FW3A Troubleshooting / FAQ

1656 posts / 0 new
Last post
Firelight2
Firelight2's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 04/08/2011 - 15:17
Posts: 5091
Location: California

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

Ah.

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

SammysHP
SammysHP's picture
Offline
Last seen: 1 hour 12 sec ago
Joined: 06/25/2019 - 14:35
Posts: 1718
Location: Germany

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

idleprocess
idleprocess's picture
Offline
Last seen: 1 hour 15 min ago
Joined: 08/17/2018 - 00:22
Posts: 93
Location: dfw.tx.us

xevious wrote:
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?


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.
MtnDon
MtnDon's picture
Offline
Last seen: 34 min 46 sec ago
Joined: 08/27/2015 - 18:25
Posts: 4403
Location: Canuk in NM

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.

xevious
xevious's picture
Offline
Last seen: 4 days 15 hours ago
Joined: 02/27/2013 - 21:55
Posts: 3394
Location: Hoboken, NJ USA

cyclops wrote:
Is it the first batch version or later ones? I hear that the later versions were having unreliable quality.
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.
Firelight2 wrote:
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.
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!
Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

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?

raccoon city
raccoon city's picture
Offline
Last seen: 50 min 47 sec ago
Joined: 10/06/2010 - 02:35
Posts: 19534
Location: Palm Desert CA ~ West Coast is the best coast

I hope you have fun here, Alexxey!

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

Thanks a lot Smile

JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

Alexxey wrote:
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?

Mainly just count all the 7135 chips. There’s 6 on the back and 2 on the front, but the front side might require unsoldering the 2 main wires to look at it. You can check that the brightness when ramping up or down is smooth.
There should be a short dip in power as you ramp between the 7135 chips and the FET. The newer firmware might be different though. Let’s see what other folks say. (My light is a very early one)

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

JasonWW wrote:

Mainly just count all the 7135 chips. There’s 6 on the back and 2 on the front, but the front side might require unsoldering the 2 main wires to look at it. You can check that the brightness when ramping up or down is smooth.
There should be a short dip in power as you ramp between the 7135 chips and the FET. The newer firmware might be different though. Let’s see what other folks say. (My light is a very early one)

All 6 are in the back, but don’t want to unsolder anything to count those on the front. Smile
Ramp is all smooth and there are two ‘blinks’, one on the high and one in somewhere in between.
JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

Alexxey wrote:
JasonWW wrote:

Mainly just count all the 7135 chips. There’s 6 on the back and 2 on the front, but the front side might require unsoldering the 2 main wires to look at it. You can check that the brightness when ramping up or down is smooth.
There should be a short dip in power as you ramp between the 7135 chips and the FET. The newer firmware might be different though. Let’s see what other folks say. (My light is a very early one)

All 6 are in the back, but don’t want to unsolder anything to count those on the front. Smile
Ramp is all smooth and there are two ‘blinks’, one on the high and one in somewhere in between.

I checked mine. When ramping up there are two short blinks. One when it switches from channel 1 to 2 and another from 2 to 3.
I’m still not sure that’s a guarantee you have 3 channels in yours. I’m not even sure which software version your light has. Hopefully someone can give you a more concrete answer.

If the ramp is smooth I wouldn’t even worry about it.

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

JaredM wrote:
80-85C. Mostly for safety of the battery.

Is it safe for leds, driver etc? Won’t affect flashlight lifespan?
Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

JasonWW][quote=Alexxey][quote=JasonWW wrote:

If the ramp is smooth I wouldn’t even worry about it.

Ok thank you nice to hear.
JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

Alexxey wrote:
JaredM wrote:
80-85C. Mostly for safety of the battery.

Is it safe for leds, driver etc? Won’t affect flashlight lifespan?

The led and driver components get heated to over 200°C during assembly. They are quite robust.

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

SammysHP
SammysHP's picture
Offline
Last seen: 1 hour 12 sec ago
Joined: 06/25/2019 - 14:35
Posts: 1718
Location: Germany

JasonWW wrote:
Alexxey wrote:
JaredM wrote:
80-85C. Mostly for safety of the battery.

