So, while I wait for a more advanced BMS, I decided to build some Zener Diode battery regulators. I found a very nice writeup here (PDF instructions here) and followed it fairly closely.
First, you start with the components - from the left, they are:
- 10 Ohm 1W resistors
- 5/16" ring terminals
- 6.8V 5W zener diodes
- 1/2A miniature light bulbs
I soldered the resistors to the light bulbs:
I then soldered half of the zener diodes to ring terminals, making the positive terminal (the stripe goes toward the terminal for the positive terminal):
I then soldered the light bulb / resistor assembly to the diode / terminal assembly. Three of them have a wire in between due to physical layout issues with the rear battery rack:
I soldered the other half of the zener diodes to the negative terminals - the silver band goes *away* from the negative terminal:
And then I soldered a 16 AWG wire in between the positive and negative terminals, heat-shrink wrapping the ends for added strength:
And here are all 26 zener diode battery regulators. 9 of them got an epoxy / potting treatment (they are the ones that live in the front compartment, they need to be a little more water tight than the rest):
Here you see the regulators on the front rack:
And here they are on the rear rack. Note the three regulators on the batteries closest to you - these are the ones which I put extended length wires on so you could actually see them (otherwise, they would be hidden by the lip of the rear compartment):
With all the regulators in place, time to test them. I connected the CamLok pack disconnects, hooked up the charger, plugged it in, turned it on, and... light! I had previously charged the lower rear rack batteries, so the lights came on immediately:
The idea behind the light is that when the lights are on you should reduce the charging current. This was with a 5A charge current - pretty low - but until I get a more automatic system, what this means is that I will have to go watch the lights and turn the amps down on the charger manually when they start to come on.
The one thing I note is that having the regulators makes monitoring the individual battery charges with the PakTrakr kind of wacky. Sometimes the batteries would register very low (i.e. 8.6 volts). This effect goes away when the batteries are not charging - I suspect the extra shunt action between the positive and negative battery terminals confuses the poor PakTrakr.
All in all, a very successful week of work. I estimate that each regulator took about 20 minutes (there were 9 solder joints on each one), so this is about 10 or so hours of work all told.