How does a BMS balance cells ?

I'm building a new pack, and would like to limit Cell Voltage to 4.05max to increase the lifespan of cells. I've been lurking at those "smart bms" out there, but they are quite expensive (would be half the price of the total pack!) and honestly, I'm not a fan of all the added functionalities. (more complexity = more bugs = more errors = more failures) The only thing I'm interested in, is a proper way to limit the max cell voltage.

Now when looking a classic cheap chinese bms ( like this one, how does the balancing work ? Is there an arbitrary threshold set upon which each P group looses or gains power, or does it simply divide the total voltage by the amount of cells (says 14s) to establish this value ? If so, is it possible to limit the voltage of each cell (or p group) by simply setting a lower charging tension (say 56.7 instead of 58.4)?

Those cheap or non-smart versions have "hard wired" comperators for each cell which starts the bleeding process at a certain voltage. same goes for low and high voltage cutoff. There are circuits which set these voltages. If you want to change the voltages you need to exchange these circuits.
See this video as an example: https://www.youtube.com/watch?v=xE3vBppr1sA&t=0s&list=LLsbZAMer8GvjB3PzlXI9m6A&index=26

izeman said:

Those cheap or non-smart versions have "hard wired" comperators for each cell which starts the bleeding process at a certain voltage. same goes for low and high voltage cutoff. There are circuits which set these voltages. If you want to change the voltages you need to exchange these circuits.
See this video as an example: https://www.youtube.com/watch?v=xE3vBppr1sA&t=0s&list=LLsbZAMer8GvjB3PzlXI9m6A&index=26

Click to expand...

Interesting. Does that mean those bms would not balance half charged packs at all ?

qwerkus said:

Interesting. Does that mean those bms would not balance half charged packs at all ?

Click to expand...

Correct. The typical cheap BMS can only balance when the cells reach around 4.2v. If you charge the pack to a lower voltage, it will never balance. New cells in good condition will tend to stay balanced for a long time. If you normally charge to a lower voltage it will take many cycles before balance becomes an issue.

What I do is normally charge to a lower voltage but occasionally check the cell voltages. If I see any imbalance, I charge to the full 4.2v/cell to let it balance. Charging to 4.2v once in a while is not going to hurt the cells as long as you discharge them shortly after charging.

fechter said:

qwerkus said:

Interesting. Does that mean those bms would not balance half charged packs at all ?

Click to expand...

What I do is normally charge to a lower voltage but occasionally check the cell voltages. If I see any imbalance, I charge to the full 4.2v/cell to let it balance. Charging to 4.2v once in a while is not going to hurt the cells as long as you discharge them shortly after charging.

Click to expand...

Are you sure this makes a difference ? According to the video, charging with a lower voltage only means the first Cells in series get charged more than the last, which in turn would mean those cells will die earlier.

No. This is not what the video says. Every cell is treated the same.
You can though change the balancing start voltage by choosing different "balancing circuits". But would you want to do that?
Risk the BMS and invest several hours of investigation and soldering just to save $20??

There are a number of threads on how BMSs work, if you look thru this list:
https://endless-sphere.com/forums/search.php?keywords=bms+work*&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
Not everything is relevant, mostly easy to figure out by title.

qwerkus said:

charging with a lower voltage only means the first Cells in series get charged more than the last,

Click to expand...

I can't imagine any way in which that would happen, with any charger, cells, bms, etc. that I've ever seen.

You could design a system that did that, but there isn't one I know of.

The exact opposite of that could certainly happen, however, since the first cell or cells usually get drained more than the rest, on cheap discrete-component BMSs, because they are used to power the BMS. Since the lower voltage means the pack is never balanced, then the imbalance caused by this gets worse over time).

(on others, the entire pack voltage is used to power it, so cells are not unbalanced by the BMS)


If all cells are in series and charged by a single bulk charger, with or without a BMS, they will all get charged the same amount.

How the cells respond to that charge is different depending on their own internal resistance, capacity, etc., so as cells age they react differently, but when they start out as new matched cells they will all end up identically charged.


The common cheap BMS just resistively drain any cell that goes above a certain voltage (depends on the BMS type, etc), until the cell drops below that setpoint. Usually it also turns off the charge port FETs while it's doing this, and turns them back on once the cell drops far enough.


There are also BMS that actually drain the high voltage cells and transfer their charge to the lower voltage cells, but those are typically more expensive, and are not any faster at balancing than the resistive types (probably slower); they're just more efficient cuz they dont' waste power as heat.

Thanks a lot for the explanation. Yes. changing the IC on a cheapo BMS seems to be out of proportion. Until now, I only build small packs, but starting with 100+ cells, a smart bms seems a good investment.

amberwolf said:

.....

The common cheap BMS just resistively drain any cell that goes above a certain voltage (depends on the BMS type, etc), until the cell drops below that setpoint. Usually it also turns off the charge port FETs while it's doing this, and turns them back on once the cell drops far enough.

There are also BMS that actually drain the high voltage cells and transfer their charge to the lower voltage cells, but those are typically more expensive, and are not any faster at balancing than the resistive types (probably slower); they're just more efficient cuz they dont' waste power as heat.

Click to expand...

That is exactly what I have in mind what an ideal BMS should do. Draining the higher energy cells to the lower ones. So they must have an array of FETs to transfer energy from group A to group B where appropriate. Just don't know that they are already available in the market. Can you please point to the exact model(s) or link ?

I also wonder how a BMS in big packs like EV/E-forklifts/E-trucks' work ? Given that the pack consists of thousands of cells to balance. Or they just work in the same concept but in a larger scale, i.e. may be balancing 10s A instead of 50-60mA ? Can you just give some real BMS models that they use ?

ichiban said:

Can you please point to the exact model(s) or link ?

Click to expand...

No idea of models, but they are sometimes called capacitive or charge-shuffling BMS, instead of resistive or draining types. Theyv'e been posted about on ES for years, so available at least that long.

Given that the pack consists of thousands of cells to balance.

Click to expand...

Doesn't matter how many cells, since like ebike packs, they generally parallel them up first, and balance them as groups like that. They aren't going to be balancing each cell, any more than the ebike versions do.

As for "models" of BMS, those are specific to the EV make and model--developed specifically for each one, and probably require the EV's onboard computer to work (which itself requires the rest of the EV to work, or at least all the sensors and other computers).

Thanks amberwolf

I'm also dreaming about making an EV from a used car.

First, determine the range needed...like 200 km/charge. Assuming avg 125wh/km energy needed for a 4-door sedan - is that a close enough figure ? So I will need 200km x 125wh/km = 25kwh. Spare 10% or so, that it will be 27.5kwh needed. That is approximate the battery size, cells capacity and types to choose from. Also that we need a/c, stereo, head+tail+signal lights.

Then, acquiring a car chassis (probably a good condition running 4-door sedan), remove the engine, transmission, radiator and all that not needed by our EV. Then find an appropriate size motor (40-50kW ?) with the RPM needed to get to the required top speed, then controller, BMS, electronics like display, protection circuits, etc.

Motor spec will determine battery serial & parallel connections. From that XsYp, the size, weight, connection and layout of the battery can be determined. And that also how & where we will install the battery in the car.

Sound simple ?

Actual work can be much more complicated. Is 18650 format the best suit to this application since they should be the best cost/performance among Li-Ions ?

Spot-welding thousands of 18650 cells using our DIY welder will be extremely fun