So is 93.7 my max capacity now?

[fnnwizard]

That is only true if only one cell is high and rest 15 are 1% lower. Then passive balancer burns that one cell down in about 10h. But OP has at least one cell substantially lower than 15 others. Your example with one cell 1% lower than others would mean that actually passive balancer needs to burn that 1Ah times 15. So balancing time is ~150 hours...

With active balancer things change as it transfers energy from the highest to lowest so deltaV comes down quite fast.

What passive BMS are you using/referencing? Serious question. Since... every passive BMS that I have ever looked at or used has individual cell discharge circuit that the respective mosfet will connect to the cell's respective resistor (bank) to do the discharge. 16 cells = 16 individual resistor banks. Each bank could be 1 or more resistors in parallel.

If there were such BMS's discharging let's say 100mAh @ 3.40-3.65V, that's < 0.4W total for the whole pack, which would be an even more extremely poor design into ~900Wh/cell. I would most definitely stay away from those, if there are such BMSs!

Show me a bms and I will show you where those resistors are .

 

[sunshine_eggo]

What passive BMS are you using/referencing? Serious question. Since... every passive BMS that I have ever looked at or used has individual cell discharge circuit that the respective mosfet will connect to the cell's respective resistor (bank) to do the discharge. 16 cells = 16 individual resistor banks. Each bank could be 1 or more resistors in parallel.

If there were such BMS's discharging let's say 100mAh @ 3.40-3.65V, that's < 0.4W total for the whole pack, which would be an even more extremely poor design into ~900Wh/cell. I would most definitely stay away from those, if there are such BMSs!

Show me a bms and I will show you where those resistors are .

I'm with you on this one. As a general guideline, most low current passive balance BMS can adjust 1Ah per 24 hours of "active" balancing.

 

[Lt.Dan]

That is only true if only one cell is high and rest 15 are 1% lower. Then passive balancer burns that one cell down in about 10h. But OP has at least one cell substantially lower than 15 others. Your example with one cell 1% lower than others would mean that actually passive balancer needs to burn that 1Ah times 15. So balancing time is ~150 hours...

With active balancer things change as it transfers energy from the highest to lowest so deltaV comes down quite fast.

There are many BMS's (actually I assume all, but I'm not sure) that will absolutely balance more than 1 cell at a time. If my Seplos BMS sees that 15 cells are over 3.45v, and 1 cell is under, then it will burn off from all 15 at once, not one at a time. In this case 150mA × 15 = 2.25a of "total" balance current.

Whereas active balancers can only discharge/charge 1 cell at a time, but typically with a much higher current (0.6a, 2a, 5a, etc)

 

[sunshine_eggo]

Batrium can do about 0.5A passive, but I'm not entirely sure it has a resistor for EVERY cell. it seems like it has one per two cells. I don't recall ever seeing any two adjacent cells balancing...



6 should obviously be burning off, but #5 is going. You'll see them hop back and forth a couple times a second.

 

[fnnwizard]

There are many BMS's (actually I assume all, but I'm not sure) that will absolutely balance more than 1 cell at a time. If my Seplos BMS sees that 15 cells are over 3.45v, and 1 cell is under, then it will burn off from all 15 at once, not one at a time. In this case 150mA × 15 = 2.25a of "total" balance current.

Whereas active balancers can only discharge/charge 1 cell at a time, but typically with a much higher current (0.6a, 2a, 5a, etc)

The Seplos 10E (48x200) bms has 20Ω resistors for each "balancing circuit" so that lines up very well with 150mA considering additional resistance in rest of components/connections and they are rating it at 3.2V, but I would think it's likely a smidge more since we balance only >3.4V.

Oh, the likely limiting factor is the heat dissapation capability for the bank(s) of resistors. Taking Seplos worst case example, we could be >10W. (15 (>3.6V) (.15A)) for that little section of PCB.

I'm with you on this one. As a general guideline, most low current passive balance BMS can adjust 1Ah per 24 hours of "active" balancing.

There are tricks to try and keep user configurable bms's (especially that Seplos mentioned above) in its continuous balancing envelope and one can get close to the rated balancing specs during balancing operations. If anyone wants to know, I can start a new thread.

 

[fnnwizard]

Batrium can do about 0.5A passive, but I'm not entirely sure it has a resistor for EVERY cell. it seems like it has one per two cells. I don't recall ever seeing any two adjacent cells balancing...

View attachment 251904

6 should obviously be burning off, but #5 is going. You'll see them hop back and forth a couple times a second.

Are you using Batrium Cellmate?

 

[sunshine_eggo]

Are you using Batrium Cellmate?

Nope. I bought shortly before they were released. WM5 and MM8, but the MM8 isn't installed as the other half of the bank is still here and there.

