First attempt there was no moon, it was set too low. Second attempt gives me 7 levels, low isn’t really all that low, but then the spring (OEM from Crelant) softened and everything got out of whack and I’m still trying to get it working again. Not a board problem, a mount issue.
Taking D to McD’s for a break, will hit it again later.
I got 1.86A at the tail initially, from a pair of Sanyo UR18650FJ’s that arent fully charged. Something like 1480 lumens. Back later, Ciao!
1.86A at the tail equates to what at the LED? I've never been any good at working that out. It should be close to 4A. It's specced to 4A, but there will always be losses that reduce the output.
- Matt
I’ve managed to kill the electronic switch, which was about half of the whole point to using this driver to keep the light working like it was designed to work. This side switch doesn’t have any notches or any other way to get a hold of the SS bezel/insert that presses the rubber boot into the head. It wouldn’t budge. So with no way to grab it I assumed it was press fit. Tried to knock it out. And broke the pcb the electronic switch was on. This was necessary as the tiny wires soldered to the pcb broke off and there was no way to access the pcb inside the host to reattach them.
At any rate, I will now be making a plug to seal the side switch up and rely on the tail clicky to run this driver. Aggravating as green snot.
Which crelant light?
Couple ways I’ve considered attacking stuck switch rings, drilling small holes on the inside of the ring that you can hook into to turn it or filing the sides of the switch ring to create flats for pilers to turn.
Sorry this happened while you were experimenting with one of my drivers :(
With 100% efficiency a 4A drive current on an XM-L2 would pull 1.86A if the batteries were at 3.87v under load. With a 90% efficiency a 4A drive current would pull 1.86A if the batteries were a touch over 4.3v under load (not possible in this scenario). I’d say he’s not getting a 4A drive current.
I’m basing that on djozz’s chart where the XM-L2 needs 3.6V at 4A which is 14.4w.
If DBCstm’s cells were just slightly discharged we can assume a best case scenario of about 3.9v (loaded). 3.9v*2=7.8v, *1.86A is 14.508W. If we assume a 90% efficiency (I assume that’s generous but don’t know what the best case for LED2001PHR is) then we get 13.0572W output. Based on djozz’s chart I’d say we’re looking at a 3.5-3.6A drive current. Possibly less depending on just how "not fully charged" the cells actually are.
Knowing the loaded battery voltage and LED voltage would help remove a lot of guesswork here.
He’s using a lower capacity battery, but I assume it behaves like this similar cell:
http://lygte-info.dk/review/batteries2012/Sanyo%2018650%202600mAh%20(Red)%20UK.html
Can this buck handle 3x stacked series 26650’s? 12~vdc
Was thinking maybe those 3C defiant throwers from Home Depot would really benefit from something like this
Maybe even a thick ground ring 20mm is in order wouldn’t ya say Mattaus?
Yeah it'll work. And I already have one with a thick ground ring although its the inductor side only.
The thing snowballed on me and I’m gonna have to start over it would seem.
The Crelant 7G5 CS that I’m experimenting on is a friends that had a dead driver. Not sure what this particular lights history is, but it was sent to me some months back for an emitter swap, I put an XM-L2 on a Noctigon in it. At that time, the spring from the driver was all catywampus, I think lap pulls had been used with the weld points not smoothed down or something. So I straightened the spring and got the light working (I only had the pill area) and returned it. When he got it back, he said it wouldn’t work. It was functioning when it left here so I really don’t know what happened. At any rate, it came back to me and I found the driver to be shot.
Yesterday, when I first completed the Knucklehead, it went through all the modes and cast a very nice amount of light. My lightbox showed 1480+ lumens. In taking a tail amp reading I was getting 1.86A and it shut down. No sparks, no smoke or funny smells it just quit. When I got it opened back up the electronic switch positive lead that had been directly soldered to pin 3 of the ATiny13A had come off. Simple enough, I thought. But I haven’t gotten it to work since. With one cell I have a bright light on the test block, but no modes. With 2 cells I have a dim light with no modes. Everything on the board looks fine, far as I can tell.
When I got this complete light with the driver acting up, it looked to me like the solder points on the switch pcb were too close to the host body. One might have been touching, or so it appeared, but the SS ring has a bevel on the outer edge, no grooves or indentions to engage it with, and isn’t coming out! In the process of building the light and troubleshooting, one of the wires pulled loose from the pcb. I couldn’t solder it back on in the hole, up against the edge of the host, and I couldn’t remove it. In attempting to remove the boot in order to better see what options I might have the boot ripped. Then I killed the switch pcb by trying to knock it out from the inside. Figured I’d just re-flash the mcu and use a tail clicky approach, but that’s not working.
The speculation as to what amperage the led was getting is fairly moot. The lightbox showed me 1480 so according to Match’s chart that’s at or above 4A. All those numbers are subject to the meter and dmm on hand so I’m not too worried about that at this point. Gotta get it working and working reliably before any of that matters. I was using the partially discharged cells in an effort to keep the possible damage if something went wrong to a minimum. Didn’t want to build a new untried driver and stick 20R’s on it, ya know? 
