I seem to have an IO EDC 18650 that just does not want to be built!
I’ll spare you the sordid details, but it looks like I may have gotten a 16mm Noctigon + XML2-T6 3A that is NOT electrically isolated from the MCPCB. Has this happened to anyone else?
I’m so frustrated at the moment I had to walk away from it. I’ll remove it and test it after I’ve settled down a bit.
If it’s not isolated, I’ll reflow it or put it on another MCPCB and see if I can salvage the LED (it’s the only decent XML2 I have left)…
Guess I may need to start checking for this up front.
When I reflow my own, I always check continuity. In the beginning, I had some problems with the solder paste wicking along the edge, inter-connecting the + or - pad to the ground pad. After the solder paste melts, I give it a few more secs/temp to settle more - think that avoided the problem thereafter.
With bought LED/Noctigons, I haven't done continuity tests but I inspect them thoroughly. Maybe I should start continuity testing them though...
That one you got is scary! Hank's been pretty good with QA, but he's not automated like Cutter is, I don't believe.
First, I thought the neg wire was getting shorted against the pill when I tightened bezel. The noctigon was getting rotated by the reflector when tightening, so I added two #0-80 screws to fasten it to the pill (I knew I kept that old 90’s flip cell phone for some reason).
Still, the light would only illuminate at one brightness level when assembled.
Then, I discovered that it works fine if I loosened the screws, and had no pressure on the Noctigon against the pill.
I’ve noticed that many of the Noctigons I’ve gotton from Hank have had “solder balls” sticking out from under the LED that prevent centering spacers from seating properly, but this particular LED didn’t have them.
The LED on the lower left looks like its rotated 90 degrees to the left of how its supposed to be oriented. This would result in a short as the LED’s pads are perpendicular to the Noctigon’s pads.
Actually, I'd rather have the solder balls than not. They show there's enough solder on the pads and it's heated enough, rather than too little. I use a razor blade or xacto knife to cut them off - but cut them from the side, not the top. Cutting from the top may cut through the trace (I did that once).
For the rotated MCPCB problem, few solutions: screws as you did, or thermal epoxy (Arctic 2 part epoxy, etc.). Also, could use shrink wrap on the silicon wires for the sections going thru the holes - adds more protection because the silicon covering cuts easy. Which method to use - depends. I try to avoid thermal epoxy - my last choice.
The problem was with the driver, and it was my fault.
I managed to overlook a very tiny solder bridge created while removing a few of the 7135 chips from the bottom side of the Qlite (yes, I know - most everyone else is adding chips rather than removing them).
Funny how it never presented a problem when bench testing - only when I assembled the light.
Lesson learned: Carefully inspect all work performed with good illumination under magnification (especially with my less than stellar eyesight).
