Suggestions on how to improve Armytek Flashlights:

[scs]

Re: Share your opinion with us!

The full chart of parts' compatibility will be available later when we finish the production of Partner v3 & Prime v2 (customized with XP-L LEDs).

Hope you're considering the XP-L High Intensity and a TIR optic optimized for it, not just the regular version behind the current TIR optic. It will give new appeal to the current line. Also consider starting a thread to take suggestions for the Partner and Prime model updates.

 

[RetroTechie]

Re: Share your opinion with us!

High temperatures will!!! damage the led. Flashlights are somewhat limited in their ability to dissipate heat.

True, but that ability to dissipate heat depends on the environment. If tail standing (free air), it's limited. If hand held, hand serves as extra heatsink and more heat can be shed.

You could test a variety of usage conditions, and see what is "low end" and "high end" in terms of how many watts can be shed. If you design for a worst case situation, that means you leave potentially higher output on the table (FWIW: I'd consider that a pretty good design choice!). If you 'think positive' and design for good cooling conditions, light will be hotter than desired (or even safe) when those conditions are sub-optimal.

For that reason, temperature control sounds nice. Safely power up when possible, limit to safe levels when cooling conditions are poor. But I don't understand why you would cycle this: when lower than what's safe (both averaged, or during a cycle), again you're leaving output capacity on the table. Which was the whole point of a temperature check, right?

So if there's temperature control, I'd simply regulate output such that the light creeps towards a temperature that's considered safe. If enough cooling capacity is available, that would go up to the specified max output. If not enough cooling, then it would be limited by whatever cooling capacity is available. Preferably small steps, so you don't consciously notice the beam adjusting as cooling capacity changes.

It's like computer cooling, there's 3 basic strategies: (with perhaps some mixing possible)
1) Pick a max temperature, and adjust CPU clock speeds etc, such that power levels matches cooling.
2) Pick an (arbitrary) max power level, constant fan speeds etc, and live with temps going up or down as load varies. Obviously max temp will be reached when ambient temperature is high and PC is doing much work. Pick those max power levels & fan speeds such that things don't get out of hand under maximum load conditions. This is my personal preference for PC cooling.
3) Pick a max temperature, and adjust cooling to maintain that as CPU load varies. This means fans constantly spinning up or down - very annoying.

Unless you include a fan in the flashlight, :devil: option 3) isn't available. Personally I don't think I'd be interested in a light with temperature sensor - I prefer to dial down max output level to something that I know is safe. Also a temperature sensor adds complexity, cost and idle current drain (however little). That said: the tech (and regulation strategies) is interesting.

 

[WalkIntoTheLight]

Re: Share your opinion with us!

Ideally, you'd start the light in max, and then monitor the increase in internal temperature. If it's rising quickly, then slowly decrease output until the temperature maintains steady just below the critical level. If it's rising slowly, hold off on doing anything.

This has the advantage of being able to very gradually decrease output, rather than having to do it all-of-a-sudden when the critical threshold is reached. It's adaptive, depending on how quickly the light is able to shed heat. It only decreases output just enough to maintain temperature, it doesn't overshoot when dropping output.

The disadvantage is that it may be too complex to put in the electronics available for a flashlight. And, it may start dropping output sooner than would be done if you wait for the critical temperature to be reached.

 

[scs]

Re: Share your opinion with us!

I'm beginning to think AT perhaps should keep it simple, lest instances of malfunction due to poor or improper design increase.

 

[FlashlightR]

Re: Share your opinion with us!

+ 1 !

 

[recDNA]

Re: Share your opinion with us!

I cannot imagine preferring a timer. I love current system

 

[TheShadowGuy]

Re: Share your opinion with us!

Ideally, you'd start the light in max, and then monitor the increase in internal temperature. If it's rising quickly, then slowly decrease output until the temperature maintains steady just below the critical level. If it's rising slowly, hold off on doing anything.

