i have been filing for an hour and the driver does not fit in the brass ring… really did not expect this to be so difficult.
I’m sorry, I forgot how it was set up originally, the brass ring that holds the driver was indeed a PITA. I used a dremel tool to enlarge the opening for the driver. It was difficult to get the solder all cut out from where they soldered the driver in at the factory. Too long ago, too many lights, I forget easy as it is but that was way back in August! lol
I ended up putting this reflector assembly in a friends light, then boring out the shelf in his reflector to accept the 3 ounce+ copper sink. Forgot about how the light came originally. I had added 2 pieces of 1/4” thick copper inside the pill to help with heat, and a copper heat sink on top of the chips inside the pill. He says it still works great, has fried the original tail cap but he hasn’t had me put a McClicky in it yet. Get the amperage up to high and the spring in the tailcap doesn’t like it.
light has been completed. very disappointed in my work, but it’s my first and was a learning experience. i’ll be tearing this thing apart again by the end of the week and doing a rebuild. words to live by: take macro shots with a grain of salt… s**t is much smaller in real life. and don’t get me started on that brass ring… grrr.
selected 4th star on driver but am having the following symptoms:
This is what i cycle through “LOW” -> “LOW (same as first low)” -> “MED” -> “HIGH”. After about 3 seconds in any mode, the QLite driver begins giving me the low battery warning… it does so by flashing 9 times and then turning off. What could be my problem here? Did i potentially give this driver too much of a cooking with the soldering iron?
Hmmmm, difficult to diagnose on here but, starting with the Zener mod to the Qlite, you put the 200ohm resistor in place of what was at the positive led lead? Not on top of but in place of? (removed the original, put the 200 ohm there?) Did you make sure the thin line on the Zener was AWAY from the ground ring? Stacked the Zener on top of the capacitor on the spring side of the board?
Resistor like this?
Zener like this?
Shouldn’t have low battery shut off, as the voltage is radically different than the driver was designed for.
Do you have pressure on the chips? I’ve found they don’t like being squeezed.
On those two images of the front side and back side of the 105C, are the ends of resistors "2000" and "472", and zener diode "D4P" supposed to be touching the ground perimeter of the driver? I have installed the resister and zener diode as instructed by you, but i am wondering if something is shorting against the brass ring that shouldn't be. Those components are awfully close to brass ring.
Might want to make sure the one marked 2000 is clear though. That one should be bringing power from the other side of the board, dont want it touching ground. If necessary file a slight detent in the ring at that position.
One end is ground for the other resistors, as is the brass ring, so yes for those that are attached to the ground ring that’s fine.
I usually cut that notch to ensure the red wire has clearance as well. A consequence of using larger ga wire.
Can you test the circuit with the driver out of the ground ring? Apply power directly to the driver via wires from the battery, see if it works outside the light, this will go a long way to show you if the ground ring is causing the issue.
Hmm so basically all the sides of components in closest proximity to the ground perimeter are in fact grounded to it?
I find the behavior to be very strange. Why are my four modes LOW - LOW - MED - HIGH, when it should be SUPER LOW - LOW - MED - HIGH. Without fail, the driver kicks into low battery warning after being on one setting for more than 3 secs, resulting in 9 flashes and then complete shut down.
I have taken some current measurements for all the settings, and they seem to be on par with what is intended. 4.3 for high and lower for the others. The two LOW settings are definitely the same current (about 320-380ma, coincidentally very similar to one 7135).
After all my soldering, and filing, i did give the PCB and circuit components a thorough wash and scrub with 91% isopropyl. This cleaned everything up very nicely, and i don’t see any shorts due to soldering.
For your viewing pleasure...
Measure the voltage at MCU pin 8, this will tell you if the zener/resistor thing is doing what it's supposed to.
What value resistor did you use? What markings are on it, what's it supposed to be, and what does it measure with an ohmmeter?
Thanks. I will double check everything when i get back home tonight. I followed DBC’s instructions very closely. I ordered my zeners and resistors from Mouser, the exact ones that DBC listed in this thread.
Resistor is marked 2000, and Zener is D4P. Zener is also properly oriented.
Sorry Dale, must have missed this post or not read it thoroughly. I will double check on my “2000” resistor. Wouldn’t that shorting against the brass ring cause a complete short resulting in the driver not working and a set of unprotected LiIons going Poof?
