I'm working on 5 x XML light with 2 or 3 x 26650 series.
Before installing I tested a KD 5*XML driver using an external setup. I first tried with two cells and got 2A per emitter on high. I cycled through the modes three times, and the driver stopped working. I don't know why, but there goes $20. I never had a chance to try it with three cells.
Okay, let's see what will happen if I hook up a KD 7*XML driver to five emitters. With two cells I got 2.6A per emitter. When I tested it with three cells and got a reading of 4.5A per emitter.
Holy Moly! 5 XML at 4.5A per emitter. Can this driver hold up? These things are costly. This test was all done externally, so I'm wondering how hot this light would get running on high. I really don't want to blow another $20 driver.
I just re-rigged the KD 5*XML driver and it seems to be working fine. It’s also giving me a reading go 4.5A per emitter with three cells. With two cells I’m reading 2.66A. Are these drivers going into DD with three cells?
I have the emitters mounted on a 2lb slab of copper with arctic silver paste. They’re the original XMLs that I removed from the light and are still mounted on aluminum stars. When I was running at 4.5A I saw smoke out of the corner of my eye. One of the stars had come up off the copper and was in the air with smoke coming out of it. It’s still working.
I think Tom discovered that what kills the big boost drivers is asking them to do too much voltage, not current. If that applies across the whole line, then the 7x driver should be less stressed driving 5 LEDs than the 5x driver.
You do have all the LEDs wired in one single series string, right?
Yup. All in a string.
They both run as advertised with two cells, 2.6A per. With three cells they hit 4.5A. I’m wondering if they’re in DD with three cells. Would DD mean that the circuitry is being bypassed causing no harm to the driver. I don’t remember TomE getting readings like this on three cells.
5 cells in series, direct drive, might give you that kind of current. But not just 3 cells into 5 LEDs. Not nearly enough voltage for that. You can confirm this yourself, measure input volts/amps and compare to output volts/amps.
I'm using a KD 7*XML driver with 6 XM-L2's in a J18, and a KD 9*XML driver with 7 XM-L2's in a J18. Both work fine with 2 or 3 cells. I also am using 4 cells in that 2nd J18 driver, and tailcap amps goes up crazy, but it survived, but don't think I've run it more than 3 mins on high in one run. I used Fujik epoxy in between the boards on these drivers, packed in well - think it helps keep them alive.
If you got 4.5A per emitter on 3 cells, that's a 7.5A tail measurement, which is (I think) about what I got on the 4 cells and 7 XM-L2's.
Could be you are over driving voltage, maybe causing the driver to lose amp regulation, dunno. But it does sound very much like how my 7 XM-L2 J18 performs with 4 cells.
I have tested the 7x and 9x drivers from KD a lot (the 5x and the 7x, I beleve to be the same driver), and I have come to the conclusion that these driver are not strictly output current regulated. Normally boost drivers will keep the same output current, but be more efficient when step up is reduced. I have not figured out exactly how these are regulated, but my emprical data suggests that when step up is reduced both output current and total power output goes up following the better driver efficiency. The downside is that output also sags when input voltage sags. In my j12 I use dedomed Noctigon MX-L2s, 7x KD 5 mode and three cells. It starts at around 4.1A on fully charged cells (MX-L2s demand higher voltage than the stock emitters), and it has survived a great deal of use. I have installed 6 mm copper under the pill to help dissipate heat, though. I have not yet tested this driver to drive three emitters from two cells, but I will very soon. 8)
Thanks for that GBD.
I too was thinking that the 5 & 7*XML driver are the same. The fact that TomE’s 7*XML driver didn’t react the same way that my 5 & 7*XML drivers did when adding a third cell makes me think that he might actually have been using the 9*XML driver.
I have the 9* driver on the way! and I’ll run some more tests when it arrives.
Actually, now I wonder if all three drivers aren’t the same and it’s just a case of how many cells you use. This LCK Driver is the same one and is good for 5 - 9 XMLs. Do you think KD has one big bin that holds the same driver 5-9?
I think the upper limit volage is really related to max step up from two, three and four cell voltage. The 7x does not really work well driving seven emitters from two cells as the 9x has trouble driving nine from three.
They will drive the claimed number of LEDs as long as they are XM-Ls, but not if they are XM-L2s. The higher Vf adds up to put the total over what the drivers can do. So with XM-L2s, you need the 9x for 7 LEDs, the 7x for 5 LEDs, etc. Just don't ask me how to drive 9 XM-L2s... :Sp
Wire them 3s3p, use a 9A buck driver from 4s voltage. Btw, I am aware that the MX-Ls they use in the multiemitters require very low voltage, and therefore work well driven by the KD, but you still need three cells to get a desent output from seven emitters, and four for nine.
Well, I decided to use the 5*XML driver to drive my five XM-L2 light. If it dies or doesn’t function well I’ll throw in the 7*XML version. I’ll post my results when finished. I have a few more 7 and 9 drivers in the mail, and I plan to do some experimenting with them. I’ve never built any boost driver lights before, so I’m entering into a new realm
