the sun is higher in the sky, your system may be producing more power than before and may be causing one cell to go into over voltage protection( OVP)

No ... Plenty of clouds today unfortunately. There was sun between 08h00 and 09h30. Now I am charging the battery using the charger from the grid instead .

But remember: when the battery went "Offline" it was in the middle of the night and all at normal voltages/current levels.

 

[Sky-HHI]

Did any of the cells go below the under voltage protection you hav set at 2.80v during the night when the battery went offline?The SOC is not very accurate in the jk bms. It is better to use a shunt ( I use a victron smart shunt ) You may need to do a top balance to get the cells all at the same state of charge, you bms is doing what it should as you cell hit low and high SOC. The batteries may not be getting a full charge during the day if one cell hits the OCP. Best to do a top balance and the go from there.

 

[silverstone]

Did any of the cells go below the under voltage protection you hav set at 2.80v during the night when the battery went offline?The SOC is not very accurate in the jk bms. It is better to use a shunt ( I use a victron smart shunt ) You may need to do a top balance to get the cells all at the same state of charge, you bms is doing what it should as you cell hit low and high SOC. The batteries may not be getting a full charge during the day if one cell hits the OCP. Best to do a top balance and the go from there.

No way I hit undervoltage protection. I have a charger in hard parallel permanently connected and the minimum voltage is set at 51V (3.18 VDC/cell) so it should never end up in a UVP scenario (unless the grid fails, but that's another topic - I would be left without any power at all and I would for sure notice it) ...

I had the feeling that after a few cycles the pack automatically balanced out, as during normal operation (say charge up to 55.0 VDC absorbtion/boost), the cell delta V is very small on both batteries. Only when charging above 55.0V Pack level the Battery 02 Cell #01 / #16 seem to do weird stuff. I just assume they are "not great" cells (they are Grade B to begin with, and as I said, they came from different batches), that's why IIRC I set 54.2VDC float and 55.0VDC absorption.

 

[Sky-HHI]

I have three jk bms and have never had any problems with them but I top balanced my cell before putting them in service, as some one suggested moving the bms and see it the problem followes the bms or stays with the battery. Best of luck, maybe someone has a better idea.

 

[silverstone]

I have three jk bms and have never had any problems with them but I top balanced my cell before putting them in service, as some one suggested moving the bms and see it the problem followes the bms or stays with the battery. Best of luck, maybe someone has a better idea.

I would have balanced them if it wasn't an absolute nightmare [in my very cramped battery box]. I see how some people do it (all cells facing the same side, simply put busbars, then after top balance turn 1/2 of the cells 180°), but I wasn't personally confortable with it.

I might do my own set of wiring harness with M6 lugs and maybe Banana 4mm connectors or something like that. I thought of doing it with some wago terminals in the past but maybe a quick connect/disconnect is better to prevent short-circuits when handling stuff...

 

[Sky-HHI]

The more connectors you have, the more problem points you have, that is why most people use their busbars. Look in the Resources of this site for a good way to top balance.

 

[silverstone]

The more connectors you have, the more problem points you have, that is why most people use their busbars. Look in the Resources of this site for a good way to top balance.

True but I don't like having to pull up and down the cells if not necessary (and in cramped places).

 

[timselectric]

The JK BMS has active balancing. No need for top balancing before assembly. (Unless you have purchased garbage cells)
"Offline", just means that it's not communicating.
If the BMS was actually off, there would be no charge/discharge happening.
Your only issue is a communications issue.

 

[silverstone]

The JK BMS has active balancing. No need for top balancing before assembly. (Unless you have purchased garbage cells)
"Offline", just means that it's not communicating.
If the BMS was actually off, there would be no charge/discharge happening.
Your only issue is a communications issue.

That is reassuring to some extent. I wonder why it happened though. The monitoring system was working quite well for a week or so (and home assistant datalogging for a few days), while the battery was "pushed" to cycle more since last week (previously it only did like 60% to 100%, now it's doing 20% to 100%).

 

[timselectric]

That is reassuring to some extent. I wonder why it happened though. The monitoring system was working quite well for a week or so (and home assistant datalogging for a few days), while the battery was "pushed" to cycle more since last week (previously it only did like 60% to 100%, now it's doing 20% to 100%).

I'm not a technology guy. I get lost on the computer stuff. I can't really offer any help on that.

 

[silverstone]

I'm not a technology guy. I get lost on the computer stuff. I can't really offer any help on that.

I'm an electrical engineer and quite good at technology, but as you can see, I'm quite new to this solar field as well.

That datalogging is quite finicky: I was in the garage the other day and just a small bump on the access point cable was enough to bring down the WiFi and lose all battery status. Tried several things, but the only solution was to reboot (if I had more patience probably just restarting the access point server/service would have been enough).

Thanks for your help once again .

 

[timselectric]

I'm an electrical engineer and quite good at technology, but as you can see, I'm quite new to this solar field as well.

That datalogging is quite finicky: I was in the garage the other day and just a small bump on the access point cable was enough to bring down the WiFi and lose all battery status. Tried several things, but the only solution was to reboot (if I had more patience probably just restarting the access point server/service would have been enough).

Thanks for your help once again .

I monitor all of my JK's with Solar Assistant.
Because, I needed simple plug and play.
Technology doesn't like me. Even if I try really hard. lol