Agro's 2020 EDC dream

I’ve spent a lot of time this and last year thinking about small zoomies. And ordering small zoomies. :wink:

A couple of days ago I’ve had a concept that I tried to visualize by making a sketch.
It turned out better than I thought it would and so I decided to share it with you. :slight_smile:

Drawing note: I use different shades of grey to denote different parts rather than different material.

So…what it is?
An almost-unambitious 14500 zoomie. :slight_smile:
What do I mean by that?

Typical zoomies have many drawbacks. Among them:

  • poor efficiency in throw mode
  • lack of waterproofing
  • mediocre cooling
  • many moving parts which is a reliability challenge

Here I didn’t try to challenge any of these, I just wanted to draw a zoomie that’s first and foremost simple.
Then I deviated from this path a bit…but in a reversible way.
So this light is not very simple but not very complex either. And it is easy to modify to make it very simple again.

Looking at the picture from the business end you can see a pretty thick lens. It’s Optolife A295MC, AR coated aspheric from a good but inexpensive lens maker.
The LED is Luminus SST-20. Good high CRI emitter that offers fair throw and output. It could be substituted for Luxeon MZ for pure flooder or Boost HL for a pure thrower.

11 mm MCPCB sits on a copper pill-head. I call it pill-head because it’s threaded and hidden like a pill when the light is in flood mode, but exposed in throw mode.

11 mm driver is a simple FET-based e-switch. SST-20 or Boost HL are not good with 18650 and FET but should handle 14500 very well, being driven fairly hard but not too hard.
The driver uses Cell-side switching, implemented simpler than tterev3 did - the switch is simply soldered on the driver board.

Going back to the head…there are small yellow squares to the sides of the main LED. These are RGB aux LEDs. The plastic tension ring that holds the MCPCB against the pill is not a typical white or black piece but just a heavily frosted diffuser.

The tail…is not very remarkable, just a cell-pusher cylinder. This tail doesn’t have a space for a lanyard hole. A surefire-style ring would be used instead.

Size? My drawings are not precise enough to tell for sure. It’s not going to be the smallest 14500 light out there but it would clearly be on the smaller side of things.
At the same time it would be sized similarly to typical AAA zoomies, much smaller than any AA zoomie on the market.

I spoke about potential complexity removal…
One could ditch the aux LEDs to make it a little simpler.
Or even ditch the e-switch and put a regular clicky at the tail. Then this zoomie wouln’t come with any novelty of its own, just yet-another 14500 light. But still quite a bit smaller than anything else in its class.

After drawing this I went back to my 2019 dream thread.
I am surprised that the dream has changed so little. :slight_smile:

I love it! For me 14500 is the best edc size. 18650 is too big, AAA depleted too soon. Tried bot, stuck with 14500 past 2 years and never looked back.

I modded all my edc’s to bistro with clicky. I love anduril, but prefer clickies for EDC to prevent accidental activation. Lockout with e-switch is too much hassle IMHO.

Zoom is really great but you are right, there are no good hosts let alone good EDC zoomies out there.
It should be waterproof. I machine washed my tool AA and AAA’s a couple of times already by accident and all were nice and clean and worked without issues thereafter :+1: :smiley:

And keep the aux lights, these are great to find the light in the dark. Illuminated tail clicky or aux lights in the front, either one will do.

I would love to see someone build this BLF ZoomEDC, surely interested!

Zoomies aren’t really my thing but I love the the detail and thought gone into this. Nice :+1:

I’m yet to wash a flashlight…though yeah, EDC lights should be the most foolproof kind as the user will have the most occasions to break them.

Waterproofing aspheric zoomies is fairly easy and may be worth it for a real light, though I didn’t include it because I thought that the tradeoffs were worth it for this concept - which was meant to be my dream. :wink:
There are multiple ways to do it but the simplest that I know would be a flat glass window fixed at the front of the pill head. The idea would be to waterproof the parts that contain electronics but don’t try to prevent water ingress below the main lens.

So what are the tradeoffs that I mentioned?

  • some extra complexity and cost
  • about 1.5 mm longer (with 1 mm thick window), a little heavier light
  • slightly lower beam intensity and output due to reflections off the window (say 3%)
  • a bit narrower flood beam
  • possibly another ring in the beam due to reflections off the window edge

For the last couple of years I’ve been thinking about ordering a custom light one day.
I used to think that I didn’t know my needs well enough to make this a good investment.

After realizing that my dream hasn’t changed much over the last year I started thinking that maybe I’m ready to do it?
I think no, I’m not yet.
But in the period of uncertainty I drew another flashlight, more suitable for a custom.

If I ordered the simplest variant of the above (with a clicky) - the costs would be high, it’s a custom after all. But not extreme. Still - with clicky UI the flashlight wouldn’t be worth its price for me.

Going a step up requires a custom driver…which adds a lot to the cost. When the light is already very expensive to begin with, pimping it up a bit more is not such a huge cost increase. So I just did it.

