Was trying to change out the optic from an EDC18 to Carclo 10507. Had a hard time removing bezel. After removing bezel and optic, I did turn it on a couple of times to see how LEDs looked without anything. Put Carclo in and Anduril was not wokring well at all, unpredictable. Did reset. Sometimes seemed back to normal, but then would not turn on with single click and other things. Now nothing works except green button. Looks like something got fried……
Maybe you’ve turned the bezel and PCB too much that provoked some kind of “rupture” in the wires or in the aux led?
Or can it be a contact issue of the PCB with the head structure?
Not owning an EDC18 I cannot help troubleshooting but maybe someone who has can do it 
Just swapping the optic really shouldn’t have caused much damage. Since the aux LED is working that means the driver has power.
Most likely one of the wires on the star came loose.
I would try the following:
- Remove the bezel and examine the wires where they connect to the star. Are the solder joints intact? Perhaps the issue is as simple as a wire on the star came loose.
- Like the FW3A, I think the EDC18 has a screw on the shelf that holds the star. The purpose of this screw is to prevent the star from rotating and damaging the wires when you screw on the bezel. In removing the bezel, perhaps you lifted up the star and it ended up sitting on top of this screw out of position rather than flat on the shelf.
- If that happened, the star might well lack sufficient heat sinking. It would overheat and probably one of the bondwires from the driver would desolder itself from the star.
- The solution is to rotate the star back into position so the screw is in its socket, resolder the loose bondwire back on and reassemble.
- If you want to be extra careful, remove the star completely, clean underneath and on the shelf and apply fresh thermal grease before reinstalling it.
If that doesn’t fix it, next thing to try is to remove the driver and take a look at the wires on it. But hopefully that won’t be necessary.
According to Illumn for the Lumintop EDC18:
CARCLO OPTICS WILL NOT FIT WITHOUT ADDITIONAL GLASS LENS IN FRONT
Maybe that’s the issue? You could stick the original optic back in and see if it works again.
Oh yeah… that’s another possibility. The opening in the metal bezel is 24mm… the same diameter as the Carclo optic. If you do not use a glass lens between the bezel and Carclo optic, the bezel won’t actually exert any pressure on the optic.
Without pressure from the optic pressing the star down into the shelf, the star almost certainly won’t have sufficient heat transfer and it will overheat. If this happens, the light will probably stop functioning when one of the driver wires desolders itself from the star.
Solution:
- Resolder the loose driver wire.
- Install glass lens.
@fxsniper
Was trying to change out the optic from an EDC18 to Carclo 10507. Had a hard time removing bezel. What method did you use to remove the bezel?
After removing bezel and optic, I did turn it on a couple of times to see how LEDs looked without anything. By "without anything", do you mean without the optic & bezel ring being put back on, or without the LEDs lighting up?
Put Carclo in and Anduril was not wokring well at all, unpredictable. Can you be more specific? Because simply installing the Carclo wouldn't make any difference, except if when attempting to get the optic's 3 feet set into the MCPCB, that you inadvertently knocked one of the 2 wires loose at their solder joints. Since it could happen, but be barely noticeable, you'll need to check to be sure both wires on the MCPCB are soldered in place securely
Did reset. Sometimes seemed back to normal, but then would not turn on with single click and other things. With Andúril, it's possible that when we believe we set it up to function a certain way, that we mistakingly don't complete the programming or set it incorrectly. This is especially true while still learning all of Andúril's functionality. This has happened to me, including when I thought I had "Reset" the programming to start over! So is it possible that you inadvertently set it to say, "Lock out" mode, or perhaps stuck in a sub-menu loop?
Now nothing works except green button. Looks like something got fried…..Although it's possible, I'm betting it's a programming error or a wire/solder joint issue!
Otherwise, the LEDs should function normally with or without the stock or Carclo optic in place.
BTW, I replaced the stock optic on two EDC18s with a Carclo 10507 to successfully achieve a much stronger & tighter beam profile, but instead of adding a lens, I sandwiched a 23mm x 1.5mm GITD O-ring between the outer edge of the Carclo and directly underneath the bezel ring for a perfect fit and nice little green afterglow!
I don’t think it’s a programming error.
- fxsniper didn’t reflash the firmware, open the back of the head, or even remove the star. It sounds like the light still has the stock firmware, stock star, stock leds, and stock solder joints.
- Lockout in Anduril still allows moonlight when the button is depressed. And when connecting power for the first time the light will briefly flash once. But fxsniper reports that the main LEDs do not light up at all.
I think the most likely scenario is one of the driver wires on the star desoldered itself.
- The likely cause is overheating from the star being out-of-position. Either it twisted slightly and rode up on top of its alignment screw, or the lack of a lens or extra o-ring above the Carclo optic causing insufficient pressure on the star.
- The behavior of the light fxsniper reported before total failure is also consistent with a loose driver wire. Initially, the driver wire might have desoldered but still be touching enough to work intermittently, until it separated fully and no longer works at all. I’ve experienced that exact same behavior in lights and a loose driver wire was the cause.
The problem should be easy to fix. It shouldn’t take more than 5 minutes. But fxsniper will need a soldering iron and basic soldering skills.