So today I’ll build a second one and see how that goes. I got 3 boards and the components to populate em. Just like in our baseball, 3 strikes and I’m out.
Come to think on it there was something odd when the light was on the lightbox. It started out with like a 3870 reading or something like that (not at my desk) and I got a pad to write down start numbers and amperage while it went through the 30 seconds, right? When I got back to it at 26 seconds it had gone UP to 4310! So it went up from some 1330 to 1480+…which is why I was taking a second reading at the tail after the lightbox. Which is when it quit. I had 7 modes or levels working fine, backwards and forwards, perfect. Put it on the lightbox and then tested the amps again and it quit me.
Did you cook the LED?
I cooked 2, the XM-L2 on a Noctigon in the original set-up, and an XM-L de-domed on a SinkPAD in further testing. Hooking it up to an XM-L on an Aluminum star and clamped to a 3” cube of 6061, I can get a single cell to give a bright light, 2 cells give a dim light, no modes to be found.
So, I built a new one this morning and just got done flashing it. Need to hook it up and see what gives…
I put it together fresh this morning, going by the list and populating it according to that. Everything went smoothly, but for one thing….it doesn’t work. When the wire touched the battery there was a light spark, nothing from the emitter at all. The emitter works, direct wires from the battery show that, but through the driver I get nada…
So what am I doing wrong? Must be something, but I’ll be danged if I can spot it!
I don’t know if this is directly related to the driver getting damaged, but I do not think that those UR18650FJ cells are up to the task. It seems to me that the increasing lightbox performance is due to the cells warming up. I assume that they are pretty old laptop pulls and I think that their internal resistance is too high. The voltage dips at first due to the high amp draw (if the 1-2A range is “high”, the cells are weak of course) and then as the guts of the cell warm up it is more able to handle the amp draw. I don’t think it’s possible in this case, but in the RC world a high ripple current caused by the switching of a brushless speed control coupled with a battery having high IR can actuallly fry the FETs on a speed control.
I seriously doubt that ripple current killed your first Knucklehead v3. On the other hand if it was me I’d definitely switch to higher end cells in order to get together some more reasonable test results.
I cleaned the mcu side of the flux that was there, which was very little. The inductor side was reflowed and has a bit more, but still very little showing. The purple mask is shiny, perhaps this is what your’e seeing? When I build FET drivers I use a solder paste mask, as I do when I reflow emitters. I try to put a similar amount of solder paste on each pad with the tip of a very sharp toothpick when building these with no solder paste mask. Hence, very little excess flux and almost no excess solder. I use Kester solder paste, squeeze a bit onto a piece of wax paper and select a very small amount for the pad, rolling the toothpick to separate it from the little dab of solder paste and then selecting less if it looks to be too much. Placing it on the pad, I twist the toothpick to place the solder paste on the pad and will separate and remove some if it seems excessive. There’s actually very little solder involved with the 40+ solder points of the 13 components on this driver.
These 2 “old weak cells” pull 4.47 and 4.45A when connected to an XM-L2 on my test bench in direct drive. That’s moment’s before typing this, after whatever I’ve been doing with these drivers. So even at ~4.1V (The 2 together show 8.20V) they’re still quite capable.
I have a list of the board numbers and I put each component beside it’s listed name. One side of the board at a time. Almost zero percent chance of getting one wrong. Unless DigiKey sent the wrong component with the correct label on it.
I had killed the e-switch, so this one is not set up for an electronic switch. Checked that too, just to be sure, but direct manipulation of the 3rd pin to ground results in no effect. I have other lights that run with a tail clicky set up with this UI so I know it works fine that way.
For initial testing it might be more helpful to know emitter current than lumen output. Maybe use two wires with a meter in between for one of the led connections. I agree about cleaning the flux. You take such great images but the detail is lost under the residue.
I don’t have much practical experience soldering. I don’t use flux. Whatever is in the solder paste is what is there. AND I wiped the boards clean with a rag and a toothpick. There is no flux or very very little.
Perhaps someone could tell me how to clean the driver so that there is no flux left. I tried de-natured alcohol and that wasn’t pretty. Around the joints turned white.
Can’t measure at the emitter as the emitter won’t light.
I’m no guru on cleaning, but I think those who are will be able to help you better if they know which Kester solder paste you are using. With regular flux there are different ones that get cleaned different ways, I assume the same is true of the stuff incorporated in the paste.
I don’t harbor a strong belief that there’s a problem hiding under there anyway, but since all your orientations are the same as the previous driver I don’t know where else to look. It sounds like you addressed all my other suggestions, so I’ve really got nuthin’. 
On the batteries thing… my recommendation stands. Weak was maybe an emotionally charged word to use, sorry for that. I do still suspect that those cells have an unreasonably high IR or something else funny about them. I don’t know of any other reason that your lightbox results would go up over time.
On the latest picture: is the voltage regulator good? It’s the most twisted thing on the board, but from here it looks 100% fine. Actually, what about the R2/U2 area though? There’s some extra solder in that area.