H17Fx driver loses low mode

[Duster1671]

I recently modified a Lumens Factory KL4 head by swapping the 319A emitter for a XP-L HI and swapping the stock 3-mode driver for a H17Fx.

After doing a runtime test I noticed that when the light heats up, I lose moonlight mode. I can keep clicking through to the higher modes, but moonlight doesn't light up.

Any idea why this might be?

 

[Jose Marin]

Funny i just got done modding a lumens factory e series tear drop head with a white 2 and h17fx. Ive installed probably around 25 h17fx in different lights and drop ins, ive probably had 2 drivers fail but it was mainly from flickering.

Ive had some from new not have the first 4 low modes when you go to programming, not sure why it does that and it does it with single/ triple and quad.

I cant say that ive had a driver temporarily lose a low mode from over heating, i also dont run them very long on turbo or really even more than 1.5 amps.

It probably doesn't help that these little p60 like heat sinks cant dissipate heat efficiently but if i had to guess youre probably just saturating the driver with too much heat which ive had other fet drivers act funny in small builds like convoy s2 because of heat.

Not sure if that kl4 head has a retaining ring or a solder in pill but if you had to solder the driver into the pill maybe add some more in spots so that heat can disipate better. Also im not sure if that head had a 10mm mcpcb like mine did but i swapped mine out with a 16mm and filled it down some to fit into the pill to help add mass to the pill

If you have another build with the same driver try heating it up and see if that light does it too or turn down that high mode to 1 amp and just do bursts of direct drive.

Hope that helps some

I recently modified a Lumens Factory KL4 head by swapping the 319A emitter for a XP-L HI and swapping the stock 3-mode driver for a H17Fx.

After doing a runtime test I noticed that when the light heats up, I lose moonlight mode. I can keep clicking through to the higher modes, but moonlight doesn't light up.

Any idea why this might be?

 

[Duster1671]

Thanks for the info!

Interestingly, this driver does not seem to have the lowest few modes either. When I go into programming and ramp through the brightness levels, I only count 20 or 22. So I guess the lowest mode, the one that's dropping out, is actually ~4 or 5.

I tried swapping the driver out for another H17Fx (I bought two together) and the second one won't go into programming at all.

I'm replacing the driver in this head because the Lumens Factory driver loses the lowest two modes and gets stuck in high when it gets too warm. I figured an H17 would fix the issue, but I guess not.

 

[Jose Marin]

Sometimes when tighting down reflectors you will get a short circuit from the mcpcb to the reflector. It will get locked in a random high mode and won't go into programming.

Before junking that 17 dollar driver make sure your solder job on the mcpcb wires aren't overly big or touching anything. Also if you don't have any insulating rings try cutting little bits of electrical tape to put over the solder connections.

 

[desert.snake]

It's possible due to the characteristics of the LEDs.

I put several different triple diodes on the same H17f, then I had to reconfigure the modes, since it could ignite a Nichia in mode 2, xpl hi only in 6, and xpl hi from the same batch and a same color temperature ignited as it should be in mode 1.

Nichia has a breakdown by voltage classes in the datasheet, I did not observe this in Cree (or did not notice):

That is, for example, class L1 will ignite at lower levels than class M2, if someone in the store scattered the coils and then put everything in one box.

Well, plus there may be some variation in the characteristics of individual instances of drivers. I once came across a driver that turned off several lower modes as soon as the ambient temperature dropped to ~ 0 Celsius. So it all turns into a lottery.

It may also be that materials expand when heated, there may be poor contact somewhere, when it heats up, the gap increases slightly and at a low voltage level is not enough to break through this gap, this does not affect strong modes.

 

[Duster1671]

Hmm, the LED not lighting up because of insufficient forward voltage would be consistent with what I'm seeing. And it seems plausible that the threshold could change with temperature.

I "solved" this by setting the first mode to the lowest level that still lights up after running on turbo for 30-45 seconds. This ended up being the fourth brightness from the bottom, or maybe level 7 or 8/24, assuming the first few levels don't light up at all even at room temperature.

On that note...when I enter programming and do 3 clicks to change brightness settings, there's a pretty long pause (3-4 seconds) before the lowest level turns on. Is this normal, or does that indicate that it's going through a few moonlight modes before reaching one that lights the LED?

If I count the levels I get up to 21 or 22 before the brightness maxes out and starts going down again. I thought that there might be too much resistance in this light to get the top couple of levels, but maybe it's that I'm not getting the bottom levels instead.

 

[desert.snake]

On that note...when I enter programming and do 3 clicks to change brightness settings, there's a pretty long pause (3-4 seconds) before the lowest level turns on. Is this normal, or does that indicate that it's going through a few moonlight modes before reaching one that lights the LED?

It seems that it is, I took the alpha, in which I put h17f and decided to look at the beginning of the modes, it seems that 1 or 2 pcs in this case does not light up for me, and one of them lights up very weakly.

 

[sween1911]

Oooh, I'm watching this one. I love the H17Fx, I have one in my 18650-bored 6P with an XM-L2. I have not seen the behavior you describe.

I do notice however that the ultra-low modes throw off the count if you're trying to keep track while ramping brightness in programming mode. There does seem to be a delay.

 

[Duster1671]

Oooh, I'm watching this one. I love the H17Fx, I have one in my 18650-bored 6P with an XM-L2. I have not seen the behavior you describe.

I do notice however that the ultra-low modes throw off the count if you're trying to keep track while ramping brightness in programming mode. There does seem to be a delay.

When you ramp in programming mode, do you count a full 24 visually-distinct brightness levels?

 

[sween1911]

When you ramp in programming mode, do you count a full 24 visually-distinct brightness levels?

I'll try this later. I know I've tried to count them as they go and I've lost count but maybe I'm not good at counting rapid flashes of light.

EDIT: Ok, I tried it a couple times up and back. Besides blinding myself and cramping my finger, I get to 23 distinct levels. I suspect the two lowest ones are very difficult to visually distinguish.

 

[Sean]

I just got a (triple 219c 4K) Laulima Ion with an H17fx driver and twice now I've lost the lowest brightness level. The first time I went back and reprogrammed it. The next day it happened again and the lowest level is no longer available. Like the driver voltage isn't high enough. I will say that the low level I lost was really low. Like small tritium vial low. Strange. Maybe I didn't lose it, maybe it just can't drive the LEDs that low now for some reason.

 

[Sean]

Well I tried more batteries and what I discovered is that when the battery starts to run down then the lowest low mode gets much dimmer. When you drop a fresh charged cell in then the lowest low mode is much brighter (but still dimmer than the next level up). Also, I didn't notice any other modes being affected.