Not a direct short, a short through a 200 ohm resistor. Study the naked PCB pic closely. BAT+ comes in from the center, goes to the top pad for the resistor(ex-diode), down to the lower pad which is closest to the GND ring, and from there to the trace that goes to MCU pin 8 (and the via in the 'fork in the road' that leads to the capacitor on the other side).
I = V / R
I=???
V=8.4
R=200
8.4 / 200 = .042, or 42 milliamps. Not enough to make anything explode (not even the resistor), but definitely enough to drain away the voltage the MCU needs to operate.
Thanks Comfy, the explanation of why it works like it does was beyond me. 
Partly my memory, and partly lack of experience/education on my part. Wasn’t sure once the question was asked if that 200 ohm resistor was supposed to be in contact with the ground ring or not, but it doesn’t make sense for it to be being as how it’s the positive supply line. That’s the bad thing about a horrible memory, instead of relying on what you know things have to be sorted by logic repeatedly.
Yes thank you Comfy! This all makes a lot more sense now. Still learning by doing. Will inspect the circuit closely tonight.
Dale, you had mentioned that the tail cap spring is the weak link at high currents. The spring is not what fails when the switch goes, am i correct? So would reducing resistance in both the head and tail springs only allow more current to flow through the whole flashlight? resulting in more current arcing at the switch mechanism itself? then in in turn resulting in the switch failing even sooner (at the expense of a brighter and more efficient flashlight)?
Oh don't worry, I think I get more out of these 'stupid questions' than anybody else. Whenever one comes up and I have to try to find a way to explain it clearly, I always end up learning something I didn't know I didn't know.
I haven’t had a switch fail, only the spring supplying it. The spring heats up and collapses, resulting in loss of contact with the cell. Bypassing the spring with braid or a wire sends the current through the copper wire/braid (path of least resistance) and saves the steel spring from doing double duty. The McClicky that I use to replace stock switches has held up at over 7.5A reliably.
Comfy has pointed out that direct drive is the most your batteries are capable of delivering. Using a driver with regulated circuitry prevents the led from seeing that maximum level. So even if you bypass springs and remove resistance everywhere possible, the regulation of your driver will be the limiting factor. If you have 4.56A of 7135 chips stacked, it can’t exceed 4.56A. Resistance in springs and such will keep you from seeing that, and the driver itself can have losses, so bypassing springs will just get you closer to the theoretical limit of your chips. It will also take that electrical delivery off the steel spring and allow the wire you put inside it to be that delivery source.
Edit: My MT-G2 in the M8 , direct drive on 2 20R cells, would pull 11.88-12.18A. Got a switch that can do that? Mattaus and Tofty are working on just that! 
Sanyo laptop pulls would only allow 7.51A, my McClicky handled that ok. Couldn’t run the 20Rs cause I knew the switch would fry. I now have a Qlite with 16 chips pulling 5.98A with 2 Panny PD’s.
I have run 12+ amps through the common Omten rectangular reverse clickys and the tall square forward clickys (same as the ones SF uses) without any issues (other than springs turning into incandescent filaments but that's not the switch's fault).
FIRST ISSUE:
4th star soldered on Qlite 105c: 5mA - 2% - 25% - 100%
Current readings are as follows for the 4 different modes:
1.85A
2.35A
2.85A
3.05A
Obviously this can’t be right. All 4 modes appear nearly identical in brightness, as would be expected by those current readings.
SECOND ISSUE:
Driver goes into low voltage protection in every mode after about 3 seconds.
Manufacturers description: “Low voltage protection: At 2.9-3V the light will switch to an emergency low mode. At 2.8V the light will flash 9 times to announce imminent cut-off.”
Here are some measurements before voltage protection kicks in. The LED is still on and bright at this time (i am unsure of what mode the driver was in when i took measurements):
6.62 Volts at batteries (2 cells)
4.59 Volts at MCU pin 8
Once voltage protection kicks in and the LED turns off, these are the measurements:
8 Volts at batteries (2 cells)
7.11 Volts at MCU pin 8
Also, from the positive spring on the driver to pin 8 on the MCU i have 200 ohms resistance. Seems to me that the resistor is doing its job.
Did i fry the MCU? i am totally lost here… been trying to get this working for almost a week now.
I am baffled.
My added zener diode is reading 656mV one direction, and 1050mV the other direction when tested with my DMM. Does this mean anything?
UPDATE: After troubleshooting with RMM the driver was deemed defective. Will try again with the mod etc… once a new driver comes in. Thanks for helping out. Findings here: How to test AMC 7135 for proper operation?