[ [quote=Firelight2]
I don’t think it’s a programming error. * fxsniper didn’t reflash the firmware, open the back of the head, or even remove the star. It sounds like the light still has the stock firmware, stock star, stock leds, and stock solder joints. Holy cow, where to begin. First of all, what did I offer that even suggested that the OP had done any of those things? Second, you are aware that the EDC18’s stock firmware is Andúril-1? Third, apparently I should have included in the subject statement above, specifically regarding the word “programming”, that I was referring to setting up the EDC18’s firmware to meet 'fxsniper’s specific needs, not reflashing Andúril-1. As I mentioned in my original attempt to offer some helpful assistance, that sometimes as we set up Andúril to our individual needs or desires, we can inadvertently find ourselves getting temporarily stuck in what I would describe as an “error loop”. This is especially true for those who haven’t become proficient in this process! In fact, when we trust that a reset of Andúril-1 will get us back on track, any of number of ‘programming’ missteps could keep us stuck in the previously mentioned error loop. Missteps might include miscounting the specific number of e-switch clicks required, or a click might not register, etc. Even a glitch in the firmware is possible, especially since such could occur at the manufacturing level* Lockout in Anduril still allows moonlight when the button is depressed. And when connecting power for the first time the light will briefly flash once. But fxsniper reports that the main LEDs do not light up at all. Well, that’s not exactly what ‘fxsniper’ said, which is why I requested clarification I think the most likely scenario is one of the driver wires on the star desoldered itself. * The likely cause is overheating from the star being out-of-position. This assumes that ‘fxsniper’ kept the EDC18 turned on for long enough at a level that could potentially cause such to occur. However, ‘fxsniper’ never mentioned having done this. Either it twisted slightly and rode up on top of its alignment screw, or the lack of a lens or extra o-ring above the Carclo optic causing insufficient pressure on the star. * The behavior of the light fxsniper reported before total failure is also consistent with a loose driver wire. Initially, the driver wire might have desoldered but still be touching enough to work intermittently, until it separated fully and no longer works at all. I’ve experienced that exact same behavior in lights and a loose driver wire was the cause. The problem should be easy to fix. It shouldn’t take more than 5 minutes. But fxsniper will need a soldering iron and basic soldering skills.
[/quote]
I agree that a wire coming loose, or already being compromised, is the most likely scenario, but "isolating" one sentence out of all that I suggested in an effort to invalidate what I had to offer is disrespectful to me and a disservice to 'fxsniper'. However, I prefer to assume that you simply misunderstood what I wrote, and that your intentions were honorable.
Because I actually have the EDC18 and have replaced the stock optic with the Carclo 10507, but also had problems with setting up Andúril-1 to my specifications, my intention is to keep the focus of this thread on helping 'fxsniper' to find a solution.
Hopefully 'fxsniper' will report back if still having problems, or will let us know what the problem was and how it was fixed!
When I changed the emitter in mine, it didn’t work when I went to put it back together. What happened was the negative was weakly soldered, came slightly apart and therefore did not work. After the proper solder, it’s fine now. 
I also have an EDC18, in which I reflowed the emitters and installed Carlo optics. I have Anduril 1 on at least 10 different lights.
Here’s why I think the problem is not a firmware configuration error as KnotSoMuch speculates.
- It seems vanishingly unlikely that fxsniper happened to just now experience a brand new never-before-discovered major issue with Anduril that nobody before has ever reported.
- Anduril 1 is complex… but it is mature. It has been out for years and is very well tested. If it were possible to “brick” a light with Anduril just by plugging in the wrong sequence of button presses, someone else would have discovered that by now. No doubt there would be mention of that as a bug in the giant FW3A problems thread or the Anduril thread. But to my recollection there have been no such complaints. I’m not aware of any way the configuration could be messed up to make the main leds not light up at all.
- If the light were left in a configuration mode, my recollection is that disconnecting power exits all configuration modes.
- You could set both min and max ramp at step 1 out of 150. If you did that, the main LEDs might be dim enough that nobody would notice them on. But even then, many functions would still work: (1) double-click for turbo mode, (2) the flash when connecting main power. (3) voltage indicator, and (4) disco modes.
- I suppose fxsniper could have set the ramp settings to 1 accidentally, then never bothered to remove the battery or test any of the functions that still work. But the chances of somehow accidentally setting the light into that configuration and then not trying anything that still works seems infinitesimally small and not supported by any evidence in this case. And again if fxsniper happened to do that, he might have been the first in the world to do so. Seems highly unlikely.
- I would be more convinced the problem could be driver-related if fxsniper had tried to remove or reflash the driver. It is quite possible to brick a driver by dislodging components from the driver when removing or replacing it, or during reflash. But none of that what was done to fxsniper’s light.
The available evidence strongly suggests the problem is a loose driver wire.
- Timing: Fxsniper reported his problems began right after swapping the stock lens for Carclo. He never mentioned anything about being in a configuration mode or not knowing how to use Anduril.
- Description of malfunction: The malfunctioning behavior of his light exactly matches my personal experience with loose driver wires. Initially, it works fine, then it may become intermittent as the wire desolders but is still close enough to touch some of the time, followed by zero light when the wire separates from the pad enough to no longer touch at all.
- There is a plausible cause: There is also a logical explanation for how the driver wire became loose: The board rose up slightly out of its socket due to either not using an extra lens or o-ring, or an error during reinstall that left the star rotated slightly and riding on its retaining screw. In either case, the star would no longer sit flat on the shelf resulting in insufficient heat transfer.
- In my experience with similar small triples with inadequate heat sinking, the star can become hot enough to melt solder extremely quickly. 1 or 2 minutes at high power is enough, even after Anduril’s thermal stepdown.
Yeah and usually it’s something simple that maybe got overlooked or not connected properly that I find. The hard part is finding out which simple stuff. 