Yes, I was going to mention that removing the 2 diodes from the PC board in the loco was not something I would have done. Those 2 diodes are not protecting anything. As Dave alluded to in his write-up, they provide a "virtual blue" voltage to the headlight. That is why the model only uses single (negative side) wire for the headlight.
This is basically how those diodes are wired on the loco (of course their "blue" output is connected to a logh bulb, no an LED, but the principle works the same.
As for the current limiting resistor, I seem to recall there might be 2 versions of the decoder and 2 different ways the negative side of headlight function output is designed. It's been a while since I last did the LED conversion, but I recall that in one instance I could not get the white LED to light up because there was only 1.5V across the positive and negative pad of the headlight. I had to modify the F0F output circuit on the decoder to get the voltage needed for the LED. I'll be servicing a Challenger soon, so I'll look into that. But if an extra resistor is desired, one side of a small SMD resistor can be soldered to one of the headlight solder pads up front on the PC board, and the headlight wire soldered directly to the other side
of the resistor.
I'm also not a fan of using thin solid-core enameled magnet wire for any connections that will flex (like up front where the pilot will move relative to the boiler, and the tender drawbar which pretty much constantly moves while the model is running. But I'm one of those people who over-engineers things, so my concerns might not worry others.
