question

Probably something(s) bypassing the shunt. Are there any connections to battery negative except the shunt?

There aren't any connections that bypass the shunt, everything goes through the monitor.

There is almost no load on the system (except for a few minor leds and background systems that can't be easily turned off, but don't consume even 10W in total)

How sure are you about this 10W estimate?

It is your DC system that consumes power. You have also losses when charging (charge efficiency factor, peukert...) Therefore you have 2 different monitoring points, on the supply side and on the battery side, they will not match, because of real world energy transformations and usage.

Hello @AJR . That is quite a conundrum. We can see that 6 watts is dissipated as heat in the solar regulator.

26.15. X 5.4 = 142W. 13.48 x 10.1 = 136W

We can also see the voltages readings of the MPPT controller and the shunt are very close, 13.48 versus 13.46.

It is the current reading of the shunt and the controller that differ significantly. It would be nice to put a DC current clamp meter on the battery wire of the MPPT controller and see if it really is 10.1 amps. About three amps is being lost somewhere. It is important to turn all DC loads off or you will never know if one of those loads is taking your 3 amps.

It may be helpful to disconnect the load side of the shunt and perform a zero calibration. Failing that being out, if all loads are disconnected, one of these devices is providing the wrong reading which can be confirmed with a DC clamp meter.

My guess is the light loads are not quite as light as first thought but this can be confirmed with a simple DC clamp meter.

Good luck.

I'll try the zero calibration and report back here if that makes any difference (I don't suspect that it will since I've done that already), if not I'll get a clamp meter out and see if I can track down the source. Thanks for the input, I'll report back to this thread if/when I locate the source.