JK Inverter BMS & Victron

[robot_zombie]

I recently setup a victron system (MP2 5kva, cerbo, 150/45 and 250/100 MPPTs) with a DIY battery (16x EVE280V3) and new JK inverter 200A BMS (no smartshunt, JK calling the shots over CAN...). Anyway, its all working pretty well aside from a few niggles... (the JK is running the latest firmware as of a month ago or so, 15.something I think), wondering if anyone else has ran into these issues with the new inverter BMS...

1) the battery capacity / consumed / remaining AH doesn't seem to be sent to the cerbo by the JK, is this normal or something is up with my system? (NaN displayed as time remaining, no capacity info in the cerbo)

2) my RCV is 3.45V, 100% soc 3.449 iirc, but sometimes if the soc is >85% or so and a big charge comes in (4+kw) from a brief burst of sun some cells momentarily jump up to 3.45V, the SOC jumps from 85% eg. straight to 100%. Bit of an issue with their algorithm, I might raise this with JK and see what they say (why doesnt this have a timer or average applied hmmm or use a tail current once ABS pack voltage is reached). I never had any issue with earlier JK BMS soc so this 'feature' is kinda annoying. I expect the correction for this issue barring a fix from JK is a smartshunt and charge control by the cerbo.

3) RFV timer... this system is setup at my parents (off-grid, 3 hours away), it seems the RFV timer is on 1 hour, although I swear I set it to 6 or 8 hrs with 1 hour ABS@3.45. I noticed this was mentioned a few times in the JK firmware changelogs as not being saved correctly, i'll be back onsite next weekend to check this but can anyone confirm the timer is working correctly with their system and which JK firmware they are running?

cheers

 

[robot_zombie]

anyone? I was thinking I might up the charge voltage to 3.46 or 3.465, it might stop the SOC jumping issue.

 

[robot_zombie]

to update, on the off chance this is useful for anyone else - adjusting RCV from 3.45 to 3.46 helped the soc jumping to 100% issue but still present, now it generally only happens >90% soc and less often.
JK need to change the trigger from a *single cell* momentarily hitting this voltage to the pack voltage hitting this value x cell count (and that should ideally be time or tail current based as well), which is how the absorption timer seems to trigger (as none of the soc jumping events have resulted in the abs timer triggering earlier... hmm!!), why it would be different, I have NFI...
and yeah the float timer works, I apparently just forgot to save it? hah
still wondering about the JK sending capacity and Ah consumed to the victron.

not sure why this issue (SOC jump) with the new JK hasnt been discussed more widely, I thought maybe its an issue with my setup but the battery is otherwise behaving well, just occasional sunbursts pumping 4-5kw in momentarily seem to be able to spike a cell or two triggering the 100% early hmmm. (new eve 280 V3's)

 

[linkcat]

You should try to make this comment to Andy from Offgrid garage, is the only way the people of JK take care of user reports!

 

[robot_zombie]

You should try to make this comment to Andy from Offgrid garage, is the only way the people of JK take care of user reports!

heh I had the same thought, I left a comment on his JK inverter BMS settings video but I guess it was too old at the time, no response as of yet!

 

[bluenomads]

I've had no problem with the cerbo bms comms - I have the same bms. I get full data, and it's passed through to the mppt.
Except today, when I lost comms and can't get it back. Be careful with updating the cerbo version, I'm not sure it was responsible but...

Agree with your 2. I had a single cell that was always vastly behind, and one or two out the front. So suddenly it jumped to 100% when one was 3.2v, most were 3.4, and one hit 3.55. Oh, and this is *after* top balancing! So I too would like a little more control than simple single cell cutoff. 2a balance isn't enough when the sun comes out.

I've also had a bad day with mine - refuses to action choices about turning charge, discharge and balance on or off, and says 2 of the cell cables can't have their resistance calculated, even though the cell voltages seem fine. I was happy, but now rethinking whether the bms is solid enough for real use.

 

[robot_zombie]

I've had no problem with the cerbo bms comms - I have the same bms. I get full data, and it's passed through to the mppt.
Except today, when I lost comms and can't get it back. Be careful with updating the cerbo version, I'm not sure it was responsible but...

