FW3A Troubleshooting / FAQ

Cool — Rayoui beat me to it; was going to suggest the same. The internal battery tube enabling the circuit is a bit finicky. I have an FW3A in 18350 format. Was working just fine. I dropped it on a hardwood floor while off and the light turns on, then rapidly kicks into ramping mode up to turbo. Button control was unresponsive. I had to unscrew both the head & tail to get the tube reseated properly for normal function. Never saw one “fall into muggle mode” like that, but who knows how sensitive is the hardware to jostling. This is why I consider the FW3A an enthusiast flashlight and for an emergency situation, I’d avoid any flashlight with an inner battery tube for connectivity… with the exception of a NovaTac that uses a reliable compression spring.

On my FW3A, the inner tube connection ring on the driver board is very close to the retention ring holding it in place. Almost touching it. Maybe something happened there when it hit the floor.

OK, I’m back to frustration mode on the FW3A. The original with 18650 is no problem. I swap batteries with no issues. I always take off the head for access.

On my FW3A that has an 18350 body, I’m running into problems again. Screw on both ends tight. There’s one quick flash in the final screw-down. And then the switch is unusable. Repeat the screw down and occasionally the light ramps up automatically and then either turns off (and stays off) or stays on—switch unusable.

This one has the tailcap parts secured by a retaining ring. All seems tight. Driver is tight. Wiped down the battery tube ends. Attach tail cap first, slide in battery, attach head. Same as described above.

I kept loosening and then tightening. Sometimes it would auto-ramp and stay on high, or it wouldn’t do a thing. And then… finally, I have control. Switch works. I set it down, head first. A moderate tap on the head. It auto ramped again.

Next, I took some snap-ring pliers and loosened the driver retainer, then re-tightened it. THAT seemed to do the trick. Rapped the light on a desk to test of instability. All seems solid now. But what really bothers me is how the driver was already tightly in place. What is actually the variable here? It leaves me wondering if it’ll happen again.

Yeah, I have a feeling this might be the same thing. The tolerance has very little room for error. Maybe there’s microscopic variations in the thickness of the retaining ring. And just the right rotation of the battery tube might bring it close enough to short.

There are several causes. First of all make sure that the driver and switch PCB are both centered. If the ring has an offset it might not work correctly. Also check for defects in the anodization of the inner tube. Pull it out, look if there are any scratches, especially at the little step. If there are any, look for a cause, remove it and tape something over the scratch. Then screw on the tailcap. Check that there is no short between the outer and inner tube. Same for the spring and inner tube. If that is OK fully assemble the light.

Never had a problem with mine, but as you said, tolerances are pretty small.

one thing i did with mine was, put a ring of very thin tape around both ends of the inner tube

the theory was, it just seemed like it might short out so the tape holds it away from the outer tube

if it did short, it would prob do what yours is doing - thinking you are constantly holding the switch down

depending on tolerances you might have to just use 3 pieces of tape instead of a whole ring of it

also maybe your inner tube has become not round - elliptical so that the long ends touch the outer tube

Eyeballing it, looks like everything is centered. Don’t see any tube bending. No scratches through the anodizing. But I have some kapton tape that’s very thin. I’ll try putting one wrapping of it on each end of the battery tube.

Hi, I need some help with my FW3A. It is not working properly after I did an emitter swap. When I insert the battery and loosely screw the head on the light turns on at a normal brightness, the tail switch doesn’t do anything at this point. When I fully screw the head on the light stays on and cannot be turned off unless I unscrew it. The modes appear to work as normal but a single click to turn off the light just puts it on a low brightness and not off.

My light is the 219C version, and I flowed the SST-20 FD2 emitters on it with a candle. Also I should mention I had forgotten that the battery was still inside the host when I was desoldering the red and black wires from the board and when lifting it off I saw a spark.

I am thinking I have probably messed it up when I forgot to remove the battery, and the thing in the driver that tells the light to turn off is scrambled. So would reflashing the driver help? (I don’t have the necessary equipment for a reflash so I would need to buy it) or should I just try to source a new driver?

I have tested the LED board with a mulitmeter and it works.
The wires of the driver don’t appear to be damaged, the wire housing is intact.
I have cleaned all surfaces of the head, board, driver, and threads, with isopropyl.
I have tried using a multimeter to check for continuity, and there are no issues with the tail / body tube.