 

[marionw]

SunGoldPower 48v100ah default BMS settings

Cell Over Voltage Alarm = 3.60vdc
Cell Over Voltage Protection 3.70vdc (BMS stops further charging)
Pack Over Voltage Alarm 57.60vdc
Pack Over Voltage Protection 58.40vdc (BMS stops further charging)
Cell Balance 3.50vdc (one or more cells have to reach this voltage before BMS starts balancing as indicated by "BL")
Balance delta 30mv

If you are routinely charging the battery to where any cell reaches Cell Over Voltage Protection 3.70vdc you will shorten the life of the battery. Some will argue that this is not an issue and others will argue that you never ever charge a cell to or over 3.65vdc. I reset the Cell Over Voltage Alarm to 3.55vdc and Cell Over Voltage Protection to 3.62vdc

Once any cell reaches the Cell Over Voltage Protection of 3.70vdc the BMS turns off the Charging MOSFETS and the cell(s) will settle to a lower voltage.

The Cell Over Voltage Protection will not be released until the offending cell(s) voltage drops to 3.38vdc.

I have had batteries where I could not get one or more cells to charge properly. I opened the battery and used an adjustable benchtop power supply set to 3.50vdc to individually charge the cell. Get the cell charged to 3.50vdc, remove the charge and wait 30 minutes. The cell voltage will drop. Charge the cell back up to 3.50vdc, remove the charge and wait 30 minutes. Each time you do this the cell "drop" voltage will be less. I also do this to the other cells until all cells settle to the same voltage after charging.

Interesting video on State Of Health (SOH) calculations:

 

[fnnwizard]

Nope. I bought shortly before they were released. WM5 and MM8, but the MM8 isn't installed as the other half of the bank is still here and there.

Ah, ok. Not familiar with those exact units and only had some white papers on them a few years back when a friend ask about using them with some LG packs since they could do either lfp or nmc's. ( btw, LG finally threw in the towel and will be building LFP soon)

Anyway, I think Batrium have 2 types of system.
1 with resistors physically on each cell and a "centralize system" which I think is what you have.

If I recall correctly, the centralize system is very unique in that it does a "poll" of each cell and targets the high V cells, but targeting 1/2 the pack = every other non adjacent cell sequentially, as you described. So in your picture the next cell to get "zapped" by the resistor would be 5, 7, 9 etc. ( it may or may not hit 11 and 13)

I'm not sure if they have individual resistors for each, but I do remember them having the highest balancing capability around. They were considered the Cadillacs and were priced as if = very expensive. Totally forgot about them!

 

[shadowmaker]

What passive BMS are you using/referencing? Serious question. Since... every passive BMS that I have ever looked at or used has individual cell discharge circuit that the respective mosfet will connect to the cell's respective resistor (bank) to do the discharge. 16 cells = 16 individual resistor banks. Each bank could be 1 or more resistors in parallel.

Forgot all about that .

But then on my cheapest Chinese premade rack batteries there are Topband BMSs which seem to be able to balance only four cells at time. Manual says balancer is capable for 30-65mA (delta 20mV).

So 1Ah for me can be ~62 hours.

While I have my batts full between 10-22 (12 hours) almost every day in summer it's not enough for that crappy balancer to balance out every batt. It seems that after Neey 4A top balance that original balancer is still good enough to keep the balance though. This is why I have installed only harnesses (~15e) to some of my worse batts (bad initial top balance), but still have only one Neey balancer.

 

[sunshine_eggo]

Ah, ok. Not familiar with those exact units and only had some white papers on them a few years back when a friend ask about using them with some LG packs since they could do either lfp or nmc's. ( btw, LG finally threw in the towel and will be building LFP soon)

Anyway, I think Batrium have 2 types of system.
1 with resistors physically on each cell and a "centralize system" which I think is what you have.

Yes. The WM5 is the direct predecessor to the WM Core. The MM8 is the predecessor to the Cellmate K9. They are centralized. Both supervisors are compatible with the individual cell monitors (blockmon, longmon, etc.).

The WM5 is basically a CORE + a Cellmate K9 in one unit, i.e., it can manage 10-15 cells without another

If I recall correctly, the centralize system is very unique in that it does a "poll" of each cell and targets the high V cells, but targeting 1/2 the pack = every other non adjacent cell sequentially, as you described. So in your picture the next cell to get "zapped" by the resistor would be 5, 7, 9 etc. ( it may or may not hit 11 and 13)

Yep.

I'm not sure if they have individual resistors for each, but I do remember them having the highest balancing capability around.

Yep. I screwed up the initial deployment of the battery and ended up with a bit more than 10% variation in SoC. While still learning how to use it, it was able to bring the dV in line in a few days:

They were considered the Cadillacs and were priced as if = very expensive. Totally forgot about them!

Yep. I have about $1K into my BMS.

 

[Ennis437]

Well guys thanks to the advice on the forum I was able to take the lid off my Sungold rack battery and charge cell 13 to 3.6v. I then let the battery settle and it fell back down to 3.3. I then charged back to 3.6, let it settle for 30 minutes, and then repeat the process until it held the charge. I'm happy to report these results after a full discharge and recharge. I am back to full capacity!