Good to see you found the problem with the driver.
When you go through problems like that & start dissecting the thing, you come away knowing a lot more about it.
+1, most of my “expertise” comes from dissecting these and trying to alter the physical shape, others know tons more about what goes on inside the mcu. The rest comes from reading threads and being corrected when I post in them.
Well, I guess that was all for nought :(…. I came home today and decided to try to replace the P7 emitter with something else (probably would’ve been an XM-L2), but then I can’t get the star out of the pill. It’s glued and I can normally twist those off with a pair of needlenose, but the star is huge, maybe 20mm or actually looks larger than 20mm, and it just won’t budge. I’ve tried using a heatgun, but still stuck.
Sigh :(…
A nifty trick to make for good heat transfer and secure but still somewhat remove able is thermal grease under the star and thermal adhesive at the edges then clamp it in place with the reflector until it cures. Put the pill in a film canister and into the spare parts bin.
Hi,
Ok, I was able to make a working light. I ended up with a kind of “orthodox” light, with an 8x7135 (380s) NANJG and an XM-L2 on 20mm Noctigon, but at least it’s working again, and it’s a heck of a lot brighter than the P7 was.
I was able to get the original P7 star out by putting a small screwdriver into the pill from the driver end into the wire holes and then used a hammer to hit the star from the bottom side. That eventually got star loose and I was able to get it out.
Then, replaced that with the XM-L2 on 20mm Noctigon and used the new 8x7135 NANJG. Unfortunately the driver has one of the MCUs with the labelling erased, so I can’t tell what that is, so I couldn’t re-burn it with minidrv :(…
What are the voltage divider values? If they and the traces are a match to the nanjg ones you should be able to transplant a burn able mcu.
Hi,
I didn’t note the resistor values, and the driver is soldered into to pill at this point, so I think I will leave it as-is for now. Thanks though!
Jim
Can someone tell me if this diagram is correct for a “normal” (non-zener modded) NANJG?
EDIT:
R1 = 19K1
R2 = 4K7
EDIT 2: I realized that the one I posted above probably isn’t right (the 7135s aren’t connected to LED, so maybe this one is better?
Hi,
As you may’ve guessed, I’ve been wrestling with the original driver (the NANJG with FET) and I just got it working again. Besides the broken ground ring problem it turned out that the small cap on the battery side was shorting (yes, it was shorting). I replaced that, and patched up the ground ring in a couple of different places, and it’s working again, for now. I probably won’t be putting this driver into a light, but I’m still waiting (and waiting) for some new NANJG drivers from FT, so this has kept me occupied at least :)!!
Jim
It’s a great way to get the details understood and yes, the 7135’s are between led- and ground. Relic’s drawing shows the center pin grounded and it is but only because it’s continuous with the ground tab(same piece of metal) not because the pad it’s soldered to is connected to ground(in many cases like yours the pads float). When designing traces led+, led-, and ground all need to have adequate trace width or vias to carry the desired current.
Yeah, and it’s amazing how, even with a so-so simple device, just about anything can go wrong!
BTW, so this now raises a question, for me at least: How much can a NANJG driver be shaved diameter-wise? Also, what’s the safest way?
For the one in question, I just used a dremel around the edges, but it seems like it broke the ground ring on the top/emitter side, which was the source of at least the majority of the problems I encountered.
Thanks,
Jim
I spoke too soon. The board seems to be working, but the brightness is way, WAY low…. I mean that even with the brightest mode with PWM set to 255, it looks like a dim moonlight mode. The 1st 2 or 3 modes don’t show anything.
The MOSFET is soldered to one of the Ground/tab pads for one of the 7135 positions, so could this dimness be because the ground path from that pad to body is bad? But when I test outside the light with my bench it is also low :(…
I think this is time to abandon messing around more with this particular driver :(!!