EG4-LL v1 BMS Communication “Float Stoped”

[ricardocello]

Background
I’ve decided to go back to open loop control on my 3 EG4-LL v1 devices, after 6 months of doing it closed loop with a Victron Cerbo GX via CAN.
The closed loop control worked reasonably well, but now my batteries are out of balance, and I have too many unanswered questions about the BMS.

I plan to put a Raspberry Pi4 in my battery rack, and monitor the three modules via RS-485 so I can at least see what the cells are doing.
I’ve found the modbus protocol description for this BMS online, but could also just look at BMS_Tools for the protocol used.

I’ve reviewed all of the threads here on absorption and float voltages, being above 3.4V to balance, etc.
There is lots of excellent advice on this site.

Observation
When I was in closed loop, if I pushed the voltage too high, the BMS would signal an Overvoltage Alarm to the Cerbo, which was really annoying.
Especially since none of the cell voltages were overloaded (watching BMS_Tools to verify). Others have discussed this at length.

Question
Looking at BMS_Tools, I don’t see any overvoltage alarms, but I do see “Float Stoped” is set.
Yes, it is presumably a misspelling. I think the Cerbo interprets any bits set in that register as Overvoltage,

Can anyone tell me what “Float Stoped” means, and what triggers it?
It appears the BMS is disabling charging, which is making it really difficult to balance one low cell.
Thanks

 

[sunshine_eggo]

Background
I’ve decided to go back to open loop control on my 3 EG4-LL v1 devices, after 6 months of doing it closed loop with a Victron Cerbo GX via CAN.
The closed loop control worked reasonably well, but now my batteries are out of balance, and I have too many unanswered questions about the BMS.

Properly configured, they should not go out of balanced. Are you 1000% certain that they were balanced at installation? New batteries being balanced is the exception, not the rule. Most are imbalanced as-received as they typically spend 30-90 days between final assembly/charging and delivery to the customer - not counting any delays in installation.

Observation
When I was in closed loop, if I pushed the voltage too high,

How were you regulating this DVCC control voltage?

the BMS would signal an Overvoltage Alarm to the Cerbo, which was really annoying.
Especially since none of the cell voltages were overloaded (watching BMS_Tools to verify). Others have discussed this at length.

This suggests strongly they were imbalanced. What was the corresponding CCL value when this occurred?

Question
Looking at BMS_Tools, I don’t see any overvoltage alarms, but I do see “Float Stoped” is set.
Yes, it is presumably a misspelling. I think the Cerbo interprets any bits set in that register as Overvoltage,

Can anyone tell me what “Float Stoped” means, and what triggers it?
It appears the BMS is disabling charging, which is making it really difficult to balance one low cell.
Thanks

Float stopped is likely the 0A CCL (charge current limit), i.e., don't send any current to the battery.

 

[ricardocello]

Properly configured, they should not go out of balanced. Are you 1000% certain that they were balanced at installation? New batteries being balanced is the exception, not the rule. Most are imbalanced as-received as they typically spend 30-90 days between final assembly/charging and delivery to the customer - not counting any delays in installation.

Well, I thought so at the time, because I looked at BMS_tools output for each of the three modules when I first got them, and I remember the cell voltage differences being less than 40 mV at 3.5V per cell (56V).

I wish I had taken screen shots a year ago. I’ve already cleared the logs. But I was a lot less knowledgable then.

How were you regulating this DVCC control voltage?

Yes, DVCC control voltage setting was very convenient.

This suggests strongly they were imbalanced. What was the corresponding CCL value when this occurred?

Only one cell is really low on one module, the others are fairly well balanced at 55V, maybe that‘s too low to know.

I probably need to hold everything at 56V for a while and reassess how bad it really is.
Or maybe creep up the voltage a little at a time so the modules don’t block charging.

Float stopped is likely the 0A CCL (charge current limit), i.e., don't send any current to the battery.

So because I have three modules, the CCL dropped form 150A to 100A to 0A, I assume because each module punts at different voltage levels.

What do you think about skipping the Quattro chargers for now and just hold the Chargeverter at 56.0 V overnight?

Thanks for responding!

 

[sunshine_eggo]

Well, I thought so at the time, because I looked at BMS_tools output for each of the three modules when I first got them, and I remember the cell voltage differences being less than 40 mV at 3.5V per cell (56V).

I wish I had taken screen shots a year ago. I’ve already cleared the logs. But I was a lot less knowledgable then.

I would call that balanced too.

Yes, DVCC control voltage setting was very convenient.

Truly!

The current control option is pretty awesome too. I've been working with a user with Trophy batteries that use the same Pylontech based protocol. It seems to me that the BMS doesn't consistently balance unless it's charging OR a cell is triggering charge protection.

I used it to limit charging to 2A, so it spends a couple hours above 3.45V with a measured current going in.

Only one cell is really low on one module, the others are fairly well balanced at 55V, maybe that‘s too low to know.

I probably need to hold everything at 56V for a while and reassess how bad it really is.
Or maybe creep up the voltage a little at a time so the modules don’t block charging.

My anecdotal evidence based on remote observation and periodic min/max voltage logging is that balancing occurs primarily when charging or in protection mode. Holding at an elevated voltage where one would expect balancing to occur wasn't particularly effective on the Trophy batteries.

Discharging to 95% and then 2A charging back to 3.45V/cell seemed to yield the best results.

So because I have three modules, the CCL dropped form 150A to 100A to 0A, I assume because each module punts at different voltage levels.

Those are triggered by the master BMS and are almost certainly based on individual cell voltage. I've watched the Trophy's do the same.

What do you think about skipping the Quattro chargers for now and just hold the Chargeverter at 56.0 V overnight?

Certainly worth trying.

 

[ricardocello]

The current control option is pretty awesome too. I've been working with a user with Trophy batteries that use the same Pylontech based protocol. It seems to me that the BMS doesn't consistently balance unless it's charging OR a cell is triggering charge protection.

I think so too.

I used it to limit charging to 2A, so it spends a couple hours above 3.45V with a measured current going in.

That‘s a great idea. I’ve just left mine at 90A, for balancing it makes more sense to slow it up.

Discharging to 95% and then 2A charging back to 3.45V/cell seemed to yield the best results.

I’ll try that next, thanks.

 

[NewGuyNoob]

I think so too.


That‘s a great idea. I’ve just left mine at 90A, for balancing it makes more sense to slow it up.


I’ll try that next, thanks.

Did you update to the latest firmware for that battery, is it the V1 [discontinued] or the V2? If V1, I have the same batteries, a firmware was released some time ago, I updated as I use them in a Schneider system. I keep my modules balanced by keeping float enabled, but changing voltage about about 90mins of absorption.

 

[ricardocello]

Did you update to the latest firmware for that battery, is it the V1 [discontinued] or the V2? If V1, I have the same batteries, a firmware was released some time ago, I updated as I use them in a Schneider system. I keep my modules balanced by keeping float enabled, but changing voltage about about 90mins of absorption.

I’ve got the v1 discontinued batteries.
I did the firmware update when they added Victron support, made my own cable, and successfully set it up.
Around the time when the youtube videos came out.

I haven’t heard if there has been another update.
I’ve had some success with the advice from @sunshine_eggo getting the one bad cell back in range.