FW3A Troubleshooting / FAQ

I doubt the flickering is related. That is usually due to a contact problem with the head not being tight enough or the battery length being wrong, etc…

The leds are soldered to the MCPCB (metal core printed circuit board). The MCPCB is pressed against the metal shelf machined into the head. The driver is on the other side of the shelf next to the battery. The tir lens and threaded bezel push the mcpcb against the shelf.

The led turning blue definitely sounds like it got overheated and that unsoldered the wire. Why, I’m not sure. I assume the bezel was tight and putting pressure on the MCPCB. It’s possible there is a piece of debris (maybe tiny solder blob) or maybe a metal burr under the MCPCB preventing it from sitting flat.

Shipping it back typically costs more than the light (a small insured package is about $40 USD from USA to China) so I doubt you or they would want to spend that kind of money. Typically they would just send you a new one and you keep the defective one.

It’s a bummer your having so many issues.

Oh yeah, I forget about the newer ones having that design change. If the MCPCB was sitting on top of the screw wouldn’t it sit really high and the bezel have a noticeable gap near the threads?

There’s no need to replace the thermal grease. It may not have even been spread out yet.

Thanks guys.
With that info i believe there is something under the MCPCB, no way i can get it sitting flat on its own. Also feels very dry underneath pushing it around, but it could be a symptom of the foreign mass.
Ill get in touch with Neal. A new light would solve both my issues, i can deal with one stuck tailcap if the others are all ok, as much as it puts a damper on my “lego games”, bound to have atleast one 18650 version of my lights at any given time.

Edit. Thermal paste is not an issue for me, 20 years of overclocking im well versed and stocked :slight_smile:

A new FW3A and extra tailcap is on the way.
Got them to throw an Osram W2 Fireflies EC01 in the box while they were at it :stuck_out_tongue:

I have 2 FW3A SST20 new, was playing with them right out the box, and noticed that in lockout mode, a double click would activate a slightly higher mode, than the standard moonlight. Cool and useful, read that it’s a feature added to the firmware.

Then I went to ramp config and reconfigured to the factory standards (Floor N = 1 and Ceiling N = 21), just to make sure they were configured correctly.

And this double click higher mode in lockout disappeared from both flashlights after this reconfig. I want it back!

The Anduril version is not so new, I can’t check it, doesn’t factory reset with 13 clicks.

Is it a bug?

Check out Mad Max’s video on how to configure this feature.

My understanding is the double-click while in lockout swaps between the floors set for ramped and stepped modes.

It’s possible you just set them both to the same value. If that’s the case, then double-clicking while in lockout does nothing.

You nailed it, thank you!

The voltage shown by BattCheck seems to indicate higher. For example, when the battery voltage is checked by the multi meter or the battery charger, either shows 3.87V. My FW3A shows 4.0V.

There is not much of discussion about BattCheck. I don’t think there is voltage calibration, either.

It is not big difference. However, I am curious to know what the other FW3A’s are doing.

Also, what does 4.0V from BattCheck mean? Does it mean like between 3.95V and 4.04V (rounded to the nearest tenth)? If so, 3.87V should show 3.9V.

There is no battery voltage calibration. It’s not going to be super accurate, but it does give you a good enough idea on voltage.

Not sure about the FW’s, but on my D4S it blinks out 3.9 from 3.99 down to 3.90 and changes to 3.8 at 3.89. Might just be that one light, can’t really say.

Yes. It provides the good enough idea about the voltage. It is useful and nice to have it.

My D18 seems to be more accurate. So, I was just curious about the others.

They all vary to a certain degree. The same is true for NarsilM lights. I don’t think the voltage detection part of the Atiny85 is all that accurate. It’s a 15 year old design.

Each chip is consistent, though. You can rewrite the code and reflash the firmware with a correction factor to make the voltage readout very accurate. This obviously can’t be done in mass production so we are stuck with what we have.

how does the FW3A indicate 4.0V?

??

i see 4.1 a lot and 3.9 - does it just skip 4.0 to avoid ambiguity in the flashing (ie how can you tell it is flashing ‘zero times’?)

wle

In Anduril the zero is a very fast blink. 3.9v is 3 slow then 9 slow. 4.1v is 4 slow then 1 slow. 4.0v on the other hand is 4 slow and 1 fast.

It’s a bit different from NarsilM which simply shows 4.0v as 4 slow blinks then a longer than normal pause and then 4 slow blinks again.

Thank you very much for the information. The tolerance of +/-0.15V is exactly what my FW3A shows. I understand the limitations. It is still a nice feature.

I added voltage calibration to Anduril - wasn't that difficult to do. TK sounded interested in it, but not sure it went anywhere. I have a modeed Anduril from the Dec 2019 base of code, so doesn't have the new temp regulation.

Hi

Is there a chart/table/equation somewhere that shows how many lumens correspond to each step level? Is it damaging to the torch to set a ceiling of the ramp higher than step 130 (say around 1500lm)?

Sorry in advance if this is a newbie question

Thanks for the help

Lumen measurements would differ depending on emitter type and color temperature. As long as the temperature calibration is set correctly there should not be any risk of damaging the light.

Thanks Rayoui. I have the FW3A with Cree XP-L.