I installed the five mode 7 x XML KD driver that I had handy, in place of the dead 9 x XML version. With three filly charged Sony US18650VTC 4 I got an initial 2.5A per XM-L2. By the time I cycled through the modes and recorded the readings, I arrived back at high with a reading of 2.37A. Pretty soon it was down to 1.92A.
Yep. With XM-L2, the 7 x driver just can't keep up with seven emitters the way the 9 X driver did, although it hasn't blown up yet. It's still insanely bright even when it gets down to under two amps, but I'll probably install another 9 X driver when it arrives. Remounting the driver is the biggest pain in the butt part of building this light.
Anyway, no rush. When I have a new driver, and feel like rebuilding, it will get done. I like the form factor of this Boruit too much not to have it running at peak intensity. My "show off Light".
As a side bar, in the OP, I mentioned how I like to use Silver Teflon Wire for hooking up emitters. Especially multi emitter lights where space is tight and you have to be aware of clearance so you can fit the reflector down flat and avoid shorting. This wire is very low profile, and not as bulky as silicone wire. very high quality also.
In this photo you can see the difference between the green 24 AWG silver teflon wire and the red and black 24 AWG silicone wire. I've started acquiring this wire in various gauges. It sure comes in handy for those tricky jobs with an unforgiving reflector.
That silicon wire looks ginormous. Besides the gauge, the thickness of the insulation can vary wildly. I bought 24 gauge silicon covered wire that is totally useless for me because the insulation made the overall thickness larger than typical 20 gauge. If you go to a true wire supply house website (only big qty), you can order wire in various gauges, various strand counts, various coatings, and various thickness's of insulation. Unfortunately most resellers don't specify all those details, so it's a mystery what you will get.
I’ve been getting getting good wire from a guy in Iowa. He’s got an eBay store with a really good selection of Teflon wire in all sizes and assorted lengths. Navships I like the silver teflon wire 19 strands.
Interesting... I think O-L is onto something with wire. He found he could reduce the gauge if the strands were less and thicker, getting the same output/amps. Also the silver coating could make a difference too. Interesting way to go - Teflon silver coated, maybe you got thicker strands - hugh help in reducing resistance and keeping a low profile.
I think OL mentioned that he did not think the silver plating made a difference, only the strand thickness. See his post #447 in the Oshpark Projects thread.
17 strand is a low strand count compared to regular silicone noodle wire, so definitely thicker.
EDIT: clearly Ouchyfoot said 19, not 17 - still a low strand count, so still thick strands. BTW nice work on all aspects of this light Ouchyfoot!
I just checked the green 24 AWG silver teflon wire I used for this mod, and it’s seven strand.
In regard to the silver coating, since the wire is tightly compacted into its Teflon sheath, making all strands in tight contact with each other, the silver coating should hep the current flow through all the strands as if it were one.
You are aware that a great majority of the sellers call silver coated when it's actually tin coated.
This silver is discussed on many forums as not actually being silver. Cost for tinned copper wire is higher than bare copper wire. Cost of the silver coated wire is higher than tinned copper wire.
Nickel coated copper wires that are normally used with Teflon insulators for the 200C+ temperature usage, which is the actual purpose of the Teflon, not 60C-80C temperature operation like in flashlights. If something really has 65C in contact with your hand you will drop it for sure (especially if it is a bigger object)
You need to know the exact resistance rating of the wire at 1Km, each wire has such rating, for the exact gauge,strands and coating. Most of the specs on informative websites are just ambiguities that do not account for any of these details.
My main purpose for using this wire was because of its low profile, all other specs were after the fact. This seller seems to carry high quality wire. All he sells is teflon wire. I looked at it under a jewelers loop, and I’m pretty sure the coating is silver.
I have a local customer who asked me to mod his J12 and J18 for him. Out of the three KD drivers, none lasted more than an hour, even being careful with them. These were the 7x and 9x drivers. The 7-emitter J18 had the 9x driver, the 5-emitter J12 went through both the 7x and 9x drivers. The first two were potted, the last 9x wasn't. There was no rhyme or reason for their failure, they were never run on high for more than 30-60 seconds and they all failed in lower modes.
My conclusion is that these drivers aren't even worth your time to install them.
