0
\$\begingroup\$

I'm currently attempting to design a passive cell balancer for a 2S3P battery pack (pictured below).

Battery Config

Being new to power engineering I'm not sure where to start looking; I know I want to go with a "resistor bleeding balancer" design but that's about it. I'm looking for a recommended IC or design for a battery pack in this configuration and chemistry (LiFePO4). Also I'm not sure how I'd wire the balancing leads to the battery pack as shown too; can anyone point me in the right direction?

\$\endgroup\$
  • \$\begingroup\$ You ask for a design suitable for "this chemistry" but don't say what chemistry that is. In order to get meaningful answers might need to add more information: Charge rate; discharge rate; how long the pack needs to hold a charge for; efficiency requirements. You can edit your question by clicking "edit" in the bottom left. \$\endgroup\$ – Jack B Jun 23 '16 at 17:58
  • \$\begingroup\$ Good catch I missed that! \$\endgroup\$ – Sam W Jun 23 '16 at 18:07
  • \$\begingroup\$ If the cells are 3.2v, it could be LiFePO4. \$\endgroup\$ – rdtsc Jun 23 '16 at 20:21
2
\$\begingroup\$

Any circuit that uses an IC - or anything other than resistors, capacitors and inductors - is not passive. So what you are looking for is an active circuit that uses resistors to remove excess charge from the cells.

The maximum charging voltage for LiFePO4 is 3.65V. During most of the charging cycle the voltage rises very slowly, then shoots up rapidly as each cell reaches full charge. Therefore the balancer only needs to limit the voltage on each cell to 3.65V.

enter image description here

With this method the cells are balanced separately without comparing one cell to another, so each cell can have its own independent balancer circuit. LiFePO4 balance modules are often supplied as a board that is simply wired across the cell (example).

enter image description here

The circuit is basically a voltage comparator which connects a high power resistor (or bank of resistors) across the cell when its terminals reach the threshold voltage. The resistor bypasses charging current around the cell so it charges slower than the other lower voltage cell(s). Ideally the bypass current should equal the charging current, though if the cells are not far out of balance it can be lower. An LED is often added to indicate when the balancer is active, and also tells you that the cell is fully charged.

Here is an example that uses a MAX921 comparator with voltage reference:-

enter image description here

For a 2S pack you need two balancers, one for each cell in series. In a 2S3P pack you can join the upper and lower rows of 3 cells in parallel, then you only need 2 balancer circuits for the whole pack.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ I'm not sure I'm understanding this schematic correctly, are the battery cells labeled? \$\endgroup\$ – Sam W Jun 23 '16 at 20:20
  • \$\begingroup\$ The circuit is wired across one cell (LiFePO4- to LIFePO4+). For a 2S battery you need two of these. In a 2S3P pack you would have one for each of the 2 rows of 3 cells in parallel. \$\endgroup\$ – Bruce Abbott Jun 23 '16 at 21:22
  • \$\begingroup\$ Would you mind explaining what the variable resistor is for that connects between V+ and GND? \$\endgroup\$ – Sam W Jun 27 '16 at 18:27
  • \$\begingroup\$ In this circuit the MAX921 compares the IN+ voltage to the REF voltage of 1.182V (+-1%). The trimpot adjusts the voltage divider output to match this when the cell is exactly 3.6V. This circuit is also used with Li-ion and Lipo cells, which need a very accurate voltage setting. It probably isn't necessary with LiFePO4 because the peak voltage isn't so critical for them. So you could remove R5 and just join R4 to R6 and it should be close enough. \$\endgroup\$ – Bruce Abbott Jun 27 '16 at 20:10
  • \$\begingroup\$ I recalculated the resistor values and it seems I made a small mistake in the value of R4. For best adjustment range it should 2.32Meg, not 2.15Meg. Best value for R4 without the trimpot is 2.26Meg. \$\endgroup\$ – Bruce Abbott Jun 27 '16 at 20:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.