So, what are the symptoms when you have a bad 7135? I mentioned this in another thread, but I have two of these 105c drivers that are dim, and won't change mode. It is NOT something wrong with the tiny13a or the programming, as I took one of them off the broken driver and put it in place of the chip on one of the working boards, and it worked fine afterwards. I thought I might have shorted something on the first one that didn't work because I changed the leads it came with, but no, there was a second one from the same batch that also didn't work in exactly the same way, and I didn't touch that one except to solder the leads to a LED, and connect it to power.
How to go about narrowing this down? Remove all but one 7135 and add them back one at a time, see if it goes from working to broken again? Or is there another component on the board that could cause this?
I had this happen before(twice actually). It happened when the outputs shorted across a metal reflector. My solder connections on the led were too tall. I think they are scrap now.
Well, I stripped all the components off, cleaned the board, checked all the 7135s (didn't find anything out of line). Put it back together with just four 7135s and now it does... something different, but still wrong. Low mode isn't low at all, it's the same as mid, and in strobe & SOS modes it never turns the LED completely off, it looks like it's just switching between mid and high. Very strange.
Nothing external is shorting, it is a bare driver with two output wires soldered to a bare 16mm board on a CPU heatsink. Driver hanging in midair between the heatsink and battery wiring. Power to the outer ring and spring on the driver from a 18650 battery box.
Look at the pad for the LED+, lower right just outboard of the little black quantum discombobulator (code for 'I have no idea what the hell that is or what it does'). You can see a tiny little solder bridge sticking out and going close to if not connecting with the ground ring, cleaned that off before I resoldered the MCU and 7135s and it still acted goofy as described in the previous post, so that wasn't the cause. I don't see anything on/in the PCB itself on either side, the MCU works fine as I tested one on a known-good board, and the 7135s check out. The only thing left is the quantum discombobulator and two resistors on the backside (pictured), and one capacitor on the other side. What's going on here?
Yes yes yes I'm fully aware this is just a $3 part but I have a compulsion to fix broken things, and break fixed things... :Sp
Discombobulator thing is reverse polarity input diode. The only way I know to diagnose 7135’s is to reinstall one, test the board, remove the chip, install another, test the board, remove the chip. Until each chip has been tested on its own. It would be nice if there was a clip or plug for testing them.
I have the board stripped completely clean. The bare discombobulator diode measures 58.6Kohm in one direction, 4.3Mohm in the other. I'm pretty sure that's not how diodes are supposed to work. Can that be bypassed with a jumper and not do anything but disable the reverse polarity protection?
The diode is fine, measure it in the “diode test/tone” mode of your meter. It should give a number between 0.200’ish to 0.600’ish in one direction. In the other direction it should be the same as the leads connected to nothing.
This is a weird one for sure. I usually have the reflector short issue which bypasses the driver (if you’re lucky and the short is on the negative side).
That solder ‘short’ from the LED+ to ground would have been a short right across the battery, so I doubt it was an actual short. Things would have gotten melty (unprotected cell), or would have not worked at all (protected cell).
Other than testing each 7135 one at a time, I dunno what else to try.
I even tried removing ALL the 7135s, and using just one 380mA version I got from I.S., made no difference. I even moved it around to different spots on the board.
Actually, the one that's in a light even does some weirdness occasionally, erratic random flickers/pulses on low & mid. It's pretty mild but still noticeable. I think paying double for 105c's from I.S. and having all of them work comes out cheaper when it's all said and done.
Cause the ones from IS worked, and the ones from FT don't. Even the ones that 'work' do goofiness that doesn't inspire confidence in their durability, the IS ones haven't done a single thing they aren't supposed to do. Should I be using some other criteria for identifying good/bad, other than works/broken?
