One of my fan fail customers sent me an extra OEM fan fail unit he had. What he found with his OEM fans were tripping his 40 amp fan breaker very quickly it would kick off then on a few second and keep repeating. He has ordered a clamp on DC current meter to read the real fan currents.
This is the first OEM fan fail I have had in my hands. I bench tested it and what it does is turn on the fan fail light when total fan current is below 20 amps. It does not compare the fan currents, inside the unit the two fan terminals are connected. The plastic case is very brittle (I guess from heat) so it shattered when I tried to open the unit. It looks like there is a reed relay in a fine wire coil that gets powered with the fan relay voltage input. Then there is four turns of heave gauge wire wrapped around that coil that carries the fan current and counteracts the fine coil.
I measured the voltage drop when both fans were powered drawing 10 amps each. One fan pin dropped 0.105 volts and the other fan pin dropped 0.080 volts. So that voltage drop is what produces the heating of the unit (1.85 watts) plus the fine coil current of 63 ma at 14 volts (0.88 watts). Of course when your fans draw more than 10 amps the heating increases.
My electronic fused fan fail drops 0.036 volts and 0.035 volts with the same 10 amp loads. My standby current is 10 ma. at 14 volts.