Rewired: Part 2

Vero Beach
No LL

I finished the wiring project yesterday, cleaned up and tied up all the loose wire with wire ties and faseners. I also installed the 7 signal wires to the Link 10 monitor.

In testing, I did find one wiring error -- a potentially dangerous one. Here's the deails. My solar panel charge controller (a FlexCharge PV7D) claims to support two independent battery banks if wired as shown below.



I wired it up as directed. However, during testing, I discovered that both banks were connected in parallel on battery switch positions 1 and 2 and BOTH. What? I disconnected the solar + from the start battery, and that fixed the problem. Now, the two banks act as they should; either bank 1 or bank 2 or both in parallel.

The controller evidently does not isolate the two + outputs from each other as the manual suggests. That's a potentially dangerous error. If banks 1 and 2 ever became seriously unbalanced, then they would try to equalize themselves through that light gauge wire for the solar panel. That wire is far too small for heavy current. It would melt and perhaps start a fire.

The Link 10 is a marvelous toy for an engineer to play with. All it does is accurately measure battery voltage and current (0.1 ampere resolution) and time. From those three measurements, it extract maximum value via some very cool engineering analysis. I would have loved to have been the engineer who designed and programmed it. I'll write more on the Link 10 after a few days experience.

I don't have it calibrated yet but within the first hour the Link 10 exposed a real problem. Running my Honda generator, I powered up the shore power connection to charge the batteries. When the voltage got up to 14.0 volts, the Link 10 showed that there was still maximum (20 amps) of charge current going in to the batteries. It should taper off to 1-2 amps trickle charge at that voltage. I'll investigate. Meanwhile, it appears that I found at least one cause of overcharging of my batteries.