Btw: there is a (too) thick solder blob at the driver pad for the negative LED wire, this might get in contact with the pill. But this would make the light run in direct drive instead of going off.
So - are you using a protected cell?
I think you have moved several amc’s from their places now…
Did you screw in the tube in the proper direction? If it is the wrong way around there won’t be ground connection from tail to head.
Yeah, looks so.
If the PCB has been heated up so much, there is a good chance that components at the other side also moved.
Sorry, forgot to answer before Flashy Mike :person_facepalming: I’m using a Samsung 30Q!
Here are the photos from the back! Despite I made a bit of a mess, it seems to me that there is no connection from the blob and the pad. Maybe I’m disregarding something?
On the other driver I received from Lexel, I made that mistake and the light only worked on max. I saw the problem and eliminated it and works fine now.
On this one, I tried to clean that and avoid a connection from the solder, but I may be wrong :person_facepalming:
Yes, everything is in the right direction! Non-anodized thread with anodized edge upwards and anodized threads with non-anodized edge towards the tail!
Nop, the chips were already that way! Here is a pic of the driver on the day it arrived (left) and today (right). Note the numbers 713, they are in the same position (one at 11h, the other at 5h)
The chips are in the same position despite I may have heated up to place the spring.
Some tests I do for trouble-shooting flashlight mods:
Tools I use:
- An old radio-shack plastic 1xAA-sized battery carrier with 2 wires coming out of it. I use an IMR 14500 in this carrier for purposes of testing.
- Two jumper wires. (6” wires each with alligator clips soldered to both ends). If you don’t have alligator clips you can still do the test by temporarily soldering the wires instead of clipping.
- A small DMM that can be set to emit a continuous beep when no resistance is detected.
Tests to perform:
- Star check. Sometimes the problem might be an improperly reflowed LED or a bad LED. Unlikely in a parallel triple, but this check is easy: Take the battery carrier and touch the leads to the bondpads on the star. If all is right with the star, the star should light up. I typically perform this check before installing any star in a light and after every re-flow. This test can be performed with the star inside or outside the light.
- Driver check. Before installing a driver in a light, I like to check that it works. Outside the light I usually solder up a junk star with an old XML on it to the driver, then use jumper wires to temporarily connect the battery carrier to the star. If the driver is for an e-switch, I’ll also temporarily solder on a momentary push button. This test lets me verify that all the electronics are working properly and not defective where I can see them outside the light. If the light doesn’t work properly, I’ll check all the solder connections and look for shorts. Still not working?…. I might swap in a different driver. If it works with that the problem is something to do with the driver.
- Ground connection to pill check. If a light doesn’t work properly, most of the time this is the problem. If the negative connection from the driver to the battery tube or pill isn’t good then the light won’t work. After assembling the pill, use a DMM to check for resistance between the pill or battery tube threads and the driver. On the driver you are using you could place one probe on the pill, and the other probe on one of the center pins of any of the 7135 chips. There is a high likelihood that this could be the problem in your case. Perhaps the driver retaining ring isn’t screwed down all the way.
- Star to driver wire connection check. Check the solder points for the driver wires to the star to make sure you’re not getting any shorts. In your case, the solder blob where the negative wire is attached to the driver seems rather large and is very close to the edge.
- Driver component check. Use a round toothpick to pick out any debris between the pins of components on the driver board and to check for loose components. Also check for any components out of position and look for possible shorts.
- Driver wire replacement. Sometimes the problem might be a broken driver lead-wire. If still not working, try soldering on fresh driver lead-wires.
- Tail cap check. With the tail cap removed, use any piece of conductive metal (such as a flat screwdriver) and bridge the back of the battery with the edge of the battery tube. If that fixes it, then the problem is in the tail cap. Check that the tail switch retaining ring is tight. If that doesn’t fix it disassemble the tail cap and check the switch for problems.
The main thing about flashlight modding is to take it slow and methodical. Don’t try to rush.
