Hi,
The choke is a simple ferrite ( not metal) e-core. It has 3.5 turns of heavy wire. It is an E65 core ( ebay?) It is simply connected in series with either of the wires going from the small heat sinks to the transformer.. it is in series with either lead...... very simple really.
The shut down is probably the mppt controller upsetting the unit. This can be fixed by bypassing a resistor on the board....
First ascertain this is the problem by running the motor without the mppt connected... if it runs indefinitely, then thats the problem.
It is sensing the high voltage currents from the mppt. A meter will not show them, and will read 26-27v.... but an oscilloscope will show that there are very high current pulses that average out against the low impedance of the battery as the 26v you see.
It usually happens when there is plenty of sun, and fuller batteries... then the pules will be added up by the chip, and it spits the dummy as high battery voltage fault.
I have not been successful in damping this ( haven't tried very hard either ), I simply put a 5v6 zenner across the" cut if low resistor" (R14) you will see near the chip.... this will stop the unit seeing a high voltage battery problem..... and your mppt controller will look after the battery for that situation anyway.
Note, it will also mess up the charging voltage of the unit, and it won't know when to stop charging.... but I don't recommend using the charging function as a rule, it can destroy the unit without modification anyway, as the software was written for lossy e core transformers with high leakage. when it is cut off from the grid or generator, the sine wave reverts to 50 hz, and can catch the transformer in the wrong magnetic circumstance ( near zero crossing is bad). In the lossy transformer, this transient will be swamped buy the fets, ...... but with torroids, they will likely kill the fets with the massive inrush currents. If you do the transformer mods to the inspire units, this is exacerbated further.... they are very tight, and have very low leakage, so the kick back will almost always kill the fets... and the drivers follow.... some 60 times the magnetising currents for a few cycles.
I have not found any load that annoys the unit as an inverter.. they are very tough in that regard.
john