Ever have a mod not go quite according to plan? Yeah, I’m having one of those nights 
I designed an aux board for the Wuben TO50R, which was my Peoples Choice Award prize in this year’s Old Lumens contest. The PCBs finally came in today. I built one up and it bench tested perfectly. Since I can’t reasonably reprogram the driver to control the aux board, I figured I’d use a JFET to turn the aux board off while the main LEDs are lit. Once voltage is detected at the gate pin of the JFET (wired to the main LED+), it would pinch off the ground connection of the aux board. Slick. And it worked great on the bench. Oh yeah, and it’s current controlled (3.0V LDO) and low voltage protected (a 2.93V reset monitor).
Well… when installed in the TO50R, apparently there’s always 5+ volts at the LED+ pad, its just a super low current that won’t turn on the LEDs. So my fancy new aux board wouldn’t light up until I disconnected the JFET and relegated it to being always-on. Bummer.
And then when I got it fully reassembled, I realized that I lost my moonlight mode. The flashlight thinks it’s on and in moonlight mode, but it doesn’t actually light up. It must have to do with where I tapped in to steal current for the aux board. The TO50R must be sensing/controlling current there. So hopefully I can remedy this aspect… I just need to take it all apart and find a new spot to steal battery voltage from. That’ll have to be another day.
Behold… the monstrosity of the TO50R driver. Triple stacked!
New aux board, fresh off the bench test
Selected placement of the aux+ connection (white wire)
Installed
First light
Behind the optic… thankfully, looks pretty good. At least I had one win for the night.
It’s a fun trek, but lots of stones on this path 
UPDATE: Eureka! So, the TO50R driver must have been monitoring current where I had tapped into the ground. I found a true ground connection and now moonlight works again! Everything is working like it should and the blue aux LEDs look beautiful under the optic. 