Unexpected Behavior of my system.

Your solar array + MPPT charge controller can output upwards of 50 Amps on a clear/cool day with a good size load on the battery bank (and/or if battery bank is significantly discharged).

For US/North America circuit breakers, I use a 1/0.80 (1.25x) derating factor for continuous loads. Circuit breakers are designed to break at 100% load (may take minutes to hours), and not break at 80% or less load...

For a 50 amp branch circuit, I would design the wiring and breaker for: 50 amps * 1.25 NEC derating for continuous current = 62.5 amps

And I would "round up" to the next standard size (i.e., 75 or 80 breaker+wire rating).

To confirm what was typed above... The electrons will flow from where they are sourced (battery bank, charge controller, etc.) to the load(s) (AC inverter, other DC loads).

When you run heavy DC loads, the charge controller should be the "first source" of energy (i.e., you want the controller outputting 50 amps, if there is enough sun, and 35 amps from the battery bank... Vs zero from the charge controller and 85 amps from the battery bank (battery bank should only supply all current at night/during bad weather).

-Bill

Another point to ponder...

This is an RV? And you have the RV "house battery negative bus" grounded to the frame of the RV? Correct?

If so, then you should not have any circuit breakers in the Negative wiring.

First, it is not needed. With a negative to ground bonded power system, the negative wiring never gets above zero volts (maybe a few tenths of volt). So there is never any worry about a negative to ground short circuit (no breaker needed).

The other reason to not use a breaker in the negative wiring (and code does not allow you to put breakers/fuses in ground bonded neutral power systems like North American home wiring)--If the negative breaker trips from over current, it does not shut off power to the devices (loads). But leaves than at +bus voltage (or +120/240 VAC for house wiring) and sets a trap for somebody working on the system (breaker off, must be no voltage in the system).

For certain power systems (large ships, floating or isolated power systems), they do put breakers in the +/- (or the two outputs)--But they use "ganged breakers" that if one side trips, the other breaker is turned off at the same time (like the 2 pole 240 VAC breakers we use in North America).

So, at the very least, you will save money on breakers... You will only need 1/2 the number as you only use them in your + Busing.

Newer DC breakers (at least those NEC/UL/NRTL rated) are supposed not to have polarity (designed to break current flow in either direction). And many "unknown" certified breakers are designed to not interrupt very large amounts of current.

For example, North American Home breakers are rated to 10,000 Amps Interrupt Current. That is the estimated/designed maximum current from a pole top transformer and down lead to the home.

When doing a "real engineering" design system, you need to look at the amount of current available in the case of a dead short.... A 12 volt battery bank for off grid solar can output a 1,000 Amps or more into a dead short. But the wiring resistance may limit the real short circuit current... (12 volts/1,000 amps=0.012 Ohms of wiring resistance--That is pretty low resistance/heavy copper wiring).

Not your brand/model of breaker (I do not know what it is)... But here is an example of what a spec. sheet can look like. Here is a Bussman marine breaker, similar to yours, specification (Northern Arizona Wind & Sun--shameless plug for our host):

https://www.solar-electric.com/mrcb-200-amp-marine-circuit-breaker.html

And the spec. sheet:

https://www.solar-electric.com/lib/wind-sun/Bussmann-MRCB.pdf

Interrupt Rating: Main Breaker Protection Interrupt Rating (5,000A@ 14Vdc, 3,000A@ 28Vdc and 1,500A@ 48Vdc).

And from the graph, it will (almost?) trip at ~150 to ~600+ seconds at 100% rated current at 70F (there is a + tolerance band too).

Since these are thermal breakers (rather than magnetic current trip), they also have a temperature trip curve... 150% to 50% of rated current (from -50F to +250F, including the "tolerance band"). It is much easier to design with a bit of margin of error/conservative (i.e., fudge factor with 1.25x heavier wire and breaker larger than expected current) than to run much of this stuff "at rating" and assume that everything is OK.

We are very used to plugging stuff into our home and everything works... But most of those items draw much less current than the circuit's rated capacity. But when we put loads on the system that are at the very margins of rated capacity, we hit problems...

For example, I take care of a small apartment building with electric hot water. The elements draw around 20 amps @ 240 VAC. They were built ~60 years ago, and the breaker panel had 20 amps breakers and 14 AWG wiring.

Not sure why, but the NEC table lists 20 Amps as OK for 14 AWG in the table but has a footnote that says not to use more than 15 amp breaker on 14 AWG wire... Anyway, 50-60 years later, on occasion a breaker will false trip for the water heater. And eventually, the false trip happens more and more often. I had the electrician pull out the 14 AWG wire (wiring looked fine, no signs of overheating), and replace it with 10 AWG wiring and a  30 Amp breaker (20 amps * 1.25 NEC derate = 25 amps). 12 AWG is limited to 20 amps, so I needed 10 AWG. Needed another breaker, so uprate to 30 amps. Now, no more false trips, and no more failing breakers over time (I hope). Interestingly, there are 4 units and only one unit had this problems where I needed to replace the 20 amp breaker....

https://lugsdirect.com/WireCurrentAmpacitiesNEC-Table-301-16.htm

-Bill "that boring engineer that types too much and still makes mistakes" B.

Here is the Midnite Breaker Spec. Sheet (made by Carling Switch?):

https://s2.solacity.com/docs/Midnite/Carling_F_Series.pdf.pagespeed.ce.0FVFif0Y8O.pdf
More information here:

https://www.solacity.com/product/midnite-solar-mnedc250/

Breaker installed correctly and reverse polarity:

https://www.youtube.com/watch?v=Cup5fMGaE2g

-Bill

Oh, I forgot to add why to never install a breaker/switch/fuse in the ground/return of a ground reference power system...

Especially, in the olden days, we had lots of ground reference communications interfaces (RS 232 was very common). Typically, these comm signals were referenced to ground/DC negative power lead. If you popped the DC negative power lead with a breaker, then the "ground reference" inside the communications device would "see" voltage and current on the ground lines and typically pop a trace on a connected device.

Grounds that were "hot" (or at least not near zero volts) between two grounds (in a data center, telephone central office, etc.), we would see lots of Wyse Terminals with blown RS 232 comm ports (the ground trace was fused).

There are some solar communication systems (controller to inverter to combo device/laptop, etc.) that could be damaged by "lifted" DC negative connections back to the battery bank.

-Bill