Yeahhhh, I should be aware of this, above all!! :person_facepalming:
Thanks so much for the advises Firelight2!! I don’t have those tools or devices as I didn’t need them before, only now they may be missing to check some points! Most of my mods so far where LED and driver swaps :person_facepalming:
Today I built another triple, with Jaxman MCPCB and a driver from MTN! All worked just fine! But now I’m struggling with this one!
I’ll try to check some of the list above:
Star check - Works fine, as it already light up, and despite all the source (MTN) is reliable!
Driver check - I should have done something like that, but the inexperience and lack of knowledge, and the “faith” that everything would be fine, made me jump that phase!
Ground connection to pill check - Don’t have a DMM to check on that :person_facepalming:
Star to driver wire connection check - +Checked on this and the solder doesn’t make contact neither with the edge of the driver, nor the pill (as far as I can tell). But can’t check much further! I’ll re-solder and see if this works. (Although the is light when I touch the negative wire on the MCPCB and the non-anodized threads of the battery tube.)
Driver component check - All seems to be in place, nothing loose and no debris on the middle!
Driver wire replacement - The wires are 22AWG (pre-soldered on the MCPCB, by MTN) seem to be fine, as the emitters lighted up and no visible damage is seen!
Tail cap check - Works fine on other flashlights, so I’ll say this is OK!
Thanks again for the suggestions Firelight2 
I think we can at least rule out a direct short between battery positive and negative, you should have noticed that. Generally I suggest to use low drain protected cells at low voltage for first tries after such a mod. It helps to reduce the damage from shorts or other faults. A current limiting power supply would be even better.
Perhaps we can see more if you post a close up of the front side of the driver, where the mcu sits. But it’s possible that your driver is fried already.
Thanks for the tip Flashy Mike!!
Hum, there probably was a short somewhere I couldn’t find out! I guess it all was due to my first soldering of the wires on the driver, it may have damaged something.
Yesterday night, after my messages here, de/re-soldered the spring and the negative wire for several times and I managed to light it up for a while touching the non anodized tube an the pill. So I placed it on the head tube and it worked, but it was flickering too much.
I took it out, filled a bit the solder on the negative wire and it stopped working again.
Later I was only able to make it work by soldering the negative wire to the driver’s edge. But it couldn’t work that way so I tried other things and the driver eventually got damaged and it stopped working :person_facepalming: 
I already order a new from Lexel to replace this one.
Thanks again to all of you for your help on this.
I don’t have enough knowledge about this electronic/electrical stuff, and I also don’t have the necessary instruments to make tests or other kind of experiences.
I just like flashlights and want to make some funny stuff for myself 
Thanks
A simple digital multimeter costs less than a new driver, and you get a decent clamp meter for reliable current measurements for less than 40 EUR. In the long run it will save your money.
I know Flashy Mike, and you are right. My question was and is always the same, is to chose one that can be both “budget” and “useful” for what I may need! I have few (very few) knowledge on how to use those instruments, and few knowledge on electronics, so I feel like I’m in a crossroad sometimes…
I never needed before, but I do understand that I may need one :person_facepalming:
Gonna search for some Unit-T!
Have you tried to test everything outside of the host? If it works, then it could be a connection issue with the host. If it doesn’t work, then I would try a different driver to see if that resolves it. And keep swapping out parts until you figure out the culprit.
Here’s how I’ve done my testing to determine the proper bleeder resistor for a lighted switch. I also use a small 14500 protected battery just in case I accidentally short something. Left picture is where I’ve solder the wires to the LED and the driver, right is where I’ve just used tape to make the connections.
Hum, I didn’t try that, I’m honest! I was so tired of soldering and unsoldering that I put it apart and didn’t pick it up since Friday!
I may take a look into that scheme and make something similar.
But my guess is that something has already fried! But, nevertheless, the doubts can be clarified if that mount is done!
Thanks for the tip NeutralFan :+1: