EG4 Lifepower4 48V Top Balancing Firmware.

[adobopr]

We found the exact same problem. It has been brought up to engineering.

Some extra info from my end. While I also have the issue where currents under 2 amps are not shown, my SOC% still updates. I wonder if maybe the BMS is actually sensitive up to 500ma but the reporting is not providing the info between 500 and 2000 mA.

Also for mine it was exactly 2.0amps where it would last report and not go under. This gave me the immediate thought that maybe there's some error somewhere where 12v = 500mA, and it gets multiplied so the 24v battery reports after 1amp and the 48V after 2 amps. This is just a gut feeling but if someone has a 24v battery that they could test this hypothesis on, it might lead to finding a fix faster.

 

[Zapper77]

Some extra info from my end. While I also have the issue where currents under 2 amps are not shown, my SOC% still updates. I wonder if maybe the BMS is actually sensitive up to 500ma but the reporting is not providing the info between 500 and 2000 mA.

Also for mine it was exactly 2.0amps where it would last report and not go under. This gave me the immediate thought that maybe there's some error somewhere where 12v = 500mA, and it gets multiplied so the 24v battery reports after 1amp and the 48V after 2 amps. This is just a gut feeling but if someone has a 24v battery that they could test this hypothesis on, it might lead to finding a fix faster.

Yes, I used the under 2 amp "feature" ? when I updated the battery firmware.

Interesting idea with the .5 12v, 1 24v and 2 48v.

 

[shopman]

The only way to accurately measure current flow in the milliamp range is with a shunt. Unless the BMS uses a shunt to measure current flow, which I highly doubt it does, there is nothing to fix and the BMS is most likely operating at its maximum resolution.

 

[MaikaiLifeDIY]

Some BMS' use a shunt to do their calculations of what is going in & out of the battery pack. Most FET based BMS' use an internal coulomb counter or sensors to do a similar job. Either way, they're measuring the current in/out and that measurement as indicated in other forums on other BMS products is typically configurable, most likely in the firmware layer.

 

[MaikaiLifeDIY]

@Markus_SignatureSolar has stated in this thread #223 that the expected current reporting threshold is anything above 500ma. Perhaps you're correct and the engineers gave him the 12v amount rather than the 48v. If you step up the voltage behavior does seem to align with a possibility of 2 amps being the result of that 500ma increase from a 12v to 48v battery.

Currently, I have batteries reporting a 92% SOC, which fuzzy math means that a 100a battery has about 92a remaining, so theoretically I should only be able to put about 8-10 amps into the battery. I only have an active current meter so I'll have to find a way to count/verify the input amps, if the battery took say 20 amps, then I would know the SOC is significantly out of alignment with the battery's actual SOC.

The reason this is so important to me is because everything I want to do with my Sol-Ark, and many other automations through Solar/Home Assistant rely on this SOC being accurate to take action at certain levels. If it's off by 5% not a big problem, but if it's off by as much as 20%, that's the difference between having a battery that is either giving me a low SOC warning or simply shutting down because it's empty.

In an off-grid situation, I simply cannot rely on a SOC value that is significantly inaccurate.

 

[adobopr]

Unless the current of the pack is measured per cell, then I don't see why it would scale with voltage. And I did see me SoC move on mine so it is detecting current changes, buy maybe it gets measure differently. Not sure.

I also confirmed you don't get boot info if the BMS detects current, and was now able to upgrade all 6 of my batteries. I now have 2 which are severely out of balance and one which is not charging anymore but saying it's at 49%. Is this a bad battery or should I do a full discharge and recharge first?

 

[Zapper77]

Unless the current of the pack is measured per cell, then I don't see why it would scale with voltage. And I did see me SoC move on mine so it is detecting current changes, buy maybe it gets measure differently. Not sure.

I also confirmed you don't get boot info if the BMS detects current, and was now able to upgrade all 6 of my batteries. I now have 2 which are severely out of balance and one which is not charging anymore but saying it's at 49%. Is this a bad battery or should I do a full discharge and recharge first?

Do a full discharge and discharge..the SOC will reset to full at 56.5v

 

[ChillnGrill]

I have the 18kpv and the batteries never get too 100% only 99%. The voltage will only go to 55.3v. Is this normal?

 

[Zapper77]

I have the 18kpv and the batteries never get too 100% only 99%. The voltage will only go to 55.3v. Is this normal?

Are you using the battery communication hub? Which battery firmware version?

 

[ChillnGrill]

Yes and latest update

 

[Zapper77]

Yes and latest update

It took somewhere between 24 and 48 hours of just top balancing before my new rack of LP4 would hit 56v... I don't recall how much longer it took to reach 56.5v. It's a lengthy process. It won't happen overnight.

 

[adobopr]

Do a full discharge and discharge..the SOC will reset to full at 56.5v

I just tried this today and my cell alarm triggered. I think I either have a super out of balance battery or a bad cell, and I'm leaning towards a bad cell. The cell undervoltage alarm triggered at total voltage of like 49.7. A few seconds later I got the SOC undervoltage alarm, but not sure if the delay has to do with a comms protocol delay. The screenshot I took like 10 secs after, and I think total voltage went back up quickly because the cell started balancing and brought the voltage up a bit.

