While this is a different engine, details like knowing the same basic TRIAC circuit is powering that light output, we might find the fix here. https://ogrforum.ogaugerr.com/...rx-led-light-problem
Here is a key answer from that:
The 680 ohm resistor is required to give the triac enough "leakage" current to fire, it won't fire without some sort of resistive load. This is common and happens on the R2LC as well.
You can get AC out of the light output in conventional mode, or when the locomotive thinks it's in conventional mode, so you need the diode. I mistakenly omitted the diode in early LED conversions, and after a few months the LED's started dying, apparently from reverse voltage spikes in excess of their reverse voltage ratings.
You need both the resistor and the diode. Also, as GGG says, for some firmware you could reverse polarity on lights, I don't know if that applies to the LCRX as well.
Again, the bottom line, a load is needed to ensure the TRIAC triggers. That is the purpose of the 680 Ohm resistor across the output of the board, then, after that, comes the diode to prevent exceeding the PIV (Peak Inverse Voltage) of the LED, and then in series, the current dropping resistor (example 1K). So a complex parallel and series circuit. I'm sure somewhere I saw capacitors could also be used, but did not yet find that previous post reference. My own testing when messing with that C420 engine was that just capacitance from my fingers on touching contacts on the LED leads was enough to trigger the TRIAC and suddenly the LEDs lit up.
Here's my take on what is needed to both trigger the output with the 680Ohm load and protect the LED from PIV.
(+) -----------1001N1->-----1K ohm-----
680 ohm LED