This has the advantage of being able to very gradually decrease output, rather than having to do it all-of-a-sudden when the critical threshold is reached. It's adaptive, depending on how quickly the light is able to shed heat. It only decreases output just enough to maintain temperature, it doesn't overshoot when dropping output.

The disadvantage is that it may be too complex to put in the electronics available for a flashlight. And, it may start dropping output sooner than would be done if you wait for the critical temperature to be reached.

I was thinking of something similar. A calculus- based algorithm using the slope of the change in temperature modified by the proportion of the current temperature to the max temperature might work. So it would start on max and very slowly decrease output as temperature rises, increasing the change as the temp rises toward max in order to reduce the change in temperature; and should have a point where it levels off and has a relatively constant temp, but changing ambient conditions would allow for higher brightness as temp tries to fall, and in ideal conditions would almost flatline at a maximum brightness for safe temps.
I kind of want to draw a graph to illustrate... Did any of that make sense?

 

[magellan]

Re: Share your opinion with us!

I was thinking of something similar. A calculus- based algorithm using the slope of the change in temperature modified by the proportion of the current temperature to the max temperature might work. So it would start on max and very slowly decrease output as temperature rises, increasing the change as the temp rises toward max in order to reduce the change in temperature; and should have a point where it levels off and has a relatively constant temp, but changing ambient conditions would allow for higher brightness as temp tries to fall, and in ideal conditions would almost flatline at a maximum brightness for safe temps.
I kind of want to draw a graph to illustrate... Did any of that make sense?

I'm no expert in this (the experts can feel free to correct me if I err), but the problem with the calculus approach is you're dealing with a series of discrete measurements which lack continuity. My understanding is a least squares approach or some other similar discrete technique works well in these circumstances.

 

[WalkIntoTheLight]

Re: Share your opinion with us!

I'm no expert in this (the experts can feel free to correct me if I err), but the problem with the calculus approach is you're dealing with a series of discrete measurements which lack continuity. My understanding is a least squares approach or some other similar discrete technique works well in these circumstances.

Yeah, while an adaptive/calculus/statistical algorithm has its advantages, I suspect it's not possible to program into a light. The light IC might not have RAM memory, which you'd need to store prior temperatures measurements to implement such an algorithm.

While the light probably has EPROM memory (Armytek has programability and mode-memory which probably uses EPROMs), you wouldn't use EPROMs for a temperature algorithm because you risk wearing out the ERPOM from too-frequent use.

So, a simpler approach might be to just have a couple of temperature thresholds, a warning and a critical. At warning, the light gradually decreases output. At critical, the light steps-down much more dramatically.

 

[BLUE LED]

Re: Share your opinion with us!

Will Armytek be using the XP-L HI LED in your lights? I am hoping it will be used in the Predator

 

[ArmyTek]

Re: Share your opinion with us!

Will Armytek be using the XP-L HI LED in your lights? I am hoping it will be used in the Predator

Yes, we are planning to release a limited edition of Predator on XP-L HI LED approximately in August.

 

[recDNA]

Re: Share your opinion with us!

Safe for use with 2 x CR123A primaries? Will it outthrow my XP-G2 Predator Pro?

 

[BLUE LED]

Re: Share your opinion with us!

Yes, we are planning to release a limited edition of Predator on XP-L HI LED approximately in August.

I was wondering if there was an update to the Predator XP-L HI?

 

[scs]

Re: Share your opinion with us!

Safe for use with 2 x CR123A primaries? Will it outthrow my XP-G2 Predator Pro?

assuming the over 30k lux for the Viking v3 is accurate, and the xpl hi roughly doubles the CD, it in the tuned predator reflector should easily out throw its predecessor.

 

[scs]

Re: Share your opinion with us!

The hotspot might be bigger than that of the xbh and the overall beam might be less specialized for hunting and more friendly for general spotting.

 

[texas cop]

Re: Share your opinion with us!

