paulway / cell-balancer-v1

Cell Balancer circuit

cell-balancer-v1

This circuit and the accompanying source code are designed to balance a group of four cells using capacitive charge shuttling.

Circuit operation

The cells can be at different states of charge, but they must all be the same chemistry - i.e. they must all work within the same range of voltages. This means the circuit can be used to balance lithium ion, lithium polymer, lithium iron phosphate, nickel metal hydride, lead acid, chromic bifurcate, or whatever other weird chemistry you want to use.

The circuit can be used while the battery is being charged or being discharged, as well as (normally) when the battery is disconnected from other circuits. In general, make sure you discharge or charge all the cells, not just a subgroup of them - this circuit cannot move much current.

This circuit is powered from the cell itself. Therefore it will use a small amount of current in operation. This will drain the battery over time, and it is recommended to not leave it permanently connected to a battery unless the battery also gets regular charging current. It will not prevent over-discharging of the cells, or over-charging them. It only acts to make sure that if you do that, you kill all the cells at the same time :-)

(Stay tuned for version 2 of this circuit, which will feature an external power input, which should reduce the current draw on the battery when the circuit is off to nearly zero.)

This circuit does not contain any voltage or current sensing, user feedback or other IO. Sensing may be a useful feature in the future, but only to provide information to the user - this circuit does not need to know what voltage each cell is. The possibility was considered of adding LEDs at each drive input to signal which cell is connected to the capacitor in each group, but discarded because of the switching speed (see below).

How it works

Capacitive charge shuttling is a technique (among many) described in this paper:

http://www.mdpi.com/1996-1073/6/4/2149/pdf‎

and has been around since at least 2001 (see Moore, S.W.; Schneider, P.J. A Review of Cell Equalization Methods for Lithium Ion and Lithium Polymer Battery Systems. In Proceedings of the SAE 2001 World Congress, Detroit, MI, USA, 5–8 March 2001.).

In its simplest sense, it means connecting a capacitor across the terminals of one cell, disconnecting it, and then connecting the capacitor across the terminals of the adjacent cell, repeatedly. The higher-voltage cell charges the capacitor, and the capacitor charges the lower-voltage cell, until the two are equal. Several similar circuits can move charge through the pack until all cell voltages are equal. Using a capacitor allows the circuit to be very efficient and waste minimal power. Depending on the turn on and turn off times of each switch, several hundred or even thousands of cycles can be made per second.

It is worth pointing out that the circuit works on balancing voltages. Most cell chemistries have a 'state of charge' versus 'voltage' graph that is not linear. For example, lithium ion cells will reach a plateau at about 3.75 to 4.0 volts, where continually applying power does not increase the voltage but still charges the battery. In this region, the cell balancing has less effect - the voltage difference between two cells may be minimal despite them being at different states of charge. Once at the same voltage, though, two adjacent cells will stay balanced fairly well.

If the rate of charge or discharge is too fast, it is possible for the balancer to not keep up with differences between cells. This can be

There is one possible failure mode which is worth noting. If two optorelays in the same row in the schematic - e.g. K1 and K2 - are turned on simultaneously, it will short out cell 1. At this point, one or more of these devices will blow up to protect the cell. Because the devices take time to turn on and turn off, there must be a 'dead time' when both optorelays in the row are turned off to prevent this short. Diodes won't help, because current has to be able to flow both ways (into and out of the cell).

The job of ensuring this dead time is given to the microprocessor.

Microprocessor

The code for each capacitor drive group is fairly simple:

1. Turn on optorelays for cell A.
2. Wait (turn on delay) + (shuttle time)
3. Turn off optorelays for cell A.
4. Wait (turn off delay) + (dead time)
5. Turn on optorelays for cell B.
6. Wait (turn on delay) + (shuttle time)
7. Turn off optorelays for cell B.
8. Wait (turn off delay) + (dead time)

The overall speed of this circuit is:

( (turn on delay) + (turn off delay) + (shuttle time) + (dead time) ) * 2

The smaller this time, the faster charge can be shuttled between two adjacent cells.

The turn on delay and turn off delay are set by the schematics of the part chosen.

In this design, the shuttle time is fixed. It must be enough to allow the capacitor to charge or discharge - if the cell is close in voltage to the capacitor it will take longer for the voltage to change in the capacitor. Yet it must not be too long - once the voltage in the capacitor is close to the voltage in the cell, very little extra charge will flow; time would be better spent switching the capacitor across to the other cell.

The shuttle time is the area which offers the most promise for future circuit improvements. Ideally, the shuttle time is over when the voltage difference between the cell and the capacitor is closer than a certain set threshold, which (too) can be tuned; this requires voltage sensing circuitry.

The dead time is a small fixed time, chosen to be enough time for the optorelay to stop conducting current and resist further current flow.
This too must not be too long - it is time in which nothing is happening and cells are not being balanced. It is suggested that this time be at least half the turn off delay of the optorelay.

Optimising the