Is it safe for leds, driver etc? Won’t affect flashlight lifespan?

The led and driver components get heated to over 200°C during assembly. They are quite robust.

But only once and with a very specific heating profile. Otherwise they will get damaged. 80 °C is a pretty safe temperature, though.
JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

SammysHP wrote:
JasonWW wrote:
Alexxey wrote:
JaredM wrote:
80-85C. Mostly for safety of the battery.

Is it safe for leds, driver etc? Won’t affect flashlight lifespan?

The led and driver components get heated to over 200°C during assembly. They are quite robust.

But only once and with a very specific heating profile. Otherwise they will get damaged. 80 °C is a pretty safe temperature, though.

They can actually handle higher heat than that, but briefly and people have ran the leds to the point of the heat melting the solder. My point was that 85° is a good upper level to not exceed due to the battery. Then there is a huge gap from 85°C to 200°C. So no one really needs to worry about the led or driver components. Worry about battery heat first (and not burning your hand). Don’t waste time worrying about the driver and led. They are quite robust.

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

JasonWW, SammysHP, thank you for detailed explanation!

JasonWW][quote=Alexxey][quote=JasonWW wrote:

If the ramp is smooth I wouldn’t even worry about it.

Actually I checked one more time, ramp is not so smooth near to moonlight level. I can see several barely visible steps (starting from moon) before it gets nice and smooth. Is it same for original version or is it a bad sign?
JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

It’s not very likely you got an old original model with all 3 channels. It’s not the newest version for sure either which is one channel (no 7135 chips). So you might have that middle version. It’s hard to say for sure. Even my early model is not perfectly smooth around the very lowest levels.

You should decide if whatever version you have is going to work for your needs and not worry so much how many channels you have. That is a tiny detail that makes a tiny difference. Even a FET only version should be pretty decent except for the very lowest levels around moonlight. It’s really not that big a deal as long as it’s operation and functionality is good.

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

JasonWW wrote:
It’s not very likely you got an old original model with all 3 channels. It’s not the newest version for sure either which is one channel (no 7135 chips). So you might have that middle version. It’s hard to say for sure. Even my early model is not perfectly smooth around the very lowest levels.

You should decide if whatever version you have is going to work for your needs and not worry so much how many channels you have. That is a tiny detail that makes a tiny difference. Even a FET only version should be pretty decent except for the very lowest levels around moonlight. It’s really not that big a deal as long as it’s operation and functionality is good.


Yeah I know, it’s just some deep desire of a perfect toy Smile
For sure it’s not the old original version, but what if Lumintop decided to make fw3a great again? I guess there is no way to tell other than unsolder something.
Comments like this made me worried in the first place actually Smilehttps://budgetlightforum.com/comment/1868702#comment-1868702
JasonWW
JasonWW's picture
Offline
Last seen: 3 days 18 hours ago
Joined: 10/22/2016 - 11:41
Posts: 12941
Location: Houston Texas

I get the feeling that was Blueswords way of saying he’s disappointed in Lumintop.
If he tried a modern one he may think it was fine.

Texas Ace Lumen Tube and JoshK Sphere calibrated with Maukka lights

Click this to go to signature links. I'm still around, just not reading many new threads.

Firelight2
Firelight2's picture
Offline
Last seen: 4 hours 7 min ago
Joined: 04/08/2011 - 15:17
Posts: 5091
Location: California

JasonWW wrote:
It’s not very likely you got an old original model with all 3 channels. It’s not the newest version for sure either which is one channel (no 7135 chips). So you might have that middle version. It’s hard to say for sure. Even my early model is not perfectly smooth around the very lowest levels.

You should decide if whatever version you have is going to work for your needs and not worry so much how many channels you have. That is a tiny detail that makes a tiny difference. Even a FET only version should be pretty decent except for the very lowest levels around moonlight. It’s really not that big a deal as long as it’s operation and functionality is good.


I thought there were basically 4 versions of the FW3A made:
  • First version: 7+1+FET driver, no tailcap retaining ring. <— good
  • Second version: 7+1+FET driver <— best
  • Third version: 1+FET driver <— bad (not efficient at mid-to-low power settings).
  • Fourth version: FET driver <— worst (not efficient at mid-to-super low power settings and lacks moonlight).

