Here is an explanation I got once when I asked, but this describes more what they do for charging and not so much for the low voltage or current cutout aspects.
Re: Headway Group Purchase
Postby GGoodrum » Tue Mar 10, 2009 8:32 pm
If you think about how a charger woks with a single cell, it is easier to understand what the BMS does. If you have a single cell and a charger that has a 3.65V limit, what happens is that the charger starts putting out as much current as the cell can accept, up to the charger's limit, 2A, in this case. As the cell gets fuller, it becomes harder for the cell to accept current at that rate so the cell voltage will rise at a fairly steady rate. When the cell gets to about 3.65-3.70V, it suddenly becomes really hard for the cell to accept current, so the voltage starts to rise at a much higher rate. Depending on the cells, this rate change in voltage increases can be quite significant. In any case, if you stopped at that point, the cell will be about 80-85% full. To completely charge the cell, what the charger needs to do is simply keep the voltage from rising above the 3.65-3.70V level. What happens is, since the voltage can't rise anymore, the cell starts reducing the amount of current that it lets in. The current will stedily drop, from the max the charger can put out, down basically to zero. It is generally assumed that when the current drops to under about 50-100mA, the cell is about as full as it is going to get. this charging profile is known as Constant Current/Constant Voltage, or CC/CV.
If cells are perfectly matched, and start out at the exact same SOC level, you can have cells in series, and just use a charger set to 3.65V x # of cells. They will all hit the 3.65V level, and start reducing the current they let in, at the same rate. In reality, cells are never exactly the same, and will drift apart ofver time, no matter how well matched they are when the pack is manufactured. As the cells become unbalanced what happens is you have some cells hitting the 3.65V crossover point (from CC mode to CV mode...) before the rest. The voltage in these cells will start to rise very quickly, and at some point, the pack's total voltage will be above the total crossover point, and the charger will start holding the voltage there. This will cause the high cells to start dropping the current they let in. The problem is that the way electricity works, cells can be at different voltages, but all the current that is circulated, must go through all the cells, all the time. If you have a high cell that starts limiting the current, it is also limiting the current available for all the cells. The net result is you end up with cells that were never allowed to finish, and they just get further out-of-balance with the high cells after each charge.
They way most BMS boards work is that there is a 3.65V-3.70V limit put on each cell. Once any cell reaches this point, a shunt circuit will keep it there, and bypass a certain amount of current on to the next cell. That way each cell can at least get whatever the shunt current limit is, so that it can finish getting a full charge. In our BMS, we turn on an LED when each cell's shunt circuit first starts to conduct current. We have logic that drives a master red-green LED that starts out red, while all the cells are still in the CC mode. Once the shunts start operating, this LED will start to turn towards green. When all the cells are completely full, the LED turns full green, and the charge current is cut off completely. In any case, the amount of shunt current allowed, will determine how long it takes to let each cell get full. For most of the typical 2-4A chargers that usually come with ebike "duct tape" packs, the shunt current is usually somewhere between 100 and 150 mA. With the standar parts called out in our BMS, the shunt current is higher, at around 500mA, or 1/2A. For ebike packs, the limit isn't usually a big deal, but with larger EVs, it does make quite a difference. If you have a 100Ah capacity, even 1/2A is probably not enough.
Anyway, my point to all this is that most BMS units don't really "balance" the cells, per se, but instead let each cell get to its 100% charge level, whatever that happens to be, and at its own pace. Functionally, it is just like using individual cell chargers, like the VoltPhreaks units. There are some units that will actually balance the cells, basically by reducing the level of the high cells, down to the level of the lowest, but I'm not a big fan of this type of approach. It works okay, to bring cells that happen to have a lower SOC at the moment, back up to the rest once balanced and recharged, but where this logic falls apart is if the cells truly do start drifting apart in capacities, IR and thermal coefficients, which eventually all cells seem to do. If each cell is protected from being over-discharged, and each cell is allowed to get a full charge, it really doesn't matter how well balanced the cells stay, in relation to each other.
-- Gary
26" Townie: Crystalyte 5304; 18-FET (4110) 100V/100A; 24s3p 88V/15Ah Turnigy 20C LiPo
26" Townie: Crystalyte 5304; 7240v2; 24s 72V/10Ah PSI
20" Dahon Mariner: 9x7 9C (soon with D/Y switching); 18-FET 100V/100A 18s2p 67V/15Ah Turnigy 20C LiPo
20" West Marine Port Runner: AF 3220-7t; HV110; S-A 3-sp hub; 12s3p 44V/15Ah Turnigy 20C LiPo
User avatar
GGoodrum
10 MW
10 MW
Posts: 2010
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA
View attachment BMS.jpg