I appreciate any help thank you.

That cannot be fixed with a new firmware, you’ve damaged the hardware and need a new driver. Sadly Lumintop doesn’t have drivers anymore and the latest version comes with a different, much inferior driver.

Probably you’ve damaged one of the AMC7135. Depending on the fault it might be possible to check in circuit which one is damaged, otherwise you have to remove one by one until the problem is solved. Then you can either live with a reduced output or find a source for AMC7135 (hard to find at the moment, that’s why Lumintop removed them from their drivers). Either way you need a hot air station to remove the affected 7135.

OK, so I’d been using my stubby (18350) FW3A with only a few hiccups… had to loosen & re-tighten the head to clear that peculiar auto shut-off when ramping up, just a couple times over the past 2 weeks.

Last night I’d used it a bit. No problems. Set it on my nightstand pointing towards me so I could see the GITD insert… and at about quarter to 6 am, I open my eyes to a blinding bright light. My FW3A had turned itself on & auto ramped to maximum!

I was lucky that I wasn’t in a deep sleep at that point and woke up in time. The head was almost too hot to touch and the table had a hotspot where the head had rested. Switch was inoperable. I unscrewed the head and cut power. Fully tightened the head & proper function was restored.

I wonder, if I did a software lockout would this have still happened?

I’ve never had this trouble with my full size FW3A (18650), just the use of the 18350 tube.

Anyway I’d forgotten to apply the kapton tape I mentioned back in April. Definitely going to give it a try now… Hopefully I can find the stuff! (My junk drawer situation is a bit fubar)

the signal tube must have made contact erroneously

which could happen with one tube and not the other

could be a pocket burner!

software lockout would stop that… if that is what it is

wle

Is it the first batch version or later ones? I hear that the later versions were having unreliable quality.

Definitely sounds like the signal tube accidentally shorted. With that short, the driver thought you were holding the button down so it ramped up to turbo and stayed there.

Some tips:

  • Remove, clean everything, reassemble. Check alignments when reassembling.
  • Make sure the tailcap retaining ring is slightly loose. That ring is just there to prevent the tailcap guts from falling out. If you leave it a tiny bit loose the tailcap pcb is able to free-float and may actually connect better to the signal tube.
  • You can prevent the issue of accidental ramp-up burning a hole in your desk by reflashing the firmware to upgrade from Anduril to Anduril2. Anduril2 has a safety feature to prevent this kind of thing happening. With Anduril2, if your signal tube malfunctions and the light ramps up to max it will pause there a couple seconds then ramp back down to min, turn off, and put the light in lockout mode.

IIRC that safety feature was added already to Anduril 1 shortly after the release of the FW3A.

That sounds quite unacceptable to me and a bit dangerous. My old man response would be to either retire that light (remove battery) or else return it to the original 18650 tube. I guess I no longer have the patience to do a bunch of investigating and repairing to make it reliable.

Ah.

Good to know. My first and second run FW3As, my FW3 Ti, and my EDC18 all lack that feature. I just tested them.

Also IIRC Lumintop didn’t update the firmware for more than a year after the fix…

Barring a software bug or unlikely corruption of the ROM, software lockout should guard against this. Given the flaky nature of the FW3As switching I use it whenever it’s not in my hand. Thankfully I’ve only had issues with it activating while clipped to a pocket whenever I’ve neglected to lockout - presumably because that switch is all too easy to bump.

Yep, I use the e-lock all the time.
ALL the time.
I have little need for an instant on all the time. That’s me, though.

It is an early one. In particular, this is the one with the ToyKeeper icon on the switch. My full size 18650 had this same trouble very early on and was resolved after cleaning and reseating the tube. Hasn’t happened again since.

Thanks. I did clean everything before, which should’ve taken care of the problem… not sure how to “check alignments” as I can’t see inside the tube when closing up. But I did not know about that tailcap retaining ring tip. This one does have a retaining ring, unlike my other FW3A. One of these days I will see about flashing firmware, but right now I’m not set up to do it. Good to know that Anduril 2 has a failsafe to mitigate that problem!

Hi everyone! I’m new here, but reading quite a while.
Just got my fw3a sst-20 4000k from aliexpress and wonder how to check did I have a “full version” with all the power channels or missing some.
Head inside looks just like on the left (I mean photo in the first post). What else can I check?