Huge thank you for this video! Looking forward to to doing my first emitter swap and now I feel (relatively) confident. 
Hi there :)…
I had “Curious cases of Benjamin Button” recently with newest batch of noctigons. Symptoms were that after re flowing everything was tested and works flawlessly(mode selection and everything) and when you assemble it to flashlight I got High mode only :smiling_imp:
So atiny is trying to change modes(i can hear high pitch noise) but high mode still remains…
That happened around 5 times out of 50 emitters so I guess it could be that what Mitko mentioned.
I did not try yet but I think electrical insulating thermal glue like AA or Fujik(if it really must be used) can resolve this issues if they ever happen to you.
Indeed, sometimes you wont be having a short connection but just a modes lost( it stucks on high), or like on some of my drivers offtime memory function disabled
That is so irritating… A lot of effort to assemble everything and than wtf happens! 
Mitko what is your method of correcting this issue?
I’ve had a similar result when there is power bleeding off to ground to led negative on the driver board. High mode always.
If there is a short of some kind in the MCPCB the best solution would be to just dump the mcpcb and use another one. Trying to prevent a short with thermal epoxy would likely result in having to use a layer so thick it would negatively effect the thermal path.
No first version is not for sure. Triple checked that.
It is pita to reflow dedomed emitter on new mcbcp since I can’t use tap from above cause I’ll damage perfect emitter.
Great video. I guess on a chemically dedomed LED you don’t tap the emitter. How do you remove the excess solder then?
You reflow them before dedome 
OK….
Thanks.
So you would solder mcpcb with domed emitter to flashlight and then you would de solder it, de dome it and then solder again?
Nah… This is not right technique.
I usually reflow around 10-20 emitters and check them on my test rig and then de dome all of them at the same time so there are ready to go.
We are talking about curious cases when everything works on testing rigs and yet it does not work when assembled in flashlight.
Edit:
I just saw you were not replying to me. But it is ok lol 
Um, no… Solder the LED to the mcpcb you want and then dedome it.
Where did you get the other part from?..
Edit: Saw your edit
Mitko and myself just want to point on this:
Wow that is great, thank you!
I particularly appreciate the intentionally casual technique, and lack of precision shown, to highlight the effect of surface tension with the wetted solder.
Very illuminating video, I had no idea that the leds could tolerate that duration of time at the solder’s melting point, or that thermal shock was the greater issue. I would have hurried it and had poor joints and raised, heat shocked emitters if I’d tried it on my own experimentally.
This is something that I’ve been putting off learning to do, I’m now quite inspired.
Thanks again.
If you reflow an LED try to stay within the heat up and cool down speed in the reflow diagram, too fast is bad for the LEDs lifetime
So the vertical side of the T is the - (right) and the horizontal side is the + (left).
How can I know the + and - of the PCB?
Hi Zoomie,
Most mcpbs are marked with + and - near the pads.
You can also verify with a multimeter by checking continuity.
Is an iron hot enough?
I have a 60W adjustable iron and it is perfectly fine for reflowing. My smaller 45W also works fine.
That’s great. Never done this before, but it looks easier than with a skillet.
Any advise about the temp setting? Cotton?