Here you can see both lights side by side.

The first clear difference is a lens, this time from Edmunds. This one is a big upgrade:

  • stronger - for better efficiency in throw mode
  • thinner (thanks to a high-index glass) - smaller size and weight
  • top notch AR coating
  • much higher precision. This is not a cheap lens but for the precision it’s very cheap.

I wanted to protect the lens by suspending it on o-ring from each side. I failed to do it, there’s one o-ring only. But I added a glued-in o-ring to the tube. It acts as a bumper, amortizing the whole bezel-with-lens from tail-first drops. Though the light will be front-heavy, I don’t think there would be that many tail-first drops.

Then there’s UCL2 0.95mm thick lens for waterproofing.
The idea is to allow water between the lenses but not deeper. Allow - but not too easily to avoid dust accumulation. That’s why you see an o-ring at the bezel (though it also adds friction, so the light doesn’t change focus unless I tell it so).

There is no space for a plastic diffuser. DC-fix ring around the UCL lens does its job instead. Also, the MCPCB would have to be held with screws.

I changed the aux LEDs for flatter ones to get some headroom. This is a cost increase but purely marginal. BTW, they are mounted on OSHPark flex PCBs and they used to be with the previous light as well. This is a necessary thickness saver.

LED? No space for a domed one. Luxeon Z ES on MosX MCPCB should do a good job. Hi CRI, warm, throws and has reasonable output. (Though if I ordered a custom I would surely evaluate a few other options anyway).

The pill was changed to Alu 7068. It is excellent thermally, 190 K/W is almost as good as any other alloy and heat storage is outstanding as well - it stores only 10% less heat than a copper pill of the same dimensions yet it weighs 1/3 of copper.
It also has superb mechanical properities, some call it aluminium with the strength of titanium. It’s a stretch but it’s not far.
And I see that it can be purchased in reasonable quantities for reasonable cost, so it’s not unobtanium. It would be used throughout the light.

There is a small thermal pad above the MCU to improve thermal regulation.

Tail? There’s a metal switch. The metal is Nickel-coated NdFeB. In other words - a magnet. The keyring attachment fits over the switch and can be used as an actuator.

I’m not making it (and it would still require a few adjustments here and there) but I like dreaming this dream. :slight_smile:

I too like this dream

There would be a little more to it than that in order to attempt the level of sealing you’re aiming for. Vacuum induced by the sliding tube is an issue.

Might be interesting to try, would be difficult to find the proper o-rings perhaps.

You mean that there should be some way for the air to enter the head? Yeah, the idea was that the o-ring on the bezel should not be too tight (actually it’s on the wrong side and it wouldn’t stick in place if it was stretched at all. I should move it to the pill anyway.) and should let some air under the shearing forces. If that’s not enough, I guess the bezel needs some perforation to make it easier for the o-ring to deform.

That’s not an issue, I checked that already. :slight_smile:
Easy and cheap any diameter (with 0.5 mm step) with cross-sections down to 0.4 mm. The light uses mostly 0.5 mm, though I opted for 0.4 under the UCL2 as this place was very tight. And the bumper is a little larger, 0.8.
I actually drew each o-ring as .1 mm larger than it should be because I was not sure if I should leave some space for machining tolerances.

Your cross-section drawing looks very, very similar to sipik sk-68. They have bad tail (no retaining ring and tiny switch), but other parts are good enough to use as a start point prototype.

I started the year with the thought that before its end I wanted to seriously improve my main EDC. My budget is fair enough, I have a nice selection of other lights yet the light I use by far the most was a $2 zoomie….because I failed to find anything better.
I wanted to find a good host and mod it to something much better than stock. And also to learn as much as possible.

I sidetracked a bit with the thought of a custom but I’m back to seeking a mod. And I’m at the host selection stage, so your SK68 suggestion has arrived at the perfect time. Thank you. :slight_smile:

So far I stayed away from SK68 because it’s large for 14500 - but I may get one later as well. It depends on how I find some smaller hosts.
Though I had an idea of carrying a SK68 as a secondary light because I heard a lot of good about it…and also because I wanted to study its head construction. I know that it’s praised for good thermals and I don’t really know how it’s constructed. I’ll think about getting one. :slight_smile:

I actually ordered different 2 smaller AA hosts already…Skywolfeye E526 and Skywolfeye B18….and it seems that both got lost on the way. :person_facepalming:
I have a SK58 clone in the shopping cart as well, may order it eventually.

But actually a 10440 light would be much closer to my drawing in a few ways:

  • similar lens size, so any optical comparisons are going to be closer
  • similar size/weight
  • similar amount of internal space. This drawing has 10.4 mm MCPCB and 11 mm driver

At the same time 10440 cells are so much worse than 14500…they won’t enable me to get a reasonable output estimation.

So…anyway, I seek AAA sized hosts as well. Just like other lights that can show me tricks that I haven’t seen before (I just got a SMS that my new Nebo Poplite waits for me picking it up).

