diy solar

diy solar

Observation about Grade B packs / Low Quality BMS and Wasted Energy while charging

robby

Photon Vampire
Joined
May 1, 2021
Messages
4,050
I have three Fortress Power eFlex Batteries and one EG4 LL Battery all in parallel using equal length wires going to Busbars and then to a Sol-Ark 12K Inverter.

My eFlex batteries (Grade A) will charge by 11am in the morning. My EG4 LL Battery (Grade B) will take until 1pm to Charge. This was something that was obvious on day one. What I have learned since then is that the Eg4 Battery wastes a bunch of PV power trying to balance the cells.

Where as the eFlex batteries will balance in 15 minutes and do so with minimal power draw. The EG4LL is continually drawing 14 Amps then discharging 10 Amps alternating between the two every 2-3 minutes and spends a good two hours doing this balancing.

The first problem is the fact that the EG4 battery has several cells that will reach up to 3.65V very quickly and several that will be at 3.35V and trying to catch up while the eFlex cells all charge up nice and evenly to 3.43V and just need a slight tweak to a random cell, maybe one is at 3.40V.

The difference in the way the balancing happens on a cheap BMS versus an expensive one is kind of interesting. The expensive one will take power from the higher charged cells and reroute it to cells that have lower voltages until the pack is balanced, this is the way I would expect it to be. It may pull a bit of power from the Inverter to complete the process but not much.

The Cheap BMS with B cells is basically charging the pack until a single cell hits 3.65V and then it stops charging. At this point several cells in the pack may still be at 3.35V and what the BMS does is bleed off some of the power 10Amps at a time to the Inverter until the 3.65V cell drops to 3.60V then it will charge the pack again for a few minutes with 14 Amps of power until the 3.60V cell once again reaches 3.65V and repeats the cycle.

The Idea seems to be that during this charging period the 3.35V cell will get up to 3.40V and the cycle will repeat until all the cells are at around 3.43V. The reality is that when the 10 Amp discharge happens the 3.35V cell of course loses energy so it ends up at 3.32V. So that 14Amp charge might bring it up to 3.37V and not 3.40V. So this cycle will rinse and repeat in a two step forward one step backward motion with each cycle using 4 Amps of extra power for nearly 2 hours.

If you switch off the Breaker on the EG4LL and are thinking that it will balance itself you can forget it. The moment the breaker goes off and the battery cannot discharge the 10 Amps the whole process stops and the cells remain at whatever voltage they were at even though the BMS is on.

The eFlex on the other hand will continue to balance the battery even when turned off.
 
Last edited:
Yep, passive vs. active balancing but the low cells should not be getting discharged at all. Perhaps what you are seeing is the voltage of the low cells "relaxing" during the period of time they are not being charged. Have you tried Top Balancing the EG4?
 
Yep, passive vs. active balancing but the low cells should not be getting discharged at all. Perhaps what you are seeing is the voltage of the low cells "relaxing" during the period of time they are not being charged. Have you tried Top Balancing the EG4?
The EG4 eventually balances out the cells although three of them will often end up at 3.7V before they all come back down to 3.43V it's the time and energy wasted that I find a huge pitfall. If I had six of these they would be wasting over 1200 Watts of power just balancing themselves. (based on one hour at 4 Amps each)

The drop in voltage could be the cells relaxing, I have no way of knowing but that would seem odd as how can the pack be putting out 55V at 10 Amps if all the cells are not active?

If by Top Balancing you mean opening the case and doing it cell by cell then no. That would void whatever warranty I might have.

This problem does not happen on every charge cycle but is guaranteed to happen if I let the battery run down to 10% SOC.

I am not all that familiar with the LV6548 and 6500ex but I am wondering what those inverters do during charging.
The Sol-Ark has zero problems moving from Consumption to Charging but I wonder if the error codes people are getting with the LV6548 and it's clones are related to this constant back and forth?
 
