Bit of a long shot but try removing the two 7135s in the L3 group.
I haven’t connected the L3 group yet.
Could it be something coming back to ground, or the neg. pathway? Like I said, I have a clear pos. pathway from battery to LED.
So measure it with the ohmmeter. Check from the LED- to the left pin on the 7135, when viewed with the pins pointing down. Don't guess, check it.
And check from the right pin to the pins of the MCU, that's the PWM circuit. If you can't find a MCU pin that's connected to that 7135 pin, backtrack and find out where it gets lost.
If the traces aren't visible sometimes you can make them easier to see by shining light low across the board from various directions. On other boards you can shine the light from the opposite side like a x-ray, don't think it will work on this one since it has so much area covered by metal on the battery side.
You could try to isolate things a bit. You could remove the solder bridge in the center of the board. Then you each of the banks will be separate. Then unsolder the led negative wire from the driver. Then touch it to each Lx- pad individually to see if any of them work. If any do, that will at least rule out some potential issues like a fried MCU, etc.
EDIT: I would follow Comfy's recommendation above first.
The pwm pins can be traced on the back. Conveniently the battery contact doesn’t cover those vias.
And jumper the LED- to GND, that should bypass all the driver circuitry and give you direct drive but would be an excellent starting point, to ensure you have solid connections for + and - to the driver.
I jumpered the LED- to ground, and got steady light. It wasn’t dim, but it wasn’t real bright either. I was looking right at it, and I’m only seeing one small spot in my eye now.
I checked the LED- to the corresponding pins, and everything checked out. I also checked the other pin and got a beep on the top right end pin for the bottom bank of 7135s, and a beep from the pin beside it from the top bank of 7135s.
I’m getting pretty weary now. I’ll check them all again tomorrow. I think I’m in way over my head here.
Both Vdd & OUT have to be isolated from ground. From OUT to GND should be way way up there, .8-.9M ohms, Vdd should measure around 50k. Anything like a direct short or low ohms on either will make it not work.
All the Vdd pins should have continuity with all the other Vdd pins, OUTs should have continuity with all the other OUTs.
Okay. I’m back in business. I checked everything I could, but like I said, my chip installation was poor at best. I started removing all the chips I installed one at a time, and retested after each removal. The board now has all the original chips, and still no light. Continuing in reverse, I rehooked the wires to the charging port…shazaaam…it works. Before I hooked it up, I tested the tabs on the port. There was no continuity between the two tabs that red wires are connected to, but there is continuity between one of the reds and the black.
I guess I should try adding chips one at a time and testing after each one, but I realize now I don’t have a clue how to mount these things properly. This is a true learning experience.
Butane soldering iron, hot air, solder paste... they snap into place just like reflowing a LED. And it comes out beautiful.
http://www.bernzomatic.com/item.html?id=16
You should be able to find something similar if not identical at HD/Lowe's. You do not use the flame, the soldering adapter has a ceramic catalyst grid in it, the fuel is burned inside and nothing but hot air comes out the front. Works great for heatshrink tubing or rewrapping batteries too.
The red and black that show continuity on the charge port… Bad Ultrafire for mixing red & black. At least you found it! Just keep those two connected if you remove the charge port again.
I have light, and all the modes, but it is dim. The reading at the led is .004A. Everything is jury rigged, but I can’t get any brightness.
The 'jumper LED- to ground' test should have resulted in no light if the BAT+ circuit was connected via the two red wires as in the picture... methinks that test was not done correctly S)
Well. What should I do now. Should I unhook from the charge port and connect the red and black.
I hooked up the black and red wire together and have dim light with modes. What I don’t get is that the leftover red wire, that’s hooked to nothing, has continuity to the pos. LED wire. If I connect it to the battery I get light.
Edit: I’m guessing that that wire would have sent the charge to the batteries.
The black and red wire I connected seem to lead to ground.
I hooked it all back up to the charger port again and still all modes, but dim. I then disconnected the LED and directly touched the battery leads…still dim. Aha!!!
I tested another XML with the battery…bright…so I hooked it up…even dimmer. Tested it again with the battery leads…still dim. I ve got so many wires and clips and junk jumbled up, I might have shorted it.
Finally, I hooked up another XML that used to be in a Roche and I got full brightness, with full modes and a reading of 2.6A on high. I added three chips the best I could. My DMM showed no shorts and connectivity with the corresponding legs of the other chips but continue to get a reading of 2.61A. I must be me. My best just doesn’t seem to be good enough. At least it’s working at the same level it was when I got it.
Thanks for the interesting thread Ouchyfoot. Believe it or not this has been totally enthralling to me and I feel your frustration. Good luck.
I have a young bartender from China named Sabrina. She understands electronics. I brought over my driver and DMM, and she figured out my problem right way. It seems I’m blind. I could swear I’m touching the legs on the chips, but it seems that sometimes I’m missing them and touching the pad below. She was great. She understood how the whole shebang worked and how everything was lined up in series. I’ll see if I can get those three chips working properly. If not, I’ll just scrape them off the board and wait till I have the means to mount them properly.