FW3A Troubleshooting / FAQ

[SIGShooter]

@DBSAR

@teacher

@CLB

@DBSAR

while the edge of the spring was not directly on the trace, as seen in the photo the spring edge was roughly 110 degrees at roughly 2 o’clock on the board, so the spring was directly resting on the trace.

I understand the reasoning that if the sharp end of the spring is directly over the trace, a sharp impact could cause the spring end to puncture the mask over the trace and cause a short. But if I am reading DBSAR correctly, the end of the spring on his light was NOT over the trace, but was at 2 o’clock. So, it was the smooth section, not the end, of the spring that was over the trace. Is it plausible in that circimstamce for the spring to have punctured the mask in a one foot fall?

If it is, then virtually all the FW3A’s would be at risk (unless the end of the spring happened to be soldered to the pad at 8 o’clock, which is something no one has yet reported seeing). Both my lights have the spring end over the 4 o’clock pad, with the spring touching the mask over the trace, and no room for sliding in a plastic insulator.

Am I understanding this issue correctly?

Hmmmmm…. good point CLB. … :+1:

it seems some have reported their springs to be in random positions, some with only one of the three pads soldered, some have two, and indeed mine had the spring edge roughly at the 3 o’clock position, but the spring surface was still against the trace mask, but only had the 8 o’clock pad soldered. it was a bunch of unlucky circumstances to what happened to mine, if toppled over off a 20” high night table, landed tail down hard on a hard-wood floor with a sharp “clunk” then it instantly turned on and began to ramp up to maximum mode on its own. I guess the weight of the 30Q had slammed down hard enough on the spring to force it through the mask to the trace. After that i don’t know what happened as i kept trying to turn it off with the switch to no avail. then tapped it a couple times on the edge of the table, then it went out but still flickered, it was really hot after running on max for the minute or so, i then began to unscrew the extremely hot head section to pull the battery out. I have no idea if something else also shorted in the head, but something obviously sent a ton of amps current through that shorted trace to burn & melt it like a fuse.

Mine has the edge at 3 o’clock with all 3 pads soldered. The pad at 4 o’clock has a lot more solder than the other 2, especially the one at 8 o’clock, so the bottom of the spring over the trace is raised up and won’t make contact with it even when I push down hard on the spring.

This is a good thing I guess

[DBSAR]

@SIGShooter

Mine has the edge at 3 o’clock with all 3 pads soldered. The pad at 4 o’clock has a lot more solder than the other 2, especially the one at 8 o’clock, so the bottom of the spring over the trace is raised up and won’t make contact with it even when I push down hard on the spring.

This is a good thing I guess

Sound like yours is good and solid with a bit of a gap over the trace with no risk of spring movement. :+1:

[contactcr]

While you can never be too safe with the spring it doesn’t seem like we are convinced this is the only variable that caused this.

Did you ever take a good picture of the driver side? Does the signal pad line up nicely where your driver is seated? Kawi had one that didn’t so rest assured i’m not just trying to agitate you DBSAR.

[Lexel]

Aux boards for the FW3A are now on sale

Prototype with ice blue, pink and warm white

[SALE] FW3A lighted TIR board multicolored with LVP, stabilized or High/low

how it began can be read here Boards can be build as Gen 1 to 1.7 now Generation 1 High/low capability but then LEDs dim to about 50% between full and empty battery Generation 1.2 LEDs are voltage stabilized High/low does not work, no LVP Generation 1.5 unregulated LED voltage + LVP Generation 1.7 stabilized with LDO + LVP [image] - version with LDO → High/low not possible version with LVP → High/low not possible → flash #define INDICATOR_LED_SKIP_LOW setting as well LDO → constant b…

[DBSAR]

@contactcr

While you can never be too safe with the spring it doesn’t seem like we are convinced this is the only variable that caused this.

Did you ever take a good picture of the driver side? Does the signal pad line up nicely where your driver is seated? Kawi had one that didn’t so rest assured i’m not just trying to agitate you DBSAR.

i didn’t take a photo or took the head apart yet, but its looks slightly off center, but no burn marks.

[pennzy]

Mine is only soldered in one place. I guess I will solder the other two , maybe rotate it also if no gap.

[KawiBoy1428]

@contactcr

While you can never be too safe with the spring it doesn’t seem like we are convinced this is the only variable that caused this.

Did you ever take a good picture of the driver side? Does the signal pad line up nicely where your driver is seated? Kawi had one that didn’t so rest assured i’m not just trying to agitate you DBSAR.

Yep the driver is quite smaller than the driver seat/pocket…some measurements…

seat ID

driver diameter

driver flat diameter

batt tube ID

driver smd/clearance pocket ID

The 2 GB and the 1 X-L HI 1A lights, the drivers are a press fit, zero slop…… this is one of two of the Nichia Lights….

[JasonWW]

@DBSAR

no doubt, from what i can gather from this, is that if the spring shorts to the trace from an impact of abrasion over time if the spring cut edge is pressing on that trace, the light will turn on from that short and begin ramping up to its maximum mode and can not be turned off until the battery is removed.

This does not seem likely. If the spring touches the trace, it’s like pushing the button. It ramps up, but then the thermal stepdown should activate.

What your describing, the thermal stepdown not activating, sounds like a driver short where the driver is bypassed and no longer in control of the LED’s.

Did you check out your driver/head area to see if anything looks shorted?