So basically, only the 1st and 2nd versions have the best driver. The 1st and 2nd versions should have 6 chips visible on the underside of the driver when the head is removed from the body.

The 3rd and 4th versions should have a flat underside of the driver with no visible chips. The final 7135 chip is on the topside of the driver so can’t be seen without removing the driver. However, a user should easily be able to distinguish a 3rd from 4th version without disassembly simply by setting to the lowest possible ramp. The 3rd version will have a really low moonlight with which you can easily stare into the emitters while the light is on from close range with dark-adapted eyes without discomfort. The 4th version won’t have moonlight and will be too bright to comfortably stare into the emitters at their lowest setting.

The original poster of this thread described looking into the bottom of the head and seeing 6 chips on the underside of the driver. This means he should have either the first or second version, … meaning he got a good one.

At least that’s my recollection.

pennzy
pennzy's picture
Offline
Last seen: 2 hours 40 min ago
Joined: 12/10/2017 - 19:45
Posts: 4546
Location: United States , Pa.

Sounds right to me.

MtnDon
MtnDon's picture
Offline
Last seen: 34 min 46 sec ago
Joined: 08/27/2015 - 18:25
Posts: 4403
Location: Canuk in NM

Four basic hardware differences sound right to me too. To those, you can add firmware variations. I have no idea how many versions were applied to production lights. I have an early April 2019, first batch with no tailcap retainer ring as well as a June 2019 version and there is a slight difference in what happens when the lights are e-locked and the switch button depressed and held.

The April ’19 light comes on in moonlight no matter how many clicks the switch gets before being held down. The June ’19 light produces genuine moonlight mode when given a 1H. Or 3H, 5H, etc (odd numbers). When the switch is given a 2H, 4H, 6H, etc. (even numbers) the light is on at a brighter level. I prefer that. It gives me a choice depending on the circumstances. That June light operation pleased me so much that I reflashed the April ’19 light with the June ’19 firmware. (That is the behavior when set to smooth ramping. When set to stepped-level operation the first, 1H produces the brighter level and the 2H produces the moonlight.) Both lights use that June ’19 firmware to this day. It does what I want and I have seen no need to change to any later version.

Scallywag
Scallywag's picture
Offline
Last seen: 18 hours 12 min ago
Joined: 01/11/2018 - 22:23
Posts: 2450
Location: Ohio, United States

I’ve ended up with two of the elusive purple TK-editions. Both have the plastic tail retaining ring, two-step lockout brightness, and full three channels.

I haven’t heard of any middle revisions with 6 7135s on the underside that don’t have three channels. So it sounds like a good one.

I rarely use them anymore. Eventually the switches got unreliable, I believe due to tube contact issues. My KR4 took over their duties. The sprung tube is just superior.

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

Firelight2 wrote:
JasonWW wrote:
It’s not very likely you got an old original model with all 3 channels. It’s not the newest version for sure either which is one channel (no 7135 chips). So you might have that middle version. It’s hard to say for sure. Even my early model is not perfectly smooth around the very lowest levels.

You should decide if whatever version you have is going to work for your needs and not worry so much how many channels you have. That is a tiny detail that makes a tiny difference. Even a FET only version should be pretty decent except for the very lowest levels around moonlight. It’s really not that big a deal as long as it’s operation and functionality is good.


I thought there were basically 4 versions of the FW3A made:
  • First version: 7+1+FET driver, no tailcap retaining ring. <— good
  • Second version: 7+1+FET driver <— best
  • Third version: 1+FET driver <— bad (not efficient at mid-to-low power settings).
  • Fourth version: FET driver <— worst (not efficient at mid-to-super low power settings and lacks moonlight).

So basically, only the 1st and 2nd versions have the best driver. The 1st and 2nd versions should have 6 chips visible on the underside of the driver when the head is removed from the body.

The 3rd and 4th versions should have a flat underside of the driver with no visible chips. The final 7135 chip is on the topside of the driver so can’t be seen without removing the driver. However, a user should easily be able to distinguish a 3rd from 4th version without disassembly simply by setting to the lowest possible ramp. The 3rd version will have a really low moonlight with which you can easily stare into the emitters while the light is on from close range with dark-adapted eyes without discomfort. The 4th version won’t have moonlight and will be too bright to comfortably stare into the emitters at their lowest setting.

The original poster of this thread described looking into the bottom of the head and seeing 6 chips on the underside of the driver. This means he should have either the first or second version, … meaning he got a good one.

At least that’s my recollection.


Thank you for such a detailed explanation!
Looks like i got 2nd version despite the fact that I bought it last week on aliexpress (SST-20 black). Or maybe there are 5th version with all channels back in place?
Also the button is all great, clicky and reliable. Hope it will stay that way.

Also I got 2 step brightness on lockout. Moon on 1H, somewhat brighter on 2H. But none of the 3H, 4H etc doesn’t work (I mean it’s all moon).

And 1 more thing. I also got bunny on the button. Whed do they start doing it?

xevious
xevious's picture
Offline
Last seen: 4 days 15 hours ago
Joined: 02/27/2013 - 21:55
Posts: 3394
Location: Hoboken, NJ USA

Does kapton tape solve shorting problems?

Recap:
My FW3A with 18350 tube would periodically start shorting out after loosening/tightening the tube for physical lockouts. I’d have to “fiddle” with it… wiping down the tube, using a cotton swab on the mcpcb… and usually that would work. But then I had a bizarre instance of my FW3A suddenly turning itself on to turbo in the middle of the night (thankfully I’d pointed it towards me before falling asleep, so the bright beam woke me up before it seriously overheated). THAT was the last straw. I stopped using it.

Someone had suggested applying some kapton tape to the very end of the tube. Kapton is strong, extremely thin tape that insulates as well if not better than your typical electrical tape. I’d misplaced my roll… couldn’t recall where I’d put it. So periodically I’d look around. FINALLY, I found it. So I applied kapton tape to the tube, aligning the edge of the tape to the edge of the forward end of the tube and made one good 360 degree wrap with some overlap. It seems to work. I can loosen and tighten the FW3A repeatedly and it works fine.

The only exception I’m finding is that if I screw down the tube extra tight, it shorts out. As long as I don’t screw down quite that firmly, it’s OK.

wle
wle's picture
Offline
Last seen: 2 hours 20 min ago
Joined: 01/07/2015 - 13:49
Posts: 3237
Location: atlanta ga

i just put 3 narrow strips of black tape at the 2 ends of the ‘signal tube’

[a whole 3/4” wide circle of tape, was too thick to go back in the light]

wle

"You never have the wind with you - it's either against you, or you're having a good day."
    Daniel Behrman, "The Man Who Loved Bicycles".
It never gets easy, you just go faster.   
-Greg Lemond.
       ,ø¤º°`°º¤ø¸,ø¤º°`°º¤ø¸,ø¤º°`°º¤ø¸

xevious
xevious's picture
Offline
Last seen: 4 days 15 hours ago
Joined: 02/27/2013 - 21:55
Posts: 3394
Location: Hoboken, NJ USA

wle wrote:
i just put 3 narrow strips of black tape at the 2 ends of the ‘signal tube’

[a whole 3/4” wide circle of tape, was too thick to go back in the light]

I actually wrapped the whole circumference of the signal tube front end with kapton tape. Just snug enough to still enter the chamber. The tail end of the tube doesn’t have a metal retainer, so there’s no issue there. But actually I’m thinking I should apply a few strips along the length of the tube to help a little more with centering.
Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

Hi! I have a very strange issue. I guess I mess something up and now I have only 1 brightness level when locked (moon). 2h doesn’t work anymore (it’s also moon).
How is this even possible and how can I fix it?

Alexxey
Offline
Last seen: 3 months 2 weeks ago
Joined: 07/22/2022 - 04:23
Posts: 11

OMG I found the answer https://budgetlightforum.com/node/68302
So 2h gives you floor of stepped mode. That’s smart, too bad it’s not in the manual. Smile

Pages