Have to agree with RMM… I blew 4 of those Drivers, tired to kept on installing it.
better just to do resistor mod to bump TR-J18 slightly
BTW do the mod like OldLumens does… 7 drivers! it’s awesome, it’s the brightest and the floodiest in my arsenals :bigsmile:
Well. I’ve got a bunch of those drivers to start blowing. I paid for em, I’m going to make sure I fry em all. Why is there such a void in the market for multi emitter drivers. They keep putting more and more LEDs into lights, and the output keeps getting lower and lower.
Do I have any other options? I’m not that in the know about TaskLED drivers, but would invest if they had the right one that fit. When $20 drivers ably last ten minutes, $40 is not a lot to pay for one that is reliable.
If you are not into making/modding some sort of regulated 7135 based monster, buy a FET driver. Use normal batteries and keep the amps around 15. The standard large Omten switches should handle it..
The problem is that the TaskLED drivers are too big to fit in most of our lights, otherwise they would probably be a good solution. It is a pity because the J12/J18 are really cool lights...just very unreliable. I actually like the form factor and balance of them even though I don't like most of the longer lights. 7000 lumens with decent runtime from 3 x 26650 is great...while it lasts.
My next solution is to remove the center 7th emitter from the J18 and run three cells with an FET driver and wire the emitters into two series circuits with three emitters each running in parallel with each other. You could also do two cells and wire them three and three if you wanted. At least I know the driver isn't going to burn out in under an hour!
Here's something interesting. I took a new 7xKD and hooked it up to three XML with two cells and got a rock solid 3.6A per emitter. The first thing I noticed was that the modes all changed instantly, and there was no whine at all.
Hmmmmmm.
I then hooked it up to seven XML and got such a faint glow it didn't register on my DMM.
Hmmmm.
I hooked it to five and everything ran to specs. I think KD sent me a 5xKD driver by mistake. That's okay though, because I was wanted to try another one. So far, I think the 5xKD may be the best of them all. I have one running perfectly with five XML2 with no problems. It's giving 4.2A with three cells. The specs for that driver are only listed for two cells. It only has some whine on low and instant mode changes.
The 7 and 9 KDs have painfully sluggish mode changes and whine like crazy.
This 5xKd running 3 or 5 XML (with two cells) makes no sound and the mode changes are instant. With 3 x XML I'm getting a rock solid 3.6A per emitter. Nice.
This could be the answer to those triple lights that only take two cells, so a buck driver won't work.
I have a FandyFire STL-V2 triple that was horrid when I got it, but installed a 7A driver wired in parallel. The driver was too big for the pill, but it fit inside the head perfectly. I just tested it, and according to the tail readings I'm getting 2.8A per emitter. Not too shabby.
Anyway, I never really used it or showed anyone because I was embarrassed by the off center emitters and reflector that is sitting too high on top of them. My skills (and wire selection) have improve a lot since then and I've been meaning to redo it. I think I'll throw that 5xKD driver in there and see how it does. I can always put the other one back in the way I wanted it. With no whining or sluggish mode changes, this KD may be more reliable than the other two. 3.6A per emitter should heat this little baby up quick. I'm kinda getting excited about this mod.
I just completed my third driver swap on my Boruit.
If you have been following this thread you'll know that the first KD9*XML driver packed it in after about 20 minutes of use. I then installed a KD7*XML driver. It worked, but trying to keep up, the current levels maintained a continual spiral downwards.
That bothered me, so I didn't use it. I had another 9*XML driver in the mail, that I have installed in the Boruit. It's back up to par, but for how long, I don't know.
I love this light, and am determined to keep it running. Whatever it takes!
If this driver dies, I'm going to search out a five emitter reflector that fits. I'm hoping I can get hold of one without having to buy a $45 light just for the reflector. The one in a TR-J12 or my 3 x 26650 Securitying should fit. I think they're all pretty standard and made in the same place.
The difference would be going from seven emitters at 2.8A each, to five emitters with larger reflector cups that will run at 4.2A each with the KD5*XML driver.
The KD5*XML driver seems more efficient without the short lifespan problems of the other two. See my thread on the KD5*XML driver.
Lets cross our fingers and hope those driver lives a little longer than the last one.
If anyone has a 5 emitter reflector that will fit this light, let me know. It's the next size up from the SRK style reflectors. Or if you can steer me towards a not too expensive retail light with a five emitter reflector. It's 66/67mm in diameter. The same size as a TR-J18 7*XML reflector.
I'm not giving up on this light.
Oh yeah. On this last driver installation I mixed Fujik and Arctic Silver to fill the space between the two driver boards.