Last edited:
What charge voltage are you using?

Ps, sorry typing as you posted.

Mike
 
What you are describing is a BMS issue rather than a cell issue. All cells will eventually begin to behave like your EG4 cells, so it is better to set up the system to handle this situation right from the start.

That is the main advantage of having the BMS in control of the charge controller. When the first cell reaches its balancing voltage the current reduces to a level that allows balancing to occur without needing to stop charging completely.

Active balancing will achieve the same result, but adding components that have the possibility of draining your cells without BMS protection isn’t something i’m interested in.

These cells should last at least 10 years, the BMS needs to be robust if it is to outlast the cells.

The single most important component in a LiFePO4 system is the BMS. The reason server rack packs like Pylontech, BYD, even Simpliphi failed early is because they set their BMS parameters incorrectly. Most server rack style batteries still have woefull BMS settings (as you have observed).

The only solution i have found is to assemble your own battery using a better BMS.
 
The EG4LL is continually drawing 14 Amps then discharging 10 Amps alternating between the two every 2-3 minutes and spends a good two hours doing this balancing.
Unless you've got some closed loop communication set up with your EG4 battery and the inverter, this isn't a battery issue. The inverter is what pushes and pulls current from the pack. The cells are not in control of where the current goes. The inverter is the device moving these 10 and 14 amps.

Also, the "wasted" energy would have to be in heat. You're claiming that somewhere between 4 and 10 amps is wasted by the battery. For easy, round math, at 50ish volts that would be a lot of heat, 200-500 watts would make the pack hot for sure.
I don't know the balance current of this specific BMS but, I'd expect something like 40-400mA that are burned off by passive balancing. Pretty standard.

And what was the cost on the Fortress? Double the cost per kWh of the EG4?
 
What is the absorption voltage you are charging with?
Absorption is normally set to 54.5V at times I have had to move it up to 56V with just the Eg4 only in order to get it to balance the cells when they get really mismatched.
 
1) the Exact Volts/AmpHours of Each Pack.
3 x 105Ah eFlex and one 100Ah Eg4LL
2) All of the Charging Profile information

I just read through the manual, there seems to be no EndAmps / Tailcurrent Value, it's important with LFP, as the Ending Amps (100AH Batt = 5A) would kick off Absorb & goto Float. It also really helps to control Runner Cells.

Try this:
Absorb, 30Min Max @ 55.2V (3.450Vpc)
Float. 55.0V (3.437Vpc)
Done most of that already. There are other people having this issue and I suspect a lot more that don't even realize it because all they are looking at is that the pack voltage is 54.5 or 55V. I can get to that every single time but the cells will still be all out of balance.
NEVER EVER Equalize !! Disable or set to "Float Voltage" for as little time as possible. (0 Min works as disable)
* ABSORB with lower running time can partly compensate for lack of EndAmps/TailCurrent value.
Nope not in all cases. If you have an Inverter that allows you to set the Equalization time and how often to do it then it's a good thing. I have had two of the Teir one companies recommend 55V Equilization voltage every 14 days for one hour. It keeps the batteries better balanced and the SOC if done by the Inverter itself is fine tuned better.
Passive Balancing is NOT intended for Large AH Batteries, it IS intended for < 75AH really.

Active Balancing @ 2A will work to 250/300AH, High Amperage for Larger AH Batteries (think paralleled cells)
Simple rule of thumb, 1A Active Balancing per 100AH of capacity.

Hope it heps Good Luck.
I agree, that is why it had not even occurred to me that the EG4LL was using a Passive balancing system until @BentleyJ said it.
 
