I connected the meter and battery directly to the LED. It measured 3.8 before tripping fuse. I tried again and the same thing happened. After a few tries I was able to get the meter to however around 3.8-3.9amps.
I’m not sure exactly what this means, but I’m guessing it means my meter has too high of an internal resistance as was mentioned earlier. I would think that an XML2 should be able to pull over four amps easy when connected directly to the battery?
Well it sounds suspiciously like three chips are not working.
True you cant get a regulated driver to pull more then DD but it is suspicious that three chips just about are the difference of what the emitter pulls DD and your measured “100%” regulated.
cold solder joint?
Damaged, moved a chip (or three)
Just not sure (this is my personal favorite)
Is there a way to test the chips? It makes it more difficult because the driver is jumpered to a 20mm contact board. Here is a link to some close ups:
The top chips should be grounded through a continuous ground plate on the bottom chips but just for kicks try soldering a piece of wire between the outer ground tabs of the upper and lower chips. If the center pin isn’t continuous to the ground ring then those upper chips aren’t contributing. You could check that with a meter first.
One time I added chips to a driver and connected the left and right pins before assembling the light. Only after assembly I realized that I forgot to connect the grounds of the chips. The easiest way is to solder the big tabs at the ‘top’ of the 7135 chip to the ground ring of the driver (or to the tabs of the chips below the one you stacked.) Visually, either all three pins at the bottom will have solder on them or the two outside pins and the tab at the top.
All of the ground tabs test as continuous with my meter to each other and to the ground ring. So soldering a wire between tabs wouldn’t be helpful right?
I am unable to see the central pins of the chips in between boards but all of the chips I can see correspondingly test as continuous. Except for the control pin of the single chip as that one is the moonlight chip. It is connected to pin one of the mou and the others are connected to pin two. I think anyway…
From your picture the center pins look soldered but not the tabs so I took a stab in the dark. It shouldn’t be an issue but I always soldered the outer tabs otherwise the current through the stacked chips is forced to flow through the stock chips in order to reach the ground ring. There is an Oshpark board for testing individual 7135’s but it only works with uninstalled chips. If you’ve tested continuity between the “like pins” on each of the stacked sets of chips(led out-led out, gnd-gnd, pwm-pwm) and all test good then it’s a choice between pulling chips one by one until you find the faulty ones or going with a new driver. Since moon mode works that’s the one 7135 you can rule out. From the pic that looks to be the one single chip on the mcu side. I’d start with the stacked pairs since they are the most suspect and easiest to get to without removing the contact board and remove a pair and retest until I got the correct current for the number of chips remaining then start reinstalling chips.
Can you measure the current draw in the other modes for us? Is it possible that the stacked chips are overheating? The big tab transfers heat to the ground ring which then transfers it to the flashlight body/head etc.
Tonight I will test the current draw from the other modes. I have already measured those and I have somE numbers in my head, but knowing me, I’d better check 
Then I think I will go ahead and solder the ground tabs and see what that does. I’m measuring the current on startup so I would think they would overheat in a matter of seconds but I could be wrong. I could see how the current may be affected by only the pins being soldered. If that is the case though, I would think Richard would know to solder the tabs? I wonder if others have mentioned anything?
Also, all this and it may still be my meter… I have an amp meter on the way though.
Well friends, I hope that the problem is solved. Before I was unable to check the chips on the inside so I decided to make some test leads out of my precision curved tipped plyers. It worked great and I found that two chips were not connected to ground. It was the two chips inbetween boards that did not have soldered tabs that were discontinuous(you can see them in the pics I posted earlier).
I soldered all tabs together and all tabs directly to the ground rings. Two of the stacked chips lost the connection between the ground pins when I soldered them. I had no idea as to how long to hold the iron there. I decided not to try and fix those as who knows what I would have done to them:/ The first tabs I soldered together were submersed in a large clump of solder. The next two were a little better. And I think I did ok on the last two:)
I have not done any testing yet; I will tomorrow. I have good hopes that this is the solution! Thanks for the help thus far!
Here is a picture of my shady handy work:)
Hello Friends! I am excited to say the problem has been solved.
The modes are now as follows:
1ma—200—1000—1800—2800—3700 yay!
Thanks for your help! I would t have had the confidence to test these things on my own. I am happy happy:)
Ps. If I put another chip over the moonlight chip will the moonlight mode bump up a bit?
:party: I’m happy the light works now!!
.
I’m NOT sure however, I did read more than once that even if a chip is added to the moonlight chip, there is no difference (not logical to me) and that there must be software changes for a difference to be seen.
Ok. I guess it makes sense. 1/2 lumen dubbled is still only 1 lumen. Not much change there. Idk if that’s the way it works or not though… Thanks again for your help.
