Update: I have this board working, testing it on some worn out 15 year 50 AH Lifepo4 cells, some slightly bloated, and down to 40% capacity, so they are all over the place; however they make great 12v batteries for powering field electronics from small PV panels.
I set the charger end point voltage to 3.5v and the balance voltage to 3.52v per cell.
Grabbing 4 series cells with no prior top balancing and using 2 x 10r smd 2w resistors in parallel as the balance load for each cell, things get very hot 80°c, as the balancing time is maybe an hour or so, allowing temperatures to rise.
Using larger through hole resistors mounted above the pcb works fine with the approx 700mA balance currents. The optimos relay chips are rated to approx double this current and get slightly warm, so they are ok.
Once the cells were balanced, loading them up with several AH of discharge, then repeating the process, worked much quicker, all cells being within a few mV of each other. This was using the CPU's internal 10bit adc.
I havent been able to test using the I2C adc chips as my order from Element14, somehow went via the USA and has been impounded by US customs with no release date, blame the orange idiot for that. Element14 are taking their time in sending me replacements
The design would work better for these larger cells if the load resistors were mounted externally an a small heatsink, even better, would have the balance currents increased to several amps. To that end I have altered the pcb some what to have low RDson mosfets driving the loads, the optimos switches now power the gates rather than the full load, the HL100N2 mosfets wont even get warm and work down to 2.5v gate drive.
Here is the new pcb, sending off today, same size 100 x 100mm
Cheers
Mike