What you are describing is a BMS issue rather than a cell issue. All cells will eventually begin to behave like your EG4 cells, so it is better to set up the system to handle this situation right from the start.
All Grade B cells behave like this. Trust me I have been Monitoring my expensive Grade A cells and they do not behave like the ones in the EG4LL. I do agree that a cheap BMS makes the problem even worst.
That is the main advantage of having the BMS in control of the charge controller. When the first cell reaches its balancing voltage the current reduces to a level that allows balancing to occur without needing to stop charging completely.
Agree
Active balancing will achieve the same result, but adding components that have the possibility of draining your cells without BMS protection isn’t something i’m interested in.
You would like the approach Fortress Power takes. It uses such a small amount of power yet does the Job really well and quickly.
These cells should last at least 10 years, the BMS needs to be robust if it is to outlast the cells.
If I get 5 years I would consider it a break even point on ROI.
The single most important component in a LiFePO4 system is the BMS. The reason server rack packs like Pylontech, BYD, even Simpliphi failed early is because they set their BMS parameters incorrectly. Most server rack style batteries still have woefull BMS settings (as you have observed).
Agree, luckily Fortress Power does not have that issue, they are constantly testing banks of batteries in their labs and send out updates on settings that they are seeing better results with. Recently they sent out a BMS update that allows manual cell recovery even is the battery has depleted down to less than 30V. On power up it will go into a stanby mode if Voltage is below the shutdown level and will allow you ten minutes to start a manual recovery charge. If no charge is detected in 10 minutes the BMS will shutdown the system again to conserve remaining power.
The only solution i have found is to assemble your own battery using a better BMS.
I am really thinking about gutting this EG4LL to see if I can actually put a good BMS in it to al least make the B cell issue more tolerable.
 
Hover over my Avatar and select IGNORE and POOFDAH Home Simpson would be a better avatar for you versus a Southpark toon.



2 REAL resources to help you gain some understanding, that's all for you.

I've cleaned out my posts in this thread & mods will housekeep it.
I deleted the Midnite Solar Thread on the Classic Controllers as well. Too Bad for those who ACTUALLY OWN & USE Midnite Solar Gear.
 
Last edited:
You linked this thread, stating worries that the cells are bad and grade B cells are much worse than grade A and waste a lot of power when charging...
But it's a pack balance issue with your grade B pack and the cells seem horribly out of balance.
So yes balance and cell matching is an issue with grade B cells, but this is why top balancing prior to assembly is done. Also who active balancers are popular.
If you can disassemble the pack, put a better BMs on. Or attach an active balancer board with voltage relay to activate above 3.4v.
If you see one cell continuously running above or below the others despite regular balances, that cell has a much smaller capacity and the rest. Take it out, test it, and ideally replace with one that has similar capacity to the rest.
 
I think Robby has provided sufficient evidence for his claims, It appears EG4 is selling ganfeng made battery which potentially could be using grade B cells but I think the intended purpose of this battery was for use in telecom cabinets to provide back up power or something similar, I do not think it was ever designed to power up vaccums and other things that require a big load.

How many other companies that have this style of battery are also using Ganfeng built? Perhaps this is why stuff like fortress is 2-3x price because they are using EV grade or professional grade.

I think as this industry matures more and more information will come to light similar to the grade A vs grade B cells. The Gold Rule still applies, you get what you pay for and if sounds too good to be true, it probably is.
 
You linked this thread, stating worries that the cells are bad and grade B cells are much worse than grade A and waste a lot of power when charging...
But it's a pack balance issue with your grade B pack and the cells seem horribly out of balance.
So yes balance and cell matching is an issue with grade B cells, but this is why top balancing prior to assembly is done. Also who active balancers are popular.
If you can disassemble the pack, put a better BMs on. Or attach an active balancer board with voltage relay to activate above 3.4v.
If you see one cell continuously running above or below the others despite regular balances, that cell has a much smaller capacity and the rest. Take it out, test it, and ideally replace with one that has similar capacity to the rest.
Very good solutions if this was a DIY pack but this is an EG4LL with a 10 year warranty (maybe) and doing any mods will void that. As for changing cells they, are all laser welded together.
 
Back
Top