FW3A, a TLF/BLF EDC flashlight - SST-20 available, coupon codes public

Thanks Jason , the party strobe is more irritating to me because it’s not bright enough, I can get both strobes to freeze a fan ,although I don’t think they’re stable (in frequency speed )because it needs adjusting now & again to keep it froozed.
Edit
It must be the 3 fans I’ve tried it on that’s not stable in speed.
Unless the frequency is changing slightly as the cell gets depleted or as the light heats up & cool ?

Ordered Ti/Cu, same discount code as Ti and in theTi drop down options. This one should hold up to my rough handling. BLFFW3T

Ordered a full ti version. Probably going to swap in 7a xpl Hi’s to keep the heat down a bit. Strange this led is not an option anymore, seemed popular.

That is a tricky question, for me at least. I have to think about it. The output frequency for the strobe is controlled by the MCU internal clock which I believe recieves power from an LDO (low drop out voltage regulator) at about 1.8v. So the MCU has stable voltage and can output stable frequencies and pulse rates. It has it’s own built in precision clock/timer.

The reduction in brightness as the battery goes down is all due to factors between the battery and led and not due to anything the MCU does. The MCU still pulses the FET and 7135 chips the same rate as always regardless of battery voltage, but the lower battery voltage across the FET and 7135 chips equals less brightness. This is a direct drive design after all.

Boost and Buck drivers actually increase and decrease the voltage going to the Leds which is why the brightness level is stable to a certain point then steps down like a staircase (Boost). With Buck it has stable output until it falls out of regulation then the output slopes down. With direct drive the output starts high and then immediately starts to slope down in a smooth line.

Yes I did, thanks for the confirmation!

It just seemed odd that turbo output was only slightly higher than the max regulated output, which is part of the reason I asked.

Won’t a warmer led make more heat , or maybe you’re comparing it to sst & not XPLHI?

Max regulated is about 2.8A and Turbo is about 9A (depending on the leds and battery). So that’s a 3x jump in amperage. This also results in about a 3x jump in lumens, from 1000 to 3000. This is a big jump (both lumens and amperage) on paper, but not a big jump to your eye. It’s like a typical jump from High to Turbo. You practically need to triple the output for your eye to see a noticable jump in output. It’s just the way out eyes and brain process light. Not as linear as you might imagine.

Yes comparing to the sst 4000k I ordered.

It sure looks like a big jump in output on my light.

I often hear the “logarithmic” rule thrown around here. I’m not sure I completely buy it. It makes a lot of sense for lower outputs, where the pupils of your eye constrict (letting in less light) as the flashlight output gets higher. I’m not sure it makes as much sense for high outputs, where your pupils are already maximally constricted. It doesn’t for me, anyway. On all my lights, I see a big difference between 500 lumens and 1000 lumens, and 1000 lumens and 2000 lumens. It’s extremely noticeable, not just a little bit. Doubling output gets more noticeable the brighter the light is to begin with.

I’m not sure I’d characterize 2000 lumens as “twice as bright to my eyes” as 1000 lumens. But, it’s closer to twice as bright than it is to 1.4x (or whatever the “rule” is).

Some people see it as a small jump and some as a big jump. It can vary from person to person. It’s basically eye balls turning light to electricity and our brain is interpreting the electricity to create an image in our mind. Kind of like tint. Two people can look at the same light and see different tints. I think the general consensus is you need to triple the lumens to make it look twice as bright. You can read more about it here. Weber–Fechner law. I try not to go too in-depth into the science if I don’t have to. Psychoacoustics are more my thing. Hearing is also logarithmic. Same basic principle.

I just want to point out that LEDs will not scale their output linearly with current, especially at higher currents. So three times the current won’t get you three times the output. But Jason is correct about the output: It’s generally accepted that it takes a quadrupling of lumen output to achieve a perceived doubling of brightness. This is a very rough number, as humans are all different and yadda yadda. But the effect is definitely not linear.

Yeah, again, I have to wonder if that “law” was developed in a dimly-lit examining room while testing people’s vision with just a few lumens of light. As I said, I completely believe it for low levels of light. It makes a lot of sense, with constricting pupils and the switch from light-sensitive rods in the retina to less-sensitive cones in the retina.

It makes less sense in bright light.

It’s barely more than double with SST-20 if I recall, so another reason why people might interpret it differently

I’m speaking in general terms of course. I did actually check the measured output of my 3D version with a 30Q at an arbitrary time of about 5 seconds. The results are 970lm and 3000lm. So pretty much a jump of 3x. I was never able to measure the current at the tail cap so I’m quoting another person who measured about 9A (3D and 30Q iirc). My memory is fuzzy. I think it was 9A, maybe it was higher? Maybe it was 10 or more amps?

For anyone who is not familiar with an led output test, here is an example from Djozz.

As current goes up, output will start to taper off. So 3 times the current is usually less than 3 times the output. Like I said, i was just trying to be be very general to try and explain something and inadvertently added complication. :person_facepalming: Hopefully people understand the point I was trying to make.

Tail standing …

I notice that my alum unit tail stands on the button, light can be activated…
My copper units both tail stand on the tail cap shoulder, light cannot be activated.

Is this just a difference in the depth of the button pocket in the tail cap machining tolerance?

That doesn’t sound good. AFAIK, it should tail-stand on the tailcap edges, not the button. Mine does, anyway.
Did you modify the button or reassemble the switch mechanisms?

I think the main point was understood. I just meant to clarify that point.
It is interesting that the top of ramp you measured is only 970 lumens, when I would expect (based on that output chart, at least) something more like 1200 from 1A per emitter.

You lose output with optics, glass, led bin, etc so you really –15% to 20% is pretty common

not at all

It seems unlikely but could the rubber gasket in the switch be missing? It would probably feel way different without it though. I guess the switch pocket could be different.