I was messing about with a light, put it back together and now when I turn it on it turns on white and then quickly fades to blue.
It used to be white xD did I reverse the polarity or overheat the LED? I would have thought that a diode wouldn’t let any current through reversed so assuming overheated…
When you say ‘messing about with a light’, what did you actually do to it? What light is it? Sounds like it’s either being overdriven, or it’s not got a proper thermal path. Did you reflow the emitter? There might not be enough solder on the central pad. Did you lift the mcpcb off the shelf? There might not be enough thermal paste under it etc.
You transplanted a led and a driver. Or did you transplant the whole pill, driver and led included.
Be careful when you transplant things from an S2+ to an S2. The threads inside the S2+ head go deeper than the threads inside the head of the S2. The pill pushes the ledboard to the reflector to the very end of the head. Inside the head of an S2 the pill might run into the end of the threads before it starts pushing the leadboard.
It’s always best to see how the MCPCB seats to the pill —you’d be amazed how warped they are sometimes —-when using thermal past I always try to sand them smooth on a flat surface — there might be a big hump in the pill where that hole is in the center also
If your LED is turning blue, you’re possibly damaging the LED. Turn it off immediately and fix the problem.
I think the problem is probably too much current, caused by a short bypassing the driver.
The 219B can only take so much current. Run it on direct drive on an 18650 and the LED will act as you describe: start out white then turn rapidly blue.
The most likely cause is a short between the negative black wire on the top of the driver and the outer ground ring on the driver. This bypasses the driver completely. The result is when you turn on the light it goes instantly to direct drive and there are no modes.
I recommend removing the driver and double-checking your solder blob where the negative LED wire is soldered to the star. Make sure that it is not bridging to the outer ground ring of the driver and that the blob won’t touch the pill when the driver is inserted into the pill.
Also check the negative (black) wire where it is soldered to the star. Make sure that the wire is securely on the bondpad and not draping over the side of the star where it could cause a short with the edge of the star.
Another possibility is a short between the center ground pad under the LED and the adjacent negative bondpad. If you didn’t reflow the emitter this isn’t too likely, but it can happen. For example, if there is some downward pressure on the LED and the heatsinking under the star is not sufficient, it might get hot enough to melt the solder under the LED when you turn on the light. If that happens, the solder might squish together and cause a short.
If the problem is not enough heat transfer to the star, usually I’ll know when a driver wire on the star desolders itself.
In this case, the light won’t turn blue… it will just stop working. When I open it up it is instantly apparent that one of the driver wires is no longer connected to the star.
The usual fix is to simply apply more thermal grease. Arctic Silver 5 is wonderful stuff. Normally you want a thin layer, but if an uneven surface is below the star you can add more below or around the edges of the star. I’ve found that even direct drive FET triples will work just fine on hollow pills if I apply enough AS5 around the edges. And it’s working fine if the driver wires do not desolder themselves.
In my experience usually emitters don’t change color immediately to blue unless too much current is applied. Usually heat takes more than a few seconds to cause damage. However, one possibility where heat might become an immediate issue is if there is solder under the 2 contact pads on the underside of the LED, but not enough under the central heat sink pad. If the central heatsink pad is not connected at all, the LED might well overheat almost instantly, without causing the wires from the driver to desolder first.
Disclaimer: I’ve had the issue of other lights desoldering themselves if not enough heatsinking between the star and pill with no damage to the LED. However, I have not experienced this in a single emitter 219B light so am not positive the 219B will behave this way.
