Yep, not going to turn battery switch off with engine running.
Back when I first got the hall effect dash ammeter I had the sensor measuring total alternator output, not like today where it islocated to measure amps into or out of battery. One day I was heavily discharged and driving intentionally keeping rpms up, and the 140 amp circuit breaker tripped after ~5 minutes of 65 to 120 alternator amps. This rather extreme load dump did not fry the diodes in the alternator and it worked fine for 4 more years once I reset the CB.
The regular parallel OEM charge circuit was disconnected for measuring the total alternator output so it was a true load dump. My meter is only rated for 100 amps, if I go over 100 amps it reads
---.-.
It was reading that at 14.1v+ at 2200 engine rpm when it went to -20 amps when the CB tripped well below its rating of 140 amps.
I bet the observation of slight discharging at float with heavy dc loads running, can be repeated without engine running. Put a 25 amp load on dc system with converter holding battery fully charged battery at 'perfect' prescribed float voltage, and the battery will still partially feed some of that 25 amp load and not remain at 100%SOC, but fall to some level slightly below that.
As I said its not much, and not really consequential, at least to me. Just an observation.
The way my rig is wired, my 1/2/both/OFF LOAD switch, is closer to my 1/2/both/OFF ignition switch, than it is to my battery and on equally thick cable, and no other wires, but the voltage sense line, goes to the battery terminals. No loads bypass the Ignition Switch or the LOAD switch to go directly to the battery. I doubt the voltage sense line's current is even measurable with any tools at my disposal, and the 20awg I am now using for it is way overkill.
My dashboard calibrate-able 00.00x voltmeter now has 20awg ground and voltage sense wires going to the battery itself well the ground still goes to my shunt's buss bar, but until recently, took a ground much more closely to the voltmeter. I employ a 10 turn potentiometer to adjust the alternator's voltage regulator, at a steady rpm it is pretty easy to dial battery voltage in to the hundredth of a volt, most of the time, exceptions being when it is near full and it bounces around a few hundredths easier.
Dash hall effect ammeter has drifted just recently, reading high, but this AGM continuing to taper to 0.0x amps at absorption or float when full is a constant. Same as on my previous NSB-27m until ~8 months before I removed it from the rig and relegated it to light cycling duty in the workshop. at end of Rv duty Amps would never taper to 0.4 and would bottom out over 4.0 amps then start rising again going as high as 13 amps if I was not there to lower voltage from 14.7v.
I kept using it until it struggled to start my engine fully charged on a warm day with a warm engine. Voltage retention on regular overnight discharges were as good as ever and it still is impressive in workshop in this regard. If I were to judge this battery's health
only by the voltage retainment on regular overnight discharges, I would be unaware of its decline. Which is surprising and unexpected, and impressive really.
Since being relegated to workshop and light cycling and a 100 watt panel facing south being the main recharging source, the amps will taper to ~1.4 before bottoming out and then rising when held at absorption, and improvement from when it was taken from my rig, where it was discharged deeper regularly.
Drifted far from the PD converter topic, apologies.