BTW here’s a known BUG / TODO / TOTHINK list for the drawing if I’m ever to continue working on it:

  • rework the magnet holder to prevent shearing it off
  • move the slider o-ring to the pillhead
  • seek suitable springs (These are OK but the tube would probably need lengthening)
  • find some good membrane for tail insulation
  • make the tail o-ring larger by 1 size, there’s enough space
  • what is the tail spring mounted to? it’s not soldered to alu…
  • if the front battery threads are anodized there won’t be electrical connection
  • verify there’s actually enough space at the MCPCB area for:
    • main LED up to 4040 size
    • At least one AUX LED, preferably more
    • 2 screws
      • beware of shorts
    • negative contact
    • 4 positive contacts
    • (note that screws can be placed below the AUX PCB)

Cut off extra fins from sk68 clone looks much easier than machining something else.
In reality there are always many physical limitations that dont allow you make it as slim as you think when make first sketch.
Lets say we have 15mm cell bore, 1.5mm tube and 18mm OD. Bezel threads can be 16mm, pill outer dia - 14mm, pill slide surface - 13mm. This allows to make place for 11mm led pcb, but not enough place for 11mm driver pcb with retaining ring threads.
You can try very fine threads everywhere, thin walls, no water sealing, and save few mms. Not sure this is what you want.

I'm interested, but what is UCL2? google doesn't turn up much, either here or the web. Your posts account for most, maybe all, mentions.

No mention of it on https://www.flashlightlens.com/, the guy that owns the "UCL" trademark. His site is a bit scary - same old look, shrinking inventory. He can be difficult to communicate with at times. I'm not aware of any alternatives .

ADDED: There’s UCL v1 with lower transmission and v2 which is better. Some sizes are v1, some v2. This is v2 and I can tell so by seeing 99% transmission.

My biggest gripe with UCL is that it’s impossible to get in EU.

Interesting idea. I haven’t thought about that.

I’m not sure either.
I’m actually not a fan of thick walls, but they ought to be reasonable.
Waterproofing is not a must but very good to have.

At the same time a few mm matter a lot - I’ve carried lights from ~12 mm to 25.4 mm in diameter on my neck so far. I know that 17 mm is not great but fine. I know that 20.5 mm is definitely too much, even with a very short light. Now I have a light with 19 mm head but thin body (haven’t measured, 15 mm?) - it’s actually fair enough. (EDIT: I just measured, the light is 18.5 mm thick at the head and 14.5 at the tube).
So I find it very important to stay as compact as possible.

I’m not a machinist and I know very little on stuff like what is a reasonable part thickness, thread pitch, thread length, tolerance. That’s why I can read the flashlight length and diameter from the drawing but so far I resisted spelling them out.
I know that if I were to build this light, the final design would need a lot of consultation with the machinist - and that it could require increasing size or dropping features. The feature that’s the most at-risk is probably flashing pads at the driver. 11 mm is already quite tight with the spring around the switch and yes, I kept the area around the switch as tight as possible not to reduce driver diameter any further simply because this feature is very important for me.

The indicator LEDs are at risk as well.
Though if I were to drop both of them I’d rather make the light larger. Though if I had to do it I’d seek if I could find a better lens (probably not, 15 mm is a standard size and there aren’t many options until 20 mm) - if I found one, there would be major changes all over the place.

Is this an open source type project? I’m not SUPER efficient at Fusion, but given days and weeks I can accomplish tasks of moderate pride. But if you’re wanting to keep your design to yourself, I 100% totally get it. I hate telling people ‘no’….and the only thing I hate more is being told ‘no’ because that means I did or requested something that tried to put someone outside their comfort zone. It’s a personality disorder. Seriously. But I love zoomies, and would love to take a deep dive into your design. I’m disabled, stay at home 90% of every day, and outside of online college classes, got into this hobby, and this site in particular, to do things like this. Either way, you put this out as a production, consider me the 1st guaranteed buy. I see what you’re doing and agree with what you say has been missing. I’m indifferent to the details, I want one and put me down as such, please sir! Thank you either way on my other request and have a good one.
(P.s. Seriously, great job, man!!)

It’s neither open source nor closed source really - because there is no source. Everything is either in this thread or in my head.
There is no CAD backing these drawings, I drew them pixel by pixel in GIMP.

What would be possible ways if this was to go forward?

  • maybe this could be turned to a BLF project. I don’t think that the latter drawing is a suitable starting point (too expensive) but the former - maybe. That would require quite a lot of interest.
  • maybe this could be turned to a short series of customs, notably cheaper than 1-off but still I would be surprised if it ended up below $300 a piece (ADDED: after some rough calculations - no, it would likely be below $300. But still expensive…)
  • somebody might make a 1-off based on this idea. I’m not ready to invest so much without being sure it’s going to meet my needs for the years to come. And so I’m going to invest in filling knowledge / experience holes with hope that I will actually make a new generation eventually.