Agree with your 2. I had a single cell that was always vastly behind, and one or two out the front. So suddenly it jumped to 100% when one was 3.2v, most were 3.4, and one hit 3.55. Oh, and this is *after* top balancing! So I too would like a little more control than simple single cell cutoff. 2a balance isn't enough when the sun comes out.

I've also had a bad day with mine - refuses to action choices about turning charge, discharge and balance on or off, and says 2 of the cell cables can't have their resistance calculated, even though the cell voltages seem fine. I was happy, but now rethinking whether the bms is solid enough for real use.

interesting re. the cerbo. The comms always worked for me just missing the consumed/ and capacity Ah, the system is running without a smart shunt, everything is controlled by the JK. Cerbo running v3.20. That RCV 100% setting tho, its real annoying, it cant even be disabled or even set higher than the charge voltage, so the issue will likely be annoying for a load of people.

 

[Patatman]

I was about to create my own topic about the SoC jumping, but looks like you have the same issue
My setup looks like this:

The different cell voltages:

If I look into the different cells, at 12:03 it hits 100% SoC, and this happens to the cells:
It then starts to balance, and keeps filling the battery until it actually reaches 100% after +- 1 hour.

And if you look at multiple days, this always seems to happen:


Looking at these graphs, I'm wondering if JK just reports 100% when floating. And after 1 hour fills up the remaining 10% SoC.
Curious to see if changing the RCV value adjusts this behaviour.

 

[robot_zombie]

yeah, similar to what I see. With the victron once float kicks in after the ABS time, it generally sits around 99-98% for me, and the system draws a little from the battery until the voltage drops to the RCF value.
The soc jumping is annoying but so far hasn't been a problem. I could see a case in winter where the battery is at 90% say, a burst of sun resets the SOC to 100%, then it pours rain for the rest of the day introducing a 10% error hmmmm combined with a few more days of overcast weather and the JK SOC which seems to (for me at least) accumulate a little positive error (ie soc>real soc) each day and the error could become a more serious issue than the JK soc drift alone! hopefully JK release some update that fixes this eg. voltage must be xx.xx for xx time or whatever!

 

[Patatman]

I'm still digging a bit deeper, but it looks like the JK bms is reporting some weird stuff.
At 11:35 it says the SOC is 100%.
After that it drops in voltage, but keep reporting 100% soc.

If I look at just 11:15 to 11:45.


I'm not sure if this is normal behaviour, since this is my first battery I've build.
Do you see similar voltage charts?

 

[kommando]

JK are constantly changing the firmware plus hardware but give no changelog, I get the jump to 100% but then it will drop to 99% within a few mins if a cloud comes over. I am on the very latest FW 15.17 for ver 15.1 board.

 

[robot_zombie]

I'm still digging a bit deeper, but it looks like the JK bms is reporting some weird stuff.
At 11:35 it says the SOC is 100%.
After that it drops in voltage, but keep reporting 100% soc.
View attachment 211616
If I look at just 11:15 to 11:45.
View attachment 211617

I'm not sure if this is normal behaviour, since this is my first battery I've build.
Do you see similar voltage charts?

yep, I see exactly the same thing. As far as I can tell, the 100% reset happens when any cell hits RCV, but the absorption timer only kicks in when the pack voltage its maintained at or above RCV x cell count. At least it seems this way, because I have not seen abnormally short termination of ABS time, and sometimes it may hit 100% an hour before the pack is actually steady at 55.3V in my case. So yeah, sometimes it might hit 100% then the sun goes away and cells drop to 3.39-3.4V like you see here. Annoying, why not just use the same logic to trigger ABS time and 100%? anyway!

IME the JK's were already 'SOC happy' in that after awhile without a full charge, the reported soc was clearly too high... rarely a problem in my case as the inverter charging from solar doesnt stop at 100%, it waits until the BMS ->(esp32)->inverter coms tell it to stop (ie. end of ABS), so sometimes in winter it might be a kwhr out and sits on 100% for awhile. but now this new 'feature' makes the BMS even more SOC happy haha :S