I started swapping this LT1 a year or two ago, ran into this problem, and shelved it, but I would love to get it working again. I started modding this before I had enough proper equipment, and managed to damage the MCPCB in a few places, where the silk screen has come off. Possibly this could be why?
My problem is that the high CCT channel (519A 5700K) works perfectly, but the low CCT channel (FL351A 1800K) is only working on a single led, the very right one in the picture. Same problem with the MCPCB installed in the light or not.
The warm led on the right side is the only one that lights up with a multimeter on diode test on the board’s solder pads, while all the cool ones work when probed or powered by the driver.
For whatever reason, the pads labelled “NW” and “WW” are reversed relative to the printing beside each led pair. I don’t remember if the stock emitters were reversed.
I can probe each warm LEDs positive and negative solder blobs on their sides, and again, only the emitter on the right lights up. I reflowed each 1800K led again, EXCEPT for the working one (If it is reversed on its pad, wouldn’t that not allow that channel to work at all?)
making sure they were on in the proper orientation according to the printing, but the problem remains.
I would prefer just to get a replacement MCPCB to rule that out, but I doubt I could get a new one. I’ll email Sofirn anyway. I feel like the board got too hot the first time I tried the swap, I was using an electric griddle back then. It looks burnt, it’s more off-white than before.
Rather than test the LEDs, I would put the meter on continuity and make sure that the 4 warm emitters have their + and - contacts connected together as they should be. You can remove the LEDs for this test, or you could use the extra solder blob that emerged from the pad as a proxy for the pad itself.
The above should let you know if the problem is with the board or emitter; if the board is ok, then remove the problematic LEDs and test them directly without reflowing to anything. If the board is problematic, there are lots of ways to fix it like scratching off solder mask and soldering on jumper wires. I’ve done terrible things to lots of my boards, including multiple reflows that were hot enough and long enough to kill the emitters on them, but the board itself is still fine.
No, the channel should still work. I’d start by following the great advice above. If I were in your shoes, those leds not having any marking would make me think I reversed the polarity. I’d remove the three and diode check them off the board to make sure they work. Then test the board. Watch for too little or too much solder under the leds.
Polarity reversal is a real possibility! Before reflows, I always check the bare LEDs to make absolutely sure that I solder them with the correct orientation. Lots of LEDs (FFLs above, also LHP531/73B) have no visible markings on the top and require either testing or examining the solder pads to determine polarity.
Thanks, I will have to try that. I don’t know why I didn’t think to test the three other emitters while I had them off the board.
Forgot to mention that the first emitter I reflowed yesterday, the one at 12:00, worked along with the 3:00 one when tested, until I reflowed the other ones.
I was using the helping hands/lighter method to reflow, as I usually do with everything else (I put smaller MCPCBs clamped on a copper plate though). I have one of those mini hotplates that would fit underneath all the emitters, but I need a more powerful USBC wall adapter for it. I accidentally cut the cord on the laptop PSU I used for it before. 60/75W I think?
I will use the hotplate next time, I used the lighter method so that I could take the emitters off one at a time. this MCPCB takes a long time to heat up enough and very quickly cools, it’s aluminum.
Probably I will try it again tomorrow night, thanks
I did make sure they were all aligned properly once I reflowed them yesterday, with the tiny notch/cathode mark on the “-” side. If I’m not familiar with an emitter I try to find the datasheet or a Köef3 test. I’ve definitely fried parts when I was newer to modding.
I will attempt this again tomorrow, and I will make sure to test the FL351As individually when they’re off the board. I know if I have to I can try to fix/bridge the traces/emitter pads. I am low on 22/24ga wire, but I do have 28 and 30ga, also very thin enamel wire.
I noticed a few spots on the board that look like this:
I’m guessing the trace layer is separating from the aluminum? doesn’t look like it goes all the way across the trace, and the positive is shared with the working 5700K channel.
I’ve just been using this cheapo multimeter, I will need to get a better tool anyway. I used the diode test setting, I’m not sure if that is the same as continuity mode on this one? I have seen both symbols side by side on fancier tools.
Huh, I didn’t realize the MCPCB is aluminum! I don’t think the separation of the trace would be a problem.
I’m not familiar with this particular multimeter, I guess you could look up the manual. Or use the setting to measure resistance, since continuity is just low resistance.
mine is from O’Reilly, but I see this model multimeter in almost every store around here rebranded a bunch. I used to have a red one that was AutoZone brand. it didn’t even come with a manual or anything.
actually, now that I remember, I do have a slightly fancier one but the leads were missing, I’ll have to see if these ones fit it.
I asked Sofirn, and like I expected, they don’t stock an MCPCB for it since I guess the light is discontinued.
I’ll mess with it more after work. even if somehow the ground for that channel is somehow bad, I guess worst case I could find a way to use four 10mm MTN MCPCBs instead. I will test the LEDs. I don’t know if they could have randomly went bad, but I don’t have any more in this CCT.
Kinda-but-not-really bump cause this new post will be a bit different than what I mentioned above:
Alright so I may have come up with a new plan. Something is wrong with this LT1 MCPCB because I have reflowed both the 519As and FL351As to SST-20 4000K and others, the warm side is just not functioning properly no matter what. I think I cooked something with the hotplate and/or lighter method.
My new plan is to sand this MCPCB down to bare Al, then solder 2-4 triple or more MCPCBs to it with LEDs flowed onto them. I MUST have this light working again.
I have some questions and am asking for advice though:
My plan is to get this MCPCB (not my picture) down to bare metal, and then attach either three or preferably four 20mm triple MCPCBs, two with 1800K and two with 5700K.
Flow emitters onto these 2-Channel Hank boards. If I do that, as in one DT8 MCPCB on either side of the stock LT1 one, would having twice as many LEDs do anything in relation to driver current?
I guess since it’s only “designed” for 8 LEDs, surely having 16 wouldn’t hurt, right? One channel on each board having its own CCT, and then connect both boards’ pads with each other (Pos to Pos, Neg to Neg each).
The BLF LT1 shares the positive, though, while it looks like the Noctigon 2CH boards share the negative. Could I just put all the LEDs on the pads backwards and solder the driver wires the same, or would that cause problems?
Basically this is a quick crude picture of my plan:
Oh yeah that’s right, thanks. Maybe then I could try to find a copper disc close enough in size I guess, I completely forgot about the whole soldering to aluminum thing. I’ve put together a diagram.
Worst case and last resort, would maybe be JLCPCB. I’d rather not go that route, I’d like to get this light working with minimal extra purchasing.
