[Solved] Flickering occur at high temperature.

Ouchyfoot,
I can’t guarantee that I will find the root cause in any time soon, but I will keep trying. :slight_smile:

Richard,
I see. So was the LED poorly reflowed or it has been damaged?

Do you have a bench top power supply, one that can push “clean” current to the driver

I too have a brand spanking new 101-AK-A1 driver I put in a Ultrafire 602C, using a crappy UF blue battery it flickers

I am beginning to think either the batteries can’t push clean power to the driver (at sustainable volts above 3.2vdc or perhaps there are bad 7135’s on the board (maybe the manufacturer pumped out a bunch of bum chips ?)

I'm not sure what was wrong with it. I think that it was a faulty/damaged LED, but I'm not sure what caused it. It could have been the reflow, but it could have also been something else.

Only other thing I have found googling is possibly bad solder joints on reflow from the factory

quote from what I found

WarHawk-AVG, thanks for your suggestion, I get your point but I don’t have any bench power supply. By the way I used unprotected Panasonic PF for my test and it flickers. I don’t know about the factory soldering quality but as for my part I did all the soldering very carefully and tried to make them as neat as possible. It’s complicated…

Richard, alright thanks for your input. I will put my focus towards the LED tonight. :slight_smile:

Have an oven? 200C for 5 minutes. :slight_smile:

(Only the driver, of course.)

You are joking, right? :open_mouth:

I don’t think its possible to get ten drivers with a faulty chip. I’m putting my money on the emitter.

People do it all the time, especially for PC components. It’s a real reflow setup.

Normally 8-10 minutes but driver is really small - it won’t soak much heat so I think 5 minutes will do.

If everything else fails, what have you got to lose - you don’t want the flickering driver anyway. Might as well try a salvage. But of course, IF everything else fails. This can be your last resort.

I see, wouldn’t this method damage the driver? I know some members here using the oven to reflow the LED and it works.

I wouldn't try and reflow the drivers, we already know that there couldn't possibly be (almost impossible, anyways) 10 bad drivers. I'm betting on the LED.

Ten!?? Dang, I missed that part :open_mouth:

All ten tested using same emitter, bibihang?

Yeah I have modded at least more than 6 or 7 Qlite drivers (didn’t really count, I leave them everywhere on my table and it is messy!) by stacking chips and most of them have flickering issue. And yes for the most recent 3 pieces of Qlite that I am working on I tested 3 of them with the same LED and ALL have the flickering issue.

Here are the symptoms:

  1. When the LED was ran on high for a certain period of time until it was very hot, and then switching it into low or medium mode the light would flickers.
  2. If the LED wasn’t hot enough it didn’t flicker on low and medium mode.

By the way, last night I was quite busy and didn’t manage to reflow the XM-L2 on Noctigon again, but I did try to hook up the Stock Qlite driver with a different LED, a XM-L on aluminum pad (salvaged from a stock light) and tested. This stock Qlite driver is exactly one of the drivers which are related to the flickers of the previous XM-L2 LED that I was working with. I cycled through all the 3 modes repeatedly for more than 5 minutes until the pill got very hot and I did not notice any flicker at all in both low and medium modes. I took a picture of my test setup in last night for you all to refer as below.

So:
The same driver vs. XM-L2 on Noctigon (reflowed by myself) = flicker
The same driver vs. XM-L on alu. pad (different LED, not reflowed by myself) = no flicker
No chip was stacked at all, it is still a complete stock driver.

Perhaps we are one step closer now, but I don’t want to get into conclusion yet as this problem is not that simple (at least to me) and more tests need to be carried out, also the XM-L2 on Noctigon needs to be reflowed again and tested.

Any opinions regarding this?

Yeah. It’s your emitter, or not a good reflow.

I see, maybe you are right from what I observed. :slight_smile:

But please bear with my ignorance, that I thought reflowing LED is nothing more than just soldering the bare LED onto the pad? The way I reflow is putting some solder on the +ve, -ve and the center pad, then heat up the entire pad until the solder were melted and put the LED on it and slightly pressing it down to push out the excessive solder. And note that I use solder, not soldering paste as most people do as I think my soldering paste is kind of low quality so I don’t bother using it.

Is there any possible issues occur during the reflow process? Or anything to take note of?

I wish I had seen this thread sooner. Yes, it's most likely the LED. If the center pad has any air gap, when the led is reflowed, then when the star heats up, the led becomes too hot and flickers. If it continues, it will fry the led. I thought maybe you got one already on a Noctigon. I have seen many, (15%+) of the leds already on noctigons have this problem when I get them. I have one right now, in the 3-MT-G2 light I just did and I have to tear it down to reflow all 3 of the MT-G2s on the noctigons.

Any more, when I reflow any leds, I always use something to push down on the led dome, (pushing very lightly), when the solder is still molten. I press and hold down till the solder cools. I also do that with all the leds I get now, seeing as lately I am seeing more and more of them come in bad to start with. I would hazard a guess that it has always been this way, but we didn't see it very often, because of lower amperages, which means less heat.

Just my usual 2 cents worth, LOL.

That’s some great explanations right there Old Lumens, except that the XM-L2 were reflowed by myself, not already on the Noctigon when it came LOL. I never know that an overheated (or poorly heatsinked) LED can flicker due to that. I remember I once overdriving a XR-E long time ago and it turns its beam colour into blue and having slight crack on the die, but it still never flicker. Guess it could be not putting enough solder on the center pad when I reflowed, anyway I will reflow the LED again and hopefully this head-scratching problem will end soon. :slight_smile:

I recall that you also use solder to reflow rather than soldering paste, right?

I saw that one guy who makes the really high power hotrods youtube videos, when he reflows, he bops the emitter on the dome with his finger while still hot, vin something something

I use Rosin Flux and 60/40 solder, because, for me, it’s easier and flows much better than paste. It’s just the way I learned, many years ago, so I do it that way.

I flux the pads on the star and on the led, then I flow solder on the star pads and sit the led on and heat from underneath with the iron, till the solder melts. Then I press lightly on the dome and hold till the solder cools. The easiest way for me to tell, is to also solder the +- pads and then when I heat from underneath, they turn shiny and when it cools, they turn dull.

Just did some tests again a moment ago. I still using the same XM-L2 T6 3C for the test but I reflowed it again on the Noctigon with more care.

Here are the results:


I think you guys are right, that it’s the LED problem. However since I’ve reflowed it again and the LED N is still flickering so I think the LED itself has been damaged. I guess during my first attempt to reflow it in several weeks ago the LED was damaged by excessive heat or some other reasons, which I don’t remember. The FT driver which was added with 4 extra chips was suspected as faulty by me, now it doesn’t looks like so as it works well with the LED S (good news isn’t it?).

Next time I will test these drivers again with another new XM-L2 LED. I wanted to test with the other drivers but it’s 3.30am now. :frowning: