Because often dead center isn’t going to be center with the bezel tightened. I’m not sure why this occurs, but at times I need to push the mcpcb all the way to one side for perfect centering, it can be a total nightmare. Also, try applying downward pressure as you tighten to the side you want the reflector to move closer to.
1 Thank
I’ve been on that road before, and I don’t have a one-solution-for-all.
- check if the bottom of the reflector is flat and clean?
- don’t (over)tighten the screws. They are not meant to fixate the mcpcb, only to prevent it from rotating and snapping off the leads.
- fasten the bezel to “almost tight” and try to knock the reflector/mcpcb sideways in the direction you want it to go
- if all else fails: glue the gasket on the mcpcb and BEWARE of the fumes of the glue that might “fog” the reflector for ever.
2 Thanks
Thanks for all your suggestions and shared pain 
I don’t mind screwing up if I can at least learn something.
When I first got the light I opened it up, tightened the screws, cleaned it out a bit and centered it perfectly with a few tries. That’s why I suspect I broke something since it now is so hard to center. Maybe I was just incredibly lucky.
I eventually got the light perfectly centered yesterday. The emitter looks like it got hit by a hammer but still works. LES is undamaged.
I should’ve tried the bezel pressure trick! I moved the MCPCB around a lot, but I figure it must’ve been something else since I managed to perfectly center the light with the screws tight.
At one point I noticed some debris in the bottom of the reflector and thought I found the issue. Managed to center the LED, but could then not repeat it. I made sure the reflector bottom was as clean as possible.
I also tried a different bezel, O-ring, gasket etc. just to rule those out. I checked everywhere for debris or deformations.
I tried the reflector knocking thing and was easily able to center it but it would always skew in another direction when I actually tightened it.
Ever since I started to suspect flux fumes to be responsible for the fogging up of reflectors I became aware of not letting anything air out on them. With the gaskets I glued I blasted them with hot air for a while before installing them so they hopefully won’t release any fumes when they get hot.
I anyway wanted to try to swap this light to an NM1. This time I’ll glue the gasket for sure. I wonder how to do this best. I don’t want to glue it before putting in the new MCPCB since gaskets and glue don’t like >300C. I also don’t want to mess with fast bonding glue in the recessed head after installation. The chance of getting glue on the emitter or misaligning the gasket seems high. In the past I had success with double sided Kapton tape. I noticed however that under high pressure the glue part of the tape can squirt out! This has only happened on the L21B so far. I suspect the larger threads and the fact that the bezel / head never fully meet allows for much greater pressure to be applied. Have to be careful.
if all else fails: glue the gasket on the mcpcb and BEWARE of the fumes of the glue that might “fog” the reflector for ever.
Kapton tape is great for this.
Yep, but see my caution about pressure. Also keep in mind double sided Kapton tape is ~0.1mm thick.
Yea. Maybe don’t tighten them at all. Just have them installed to prevent MCPCB rotation.
If the optic pushes down on the MCPCB when the bezel is installed then the MCPCB does not need to be tightened with screws. The optic and centering gasket should center the LED unless the pocket for the MCPCB is not large enough to allow it to float enough. A sufficient amount of fresh thermal paste might help it float into place as well. Tightening the screws may cause problems with centering, causing radial force on the LED and optic.
MCPCB screws are never needed for better contact to the host? CPU coolers have high mounting pressure to ensure best thermal performance, I was thinking maybe that was at least part of their function in high power lights.
I was told years ago by a well known member here called Texas Ace, he said it’s okay if the reflector is placing enough pressure onto the mcpcb, but screws are better. I always screw them in tight, moreso because the lights are weapon mounted and I can’t risk them falling out. I’ve had units arrive with them rattling around inside.
The bezel IS a screw.
On a light where the bezel/pill and optic combination does not apply down force to the MCPCB, then screws are needed, or thermal adhesive.
The MCPCB screws are probably what is causing people problems with centering, although it could also be that the MCPCB pocket is not large enough to allow the MCPCB to float as much as is required.
The issue with the screws is that they are flat head screws. If you tighten them unevenly, the screw that is tightened more will push the MCPCB further in a certain direction. The taper of the screw is a wedge, and pushes the MCPCB away from the screw.
Even if you tighten the screws evenly, the manufacturing tolerance of the MCPCB, and how “on center” the LED reflow is, are additional factors at play.
A manufacturer could use a jig to center the LED dead center within the flashlight head, before tightening the MCPCB screws, and they could be using button head screws instead of flat head screws. Or the optic could be designed to float within the head. Or the MCPCB can float on center and be pressed down by the optic / bezel / pill combination rather than by MCPCB screws.
1 Thank
Decided to just buy Nylon screws and leave them loose, this way I don’t have to worry about a short circuit if they come out.
This 22mm 12A FET driver for the L21B keeps failing. The component shown keeps blowing and creates a short circuit, little bit dangerous. Simon uses this with the SBT90.2. It’s rated at 12A and I measured 22A, so I think this chip isn’t rated for the high current.
2 Thanks