I also thought these had passive balancers but this does look like it's moving current between cells and not just discharging them. Am I reading this wrong?

This firmware did fix a bunch of stuff on my other batts, but I tihnk it may have also uncovered a bad battery. I can say I'm definitely liking this firmware update.

 

[Zapper77]

I just tried this today and my cell alarm triggered. I think I either have a super out of balance battery or a bad cell, and I'm leaning towards a bad cell. The cell undervoltage alarm triggered at total voltage of like 49.7. A few seconds later I got the SOC undervoltage alarm, but not sure if the delay has to do with a comms protocol delay. The screenshot I took like 10 secs after, and I think total voltage went back up quickly because the cell started balancing and brought the voltage up a bit.

I also thought these had passive balancers but this does look like it's moving current between cells and not just discharging them. Am I reading this wrong?

This firmware did fix a bunch of stuff on my other batts, but I tihnk it may have also uncovered a bad battery. I can say I'm definitely liking this firmware update.

I think...that qualifies as a full discharge.

You can only top balance or bottom balance...not both

Now you just need to fully charge and I suspect you may need several hours if now days of top balancing...seeing how that one cell seems oddly low.

 

[adobopr]

I think...that qualifies as a full discharge.

You can only top balance or bottom balance...not both

Now you just need to fully charge and I suspect you may need several hours if now days of top balancing...seeing how that one cell seems oddly low.

Ok, I'm gonna charge it with my bench power supply and just leave it isolated from the pack for a few days. while the bench PSU handles top balancing.

Will report back progress and thanks for all the help.

 

[Zapper77]

Ok, I'm gonna charge it with my bench power supply and just leave it isolated from the pack for a few days. while the bench PSU handles top balancing.

Will report back progress and thanks for all the help.

Check on it every 4-6 hours. After reaching Cell OV.

56.5 or 57v is a good target goal to start with. But don't get discouraged if a Cell goes to OV when under 55v.

 

[adobopr]

Check on it every 4-6 hours. After reaching Cell OV.

56.5 or 57v is a good target goal to start with. But don't get discouraged if a Cell goes to OV when under 55v.

That's a good idea. I isolated the battery but left it on comms while charging so I can monitor through SolarAssistant. Everything looking good up until now. Hope this battery gets itself back together after all this.

 

[adobopr]

Some updates in case anyone also runs into a huge balance difference like me.

I ran my lab bench at 56.2V for like 18 hours, but at some point I actually took the battery and paralleled it with the rest of the pack so it would charge a bit faster, then continued with 56.2V. I think it hit real 100% around 6 hours ago and it's been doing some type of cell balancing but it's been oscillating per my grafana graphs. This graph tracks the max cell voltage difference in a pack.


I did just raise the voltage of the lab bench to 57V which is that little which is when it went up to a 234mV difference, it was ~150 @ 56.2. I'm hoping the higher voltage triggers some better top balancing which is why I increased it.

Here's what SA is currently reporting.


This battery only reports about 11 cycles for some reasons while all the other ones are at around 230. I always thought the BMSs in general were wonky, but after the update they're all behaving pretty well, so I wonder if this 11 cycles is legit and has to do with either a bad cell or an imbalance that never got fixed. If this is a passive balancer only and cannot bring that last cell up, this will take a really long time to get that average down to the lost, so it could take a while until I know if the cell is bad or just very out of balance.

Will update tomorrow with more progress details.

 

[Zapper77]

Some updates in case anyone also runs into a huge balance difference like me.

I ran my lab bench at 56.2V for like 18 hours, but at some point I actually took the battery and paralleled it with the rest of the pack so it would charge a bit faster, then continued with 56.2V. I think it hit real 100% around 6 hours ago and it's been doing some type of cell balancing but it's been oscillating per my grafana graphs. This graph tracks the max cell voltage difference in a pack.
View attachment 164504

I did just raise the voltage of the lab bench to 57V which is that little which is when it went up to a 234mV difference, it was ~150 @ 56.2. I'm hoping the higher voltage triggers some better top balancing which is why I increased it.

Here's what SA is currently reporting.

View attachment 164506
This battery only reports about 11 cycles for some reasons while all the other ones are at around 230. I always thought the BMSs in general were wonky, but after the update they're all behaving pretty well, so I wonder if this 11 cycles is legit and has to do with either a bad cell or an imbalance that never got fixed. If this is a passive balancer only and cannot bring that last cell up, this will take a really long time to get that average down to the lost, so it could take a while until I know if the cell is bad or just very out of balance.

Will update tomorrow with more progress details.

Took around 72 hours before I was comfortable with the voltage difference. 4 of my batteries were good after 24 hours. 5th after 48 and (the 6th) I stopped after 72 hours...6th was close enough for me.

 

[adobopr]

Took around 72 hours before I was comfortable with the voltage difference. 4 of my batteries were good after 24 hours. 5th after 48 and (the 6th) I stopped after 72 hours...6th was close enough for me.

Did you cycle them as normal and just let them top balance during sunlight topup or did you leave them topped/charging like I am?

 

[Zapper77]

Did you cycle them as normal and just let them top balance during sunlight topup or did you leave them topped/charging like I am?

Left them charging. I cycled the breaker a few times and a few small discharges of around 5%. Mostly I just left them on charge.