logic error while troubleshooting isolator charging

A few weeks ago I started to notice anomalies with alternator charging. It seems like a long time to put it off but…

  1. I wasn’t sure what I was seeing (is it normal behavior?)
  2. I don’t drive much so it doesn’t come up often
  3. copious solar is taking care of my needs

I suspected the Battery Doctor voltage-sensing relay (VSR) was dying after 3 years of earnest use.

In correct operation the VSR should close the circuit (click!) to tie the chassis and “house” electrical systems together when chassis-side voltage is >=13.4v. When the house bank is deeply cycled (like 50% DoD) the bank will pull ~44A until alternator voltage (Valt) is reached. At that point current acceptance tapers off due to both a reduced voltage delta and the nature of lead batteries in quasi-Absorption.

The combined circuit is maintained until the VSR detects voltage under 12.8v for at least 60 seconds, at which point it assumes alternator charging has ceased.

what I was seeing, or thought I was seeing

  • chassis voltage 14.1v per OBDii reader, normal for my van
  • house battery voltage in the twelves per shunted battery monitor, normal when solar isn’t charging the bank (I observed the issue when it was dark, then did the actual troubleshooting with solar panel array circuit breaker tripped so I would have simulated darkness)
  • VSR variously
    • failing to connect
    • connecting intermittently and unstably
    • failing to pass power when connected

My suspicions were more specific now, centered on the solenoid inside the VSR. I suspect burnt/broken contacts, insufficient movement, etc.

what I did

I took everything out of the van that was blocking access and removed the VSR. I put a plain continuous-duty solenoid in its place for now; I didn’t need a trigger wire since I was already using a high-voltage disconnect with the VSR to prevent backfeed high solar voltages to the chassis. A tweak ot the HVD settings would connect the solenoid at 13.4v and disconnect under 12.8v just like the VSR.. . Battery Doctor has a 5-yr warranty, so I figured I could ship it to them for evaluation and possible replacement. If they replaced it I’d toss it back in the mix and if they denied replacement I’d leave the solenoid in there.

There was no change in behavior. WTH? I traced the alternator positive (charging) wire back under the floorboard; the fuse was good and the connection was clean and tight. I verified starter battery voltage matched OBDii claims.

/me was confused

Finally I explicitly measured voltage between the alternator charging wire to the solenoid and house ground, which should have been 14.1v; it was 12.9v, which didn’t match either chassis or house voltage. What?

So I tested alternator wire to chassis ground and got 14.1v. Oh, I see.

what was really happening

The house-side chassis ground connection had loosened, with different impacts on each system.

house power

The house system is set up like this:

  • all “grounds” (negative returns) including battery bank -> ground bus -> chassis ground

The entire Everything-But-Isolator system was working fine because everything went back to a common ground, the ground bus. The house system could be floating in mid-air and not care about chassis grounding as long as the negative bus was working.

isolator charging

Isolators only connect the positive side; grounding is through the chassis.

  • isolator -> chassis ground

Isolator charging was wonky because it needed the chassis ground to measure voltage and run the internal solenoid itself.

the solution

I corrected the chassis grounding problem and the solenoid started working as expected. At this point I was tired and annoyed with myself but I yanked the solenoid and re-installed the VSR, which worked fine again. I closed the solar array breaker and let the batteries take their fill.

I put all the stuff back in the camper and vowed to troubleshoot more rigorously in the future.