[JasonWW]

I think I’ll try shorting the spring to the trace to try and duplicate the sceneries described. I’ll scrape away a bit of the solder mask and solder a single tiny strand of a copper wire from the trace to the spring. I’ll see if it burns through like a fuse or what exactly. It might be next week before I can do it.

[teacher]

@JasonWW

I think I’ll try shorting the spring to the trace to try and duplicate the sceneries described. I’ll scrape away a bit of the solder mask and solder a single tiny strand of a copper wire from the trace to the spring. I’ll see if it burns through like a fuse or what exactly. It might be next week before I can do it.

. Whoa…. .
That’s “taking one for the team” to a whole new level!!

[JasonWW]

@DBSAR

if your trying to prove i am faking this or call-me-out on what happened to my FW3A , you are so very damn wrong, but good luck is what i have to say on your experiment to prove what you want to do.

Not at all. Like I said earlier, in your case it sounds like the driver got bypassed somehow. From what I can see as to how the switch pcb is designed, shorting the spring to the trace should only simulate the button being held down. The thermal stepdown should activate. I assume it will, unless a held down button somehow bypasses the stepdown software. Toykeeper probably knows what the code will do.

High current usually goes through the vias on the pcb. Maybe the trace is weaker in comparison to the vias and that’s why it got overheated. I really don’t know. It’s a mystery.

I figure using a tiny wire strand will act as a fuse so I don’t risk burning the trace. I can fix the trace if it should burn, but that’s extra work. :confounded:

[Rdubya18]

I can’t lose this nagging feeling that this is an effect of a loose driver retaining ring . I am in no way as knowledgeable as most of you , but the lights behavior when the ring is loose is exactly how DBSAR’s did . Is it possible with the current running wild , and a shock ( blow ) to the tail at that moment , caused the trace to short ?

[contactcr]

@JasonWW

@DBSAR

if your trying to prove i am faking this or call-me-out on what happened to my FW3A , you are so very damn wrong, but good luck is what i have to say on your experiment to prove what you want to do.

Not at all. Like I said earlier, in your case it sounds like the driver got bypassed somehow. From what I can see as to how the switch pcb is designed, shorting the spring to the trace should only simulate the button being held down. The thermal stepdown should activate. I assume it will, unless a held down button somehow bypasses the stepdown software. Toykeeper probably knows what the code will do.

High current usually goes through the vias on the pcb. Maybe the trace is weaker in comparison to the vias and that’s why it got overheated. I really don’t know. It’s a mystery.

I figure using a tiny wire strand will act as a fuse so I don’t risk burning the trace. I can fix the trace if it should burn, but that’s extra work. :confounded:

No idea why he is so defensive about this but i’m with you i’d rather know that we found the real cause and not a side effect of the primary issue. Sending people down some rabbit hole of spring soldering only to find out the signal tube bridged something on a loose driver (hypothetically) is counter productive. If a revision is to be made to assembly or otherwise the real cause needs to be determined.

[pennzy]

No harm in getting to the bottom of this. I’d like to know if I need to lock mine out or exactly what mod is needed or if this is just a fluke.

[contactcr]

@pennzy

No harm in getting to the bottom of this. I’d like to know if I need to lock mine out or exactly what mod is needed or if this is just a fluke.

That’s the thing with e-switch though. Firmware lockout does nothing if it shorts. Physical lockout is hard to do on this light and not recommended anyways cause nothing is glued. Until we know more I would suggest just making sure the driver is lined up and centered and the retaining ring is not loose.

[JasonWW]

DBSAR, did you change your light to ramp to Turbo or does it only ramp up to 1000 or so lumen like stock?

I was testing to see if the thermal protection would still kick in when you hold the button permanently, but I forgot Anduril doesn’t ramp to Turbo (at least not in stock form).

[JasonWW]

Okay, I found time to test this today. Nothing bad happened. It just ramps up to the stock level. No Turbo and no high current going through the tail circuit. There’s no need to worry about the spring shorting the trace.

I took some pics and a video to show what happens. I’ll put it in the next post.

My point is something else happened with DBSAR’s light that caused it to go to Turbo and burn that trace. This is still a mystery. I suspect something in the driver came loose due to the fall and the driver got shorted somewhere in the head. I don’t really know, I’m just trying to eliminate things to narrow in on the true cause.

[WalkIntoTheLight]

Interested to find out the root cause, if not just the tail trace. Still, it sounds like it’s worth soldering the spring properly, to prevent it from damaging the tail trace.

[Nev]

It may be just a bad design & did what it did because this torch can produce loads of lumens but it’s not really built to handle it , mine started smoking & melted the feeble wires , I think lemontop ought to watch out , because selling this little time bomb to muggles is gonna end up in disasters & people getting hurt.

[JasonWW]

Here are the pics.

Copper strand added.

Here is the video. Sorry about the length.

Note, at 6:10 it seems the inner tube temporary lost contact and reconnected. This simulated the switch being released and pressed again quickly. You can see it ramp down.

FW3A Test - Spring Shorted to Trace

FW3A Test - Spring Shorted to Trace

Something else I just noticed. Most of you probably already knew this, but I didn’t. The metal dome has 2 tabs that keep it located. I thought the tape held it in position which seemed fragile. In reality, the tabs lock it into place and the tape is only there to aid assembly. Neat.

.