I would like to see a 26650 format light added to the collection. At night I often use a light so much that its dimming greatly by morning. With a little increase in girth Capacity would jump from 3500 mah an the best 18650's to 5500 mah on the best 26700's. I much prefer single cell length lights for the pocket or belt. Any longer tend to get left in the truck.

 

[Raysbeam]

Re: Share your opinion with us!

Thank you all for your opinions and advice! That has given us the ideas and grounds for reconsideration of thermal control in our lights =) We are going to discuss the possibilities with our engineers and make some changes in future flashlights. What conclusions we can draw now:


1. Of course we are not going to use any timers! Constant brightness is the main idea we're going to "fight for till the end".


2. Thank you all for your opinions and advice! That has given us the ideas and grounds for reconsideration of thermal control in our lights =) We are going to discuss the possibilities with our engineers and make some changes in future flashlights. What conclusions we can draw now:


1. Of course we are not going to use any timers! Constant brightness is the main idea we're going to "fight for till the end". We will definitely keep the full stabilization so that in the conditions of good air-cooling (in winter, on a bike or underwater etc.) Pro-flashlights will deliver constant light even with almost discharged battery.


2. We use the similar algorithm to maintain the temperature under the critical limit as Zebralight's PID but in our lights brightness "ups & downs" are larger and thus noticeable. We will make them smaller so they'll be gradual & unnoticeable. That will enable us to provide the maximum allowable brightness as well as keep the flashlight and your hands (or head) safe =)
3. We do not plan to decrease the Max outputs in this situation as that what is very important for many of our customers. We have plenty of models with different brightness & power sources so everyone can choose what he prefers. The thermal management issue regards mostly to high-power & compact lights with constant brightness such as Wizard or Viking Pro, that are too bright not to overheat & and too small to loose heat fast enough.

This is another reason your lights are the best, in my opinion. I have always been disappointed when Chinese companies claim 1000lm for 2h and after 3 mins my light was running 230lumens, sold them and will only buy ArmyTek from now on.

 

[Raysbeam]

Re: Share your opinion with us!

I would like to see a 26650 format light added to the collection. At night I often use a light so much that its dimming greatly by morning. With a little increase in girth Capacity would jump from 3500 mah an the best 18650's to 5500 mah on the best 26700's. I much prefer single cell length lights for the pocket or belt. Any longer tend to get left in the truck.

I always thought this also but now I'd rather the 18650 battery tube and bring a second cell and have 7000mAh.

 

[Phlogiston]

Re: Share your opinion with us!

Fair enough, but that gets into the realm of what you're willing to carry around with you and whether you're willing to accept the risk of being caught out in some way as you swap the cell. Besides, you could bring a spare cell for the 26650 light too, and have 10000mAh or more.

26650 lights also open the option of having a simple adaptor that lets you put an 18650 in at a pinch, whereas you can't use a 26650 with an 18650 light at all.

I see a place for both formats. In fact, I'd even be interested in a 32650 light, if I could get decent cells with a worthwhile increase in capacity.

 

[clemence]

Re: Share your opinion with us!

TIR optics please. Even with such high impact rating, I still have no confidence dropping a flashlight with glass part in it. I'd feel far more secure if the entire optical parts are made from PMMA or silicone lens. Imagine drop it with the glass facing any pointy shaped hard object....
TIR optics also give better overall control for designer to create limitless beam shape with minimal lumen output loss, let alone the better tint mixing. I imagine a super tight Predator with specialized TIR optics and XP-L HI neutral tint comes bundled with 40 & 80 degree Luminit difusser.
For EDC and headlamps product line consider using higher CRI LED such as the Nichia 219 series. Maybe three emitters with reasonable current would get about the same brightness and efficiency with superior CRI compared to one high power low CRI emitter pushed too hard. I often do many color related works at night using headlamp as my only light source. Sometimes I paused just to make sure the stripes in a bunch of mud covered cables I work with were the right color.