JB Kwik may be as good as the original, thermally, but it still sets too fast for me, especially if it's something complex or a lot of parts to do at once. With the original you can mix up a big batch and take your time. If you mix too much, or if you need to let one part cure before you add another part, you can stick the mixed epoxy in the freezer and it'll stay stable for a few days. Pull it out, let it warm up, and it's like you just mixed it 5 minutes ago.
And, JB Kwik doesn't come in the giant 5oz tubes. The Kwik is best saved for emergency/temporary stuff; if it's permanent, who cares what the curing time is.
+1 - I don't think epoxy will make any difference here. The AS5 should have worked. If it didn't, comfy is probably right. I've done this with a couple of HD 2010's that had the same problem.
Well, I backed off the pill a half turn and that seems to have done the trick. The hotspot on this torch is still horrible, even after centering the mcpcb. Someone suggested that I replace the mcpcb with a 20mm/U2 emitter and use thicker wires to make the light brighter. I wonder if doing this will improve the hotspot as well?
Good! That is a very common issue on these lights. Backing out the pill a 1/2 turn usually makes a big improvement in the measured throw as well. The hotspot look could be just the way it is - I don't think the stock HD 2010's hotspot is typically well defined, but still should be intense I hope.
I’ve always had very good luck simply using thermal compound. Don’t use so much that it oozes out the side but just enough for decent coverage. Then just put a couple tiny drops of super glue gel on opposite corners of the star. It’ll hold the star in place very solidly when cured. Also it can be pried off easily without damaging the star. And using pure thermal compound also means better thermal conductivity than a thermal adhesive.
Thermal pastes need high clamping loads to work, the superglue makes that impossible. You'd do better to skip the glue and let the reflector clamp the star to the pill. Just make sure when it's all stacked up the bezel doesn't bottom out against the head before the reflector/MCPCB are clamped down tight.
I would not us JB Weld for this… makes the star very hard to remove if you even need to. I use a silicone based thermal adhesive. You can easily use long nose pliers to break the bond by twisting the star off.
How often do you need to remove the star? JB Weld will tolerate the heat needed to reflow the LED with the star still glued to the pill, done it many times, blowing hot air in the underside of the pill and swapping LEDs. Not really any less convenient than soldering a copper star to a copper/brass pill.
It's still removable, fairly easily if part of the star overhangs one of the wire holes - a small punch and a sharp thwack with a screwdriver handle and the star pops right off.
Good spotting, thanks for the head’s up. I was just going to go place a bid then I thought maybe we should not get into a bidding war in the same auction amongst ourselves and raise the price so it ends up costing more than Fasttech. If anyone here is already one of the 2 bidders let me know, otherwise I’ll throw a bid in.
I never use Thermal Adhesive or any adhesive under the star always AS or similar. If I need the star to be fixed either screws or just Fujik around the edge of the star with compound under it.
Why does it look like that Fujik is just plain old white silicon...the stuff you can get at lowes
I remember I bonded a CPU heatsink down on an old PIII with clear silicon that I had left over from caulking a sink in...and it worked perfectly (of course a little dab will do ya)