safe said:
Is anyone going to get back to the "balancer" question?
Yes. I am.
The idea of balancing is to try to keep all of the batteries at the same level of charge. This could be considered the goal of any charging system - except that nearly all multi-cell battery chargers don't actually try to directly achieve this goal. Nearly all multi-cell chargers out there charge the battery pack until an "end goal" is met that indicates the battery is fully charged. They don't spend any effort trying to make sure that each cell is charged, they just watch the pack overall to detemine if the cells are charged or not. So if the pack appears to be charged by the charger, then it is assumed that all of the cells are charged... even if not all of them are. The goal of a balancer is to make sure that each cell is charged to a similar level as all the other cells in the pack.
Balancers are useful with all battery types, but they are particularly important in lithium-based battery systems that don't have battery management systems (BMSs) because out of balance cells can lead to safety concerns.
The way that lithium ions batteries charge is pretty straightforward - particularly compared to nickel-based batteries - you apply a constant current, until a voltage threshold is met, usually 4.2V per cell, at which point the charger switches to a constant voltage mode where the current is modulated downwards to hold the voltage at 4.2V. Imagine a scenario where you get a pack of 20 3.7V lithium-ion cells and you hook them all end-to-end (20 cells in series) to create a 74V pack. In a 74V lithium ion charger you would apply a constant current to the pack - usually 1C - until the pack reaches 84V at which point, the charger would modulate the current to hold the voltage at 84V until the current drops to a small value and the charger deems the battery to be charged.
But imagine that in these 20 cells, a couple self-discharge a bit faster than the others. So the charger charges until 84V, but for a couple of cells that are not as charged as the rest because they lost more due to self-discharge, they are a bit under charged at 84V, and the rest are a bit overcharged at 84V. The pack reaches 84V, but some of the cells are at 4.1V and some others are at 4.3V. Over time, these cells become more and more "out of balance". The BMS will disable overcharging at 4.35V, but still you can have much of the pack at 4.35V - overcharged - and a handful of other cells at 3.75V and the charger thinks this pack is fully charged (15 * 4.35V + 5 * 3.75 = 84V).
So, the BMS stops the pack from overcharging a cell, but then when you discharge the pack, the BMS will disable the pack when it is overdischarged (typically any cell drops below 2.3V). So you have some cells that are only getting charged to 50% capacity at 3.75V and then the whole pack shuts down if any cell reaches 3.0V. So, now your effective pack capacity is 50% of it's rated capacity. But this effect isn't permanent(*). All you need to do is "balance" them and you'll be back close to 100% capacity again.
A balancer essentially looks at each cell individually and compares it against all of the other cells and then adjusts the voltage of them to equalize to the same voltage value (and hopefully to the same state-of-charge). The way that they work is they compare the voltages of the cells and then for all of the cells that are higher than the lowest cell, they shunt in a variable resistance in parallel with the cell to drain it faster. This can be done while charging or discharging. More advanced techniques use charge shuttling to charge - or discharge - cells by connecting cells together.
Any easy way to balance a pack is to switch to parallel charging them as individual 3.7V cells periodically. So you take your hypothetical 20 series pack and charge it all in parallel as one 20-cell parallel pack. This takes a long time though. Another alternative is to do what Xyster does and always charge them as individual cells.
There's a paper on the subject here (found with Google "lithium ion balancer"):
http://www.americansolarchallenge.org/tech/resources/SAE_2001-01-0959.pdf
Patrick Mahoney
Fort Collins, CO
(*) Well, except that 2.3V is horribly low in my opinion. Way too low to not cause permanent damage, based on my experience. One of the first things that I'd do if I bought one would to be reprogram it for a higher "low voltage cut-off".
Do you know something the battery "specialists" don't know?
