200A PWM controller intermittent problems

[andidan]

So I got these two unnecessarily huge 200A PWM speed controllers. ($85 each, see my part overview here - ).
Details:
Using on a 55lb thrust trolling motor that uses max 50 amps. I oversized because I had a PWM quit on me 2 miles from the boat ramp and had to paddle. I assumed these were safe because of sheer size/Amperage.

Using DC House 100Ah LiFepPO4 battery. Battery not the problem. Hooked the trolling motor directly, no issues.

The PWM is in a box that has 2 exhaust fans on it. I have a temp probe in the box and it was max 33C with 31 ambient. Plus it has 2 fans on the MOSFETs heatsink.

Barley getting warm at 50A continuous. I tested with a 12v heater that pulled 48A for an hr continuously. No problem.
Problem - both PWM's have this weird problem: 3 out of 10 times, when I "accelerate" by rotating the pot, the current will shut off and restart in a second. And shut off and restart. At one point it did it 5 times in row. Could happen at 20A or at 40A. Then it works fine, once it's set on power, even if it's max. I managed to replicate it on the bench using 2 10A pumps. It's like a glitch - stop and restart almost immediately. On a boat you feel it. It's like braking abruptly and accelerating immediately.
What I did:
I changed the pot, changed the Fwd/Rev switch, checked all connections and wiring, and finally I checked with a smaller PWM (the small one in the video). Problem solved. Went all day on the lake, problem never appeared again. But I don't trust the small one because it warms up with continuous use.
The board has 2 MOSFETS for FWD and 2 for reverse, HY4008, rated at 200A. Also a relay for each direction rated at 200A. There are also 2 Schottky diodes, 1 for each channel, MBR30100PT. The diodes are rated at 30A. The problem occurs ONLY during acceleration, when there's a spike in Amps, and it's definitely over 30A. Thinking diodes are undersized, I replaced the 2 diodes with 60A equivalent (MBR60100PT), but the problem persists.
Now, is there any idea out there about what part of the PWM could create that current drop/restart? It would be a shame to throw away 2 potentially good industrial size PWM's if maybe there's a simple fix. I'm in contact with seller to get refunded, given the steep price. Thanks!

[m k]

It can be a program error of a PWM controller, but then it must be always there.

If its style is intermittent it most likely is a contact problem.
But it can also be inside a component.

The problem component doesn't need to be big.
It can also be an overcurrent feedback something, like a small resistor.