Newbie needs advice on 3x XM-L build

[Diving Gecko]

Hey all,
I have been reading a lot lately but with that comes questions so I a, hoping some of you will chip in.

Basically, I am setting out to build a dive torch almost from scratch.
I will have access to lathes, milling machines and that part of it, I will have help in.
The problem is in the electronics departments so I would be grateful for your thoughts.

What I need it to do
I wont need it to be able to run for long at all and most of the time, I think it will only need to run at 50-75% of its full output. If it can run for 30-45 mins at 75%, it should def suffice.
I don't need a hard, narrow hotspot or a lot of throw. A nice even beam at a tight to medium angle is probably ideal.

Lenses
Starting at the front end, I would like a 3up XM-L lens/optics on it, but I have not been able to find much. The only ones I have found are LEDIL in 35mm diameter with 19/25/42 degrees of half angle. (Strangely enough the LEDIL series is called "cute"...)
Do you guys know of other options for triple XM-Ls?
Or is it just as easy as just getting three single lenses?
I know for XP-G the triple options are more and it would be good to have a 3up lens for XM-L that was a bit tighter than the minimum 19 degrees in the LEDIL Cute series.
Also, is there much of a difference (other than physical size) to 35mm vs. 50mm lenses (or even smaller)?
I do like the idea of the 3up combined lenses though, as then there would be a possibility to change the lenses easily for another type of beam.

LEDs
After the lenses, we get to the emitters.
What is the norm in color for underwater torches? High white, neutral or warm?
Is there any scientific reason for either or is it just down to personal preference?
I'll likely try to make the heat sink part of the head. Should be pretty doable when machining from scratch.
I read an estimate somewhere that a single(?) SST-90 would likely not need more than a few square inches of material in contact with water and that wings aren't really needed on UW lights. Do you guys agree?

Driver and Batteries
Well, the standard seems to be Taskled, so I think I'll go with that. But which one?
And now, a completely silly question: Say I run the 3 XM-Ls at 3A is that then still just a total of 3A that the driver has to be able to handle or is it 9A?
According to my rough calculations, I would need 13 watt hours to run 3 x XM-Ls @ 75% for 30 mins. That should be covered by two OK quality 18650 cells. But since the handle part of the torch will be lathed and thus circular I might as well go with three cells as there wont be an increase in width. (I would like to keep the torch compact and def only the length of one cell + head).
So, what driver would you recommend for 3 x XM-Ls and 3 x 18650s?
And how would you run those 18650s?
I would also like the driver to be as user programmable as possible.


I know this is a lot of questions but I would really love your input.

All best,
David

 

[alpg88]

taskled. as for wich one, if energy of 3 cell is enough for your needs\runtime, use lflex, if you'll be using 4-5 cells use b3flex. i used many different drivers, taskled, are the best, imo, der witchel driver is also very good one, but taskled is easier to heatsink, even thou i have derwitchel driver in my 3xp7 mag, without any heatsinking, and it works great for years.
p.s. you can wire 6 cells 3s2p and use lflex. of course your leds must be wired in series.

 

[Packhorse]

Your post is very hard to read. Breaking it down into paragraphs would help so much!

30-45 minute run time is not much. If you were to use 4 18540 2500ma cells you would get about 37 watt hours. That enough to run 3 XML's at full power for an hour+. Or consider using 26650's. 2 of those should run for 45 minutes at full power. But to be honest I think you will end up being very disappointed with such a low run time. Myself, I wouldnt jump in the water for a night dive with a torch that had less run time than I expected dive time.

If you only plan to run them at 50 why use 3 XML's? You would be better off using 3 XP-G's. Or if you were after a tight beam then maybe a single XML.

DX have a range of 50 mm reflectors for multi emitters. 3 4 or 6. May be worth looking at .

As long as you can get the heat from the LED to the body in an efficient manor you dont need a lot of surface area exposed to the water. If you intend to use the light out of the water its a totally different story.

 

[DIWdiver]

I'll bet Diving Gecko is running windoze 7 and his enter key doesn't work on CPF. Mine doesn't either.

I get a new line as follows: open Notepad; press enter twice; hold shift and press up arrow twice; press ctrl-c.

Now, anywhere you want a new line, press ctrl-v. It's a bit of a pain, but really only takes a few seconds once you know what you are doing.

As far as 3A vs 9A, it depends on how you wire the LEDs. If wired in series, they will want 3A at around 10V. If wired in parallel, they'll want 9A at 3.35V. Either way it's still around 30 W.


The light I'm building now uses a single XM-L and jspeybro's 25mm aspheric Fresnel lens. This can be focused any way you like it from quite tight (but not quite as good as Ahorton's aspheric, which is larger to quite wide. Three of these would fit in a reasonable sized head.

 

[Diving Gecko]

Your post is very hard to read. Breaking it down into paragraphs would help so much!

So sorry for that, it did look horrible. It looked fine when I first wrote it and sent off. But since it was my first post here, it had to go through authorization - maybe it got messed up in those extra steps? I am on a Mac and Firefox and just did new line breaks which seem to stay.

30-45 minute run time is not much. If you were to use 4 18540 2500ma cells you would get about 37 watt hours. That enough to run 3 XML's at full power for an hour+. Or consider using 26650's. 2 of those should run for 45 minutes at full power. But to be honest I think you will end up being very disappointed with such a low run time. Myself, I wouldnt jump in the water for a night dive with a torch that had less run time than I expected dive time.

I see what you are saying and I totally agree with you thoughts on not having enough juice for the duration of the dive.

But... this is not really for scuba nor for night diving. Truth be told, I am hesitant in saying what it is for just because a lot of people have, what I believe, a "wrong" idea about it...
I am talking about spearing which when done freediving and within proper constraints is without a doubt a much better way of fishing for your own needs than buying commercially caught fish.
I spear maybe a handful of times a year and never take more than what I and a few friends would be able to eat that night.

So, the intended usage is for a high power light that can be used to look under ledges or in small holes during day time fishing.
As such, it would only need to stay on about 1 min. at a time for as many dives one would do that day. And having it small would be great. There is nothing in the market like this (nor would it sell due to the limitations you pointed out) and I would love to make myself one.

But I also scuba and I love nightdiving so I'll look into the 26650s as I could quite easily make two different lengths handles. One for one cell for a compact freediving torch and one for two cells for scuba and night diving.

If you only plan to run them at 50 why use 3 XML's? You would be better off using 3 XP-G's. Or if you were after a tight beam then maybe a single XML.

DX have a range of 50 mm reflectors for multi emitters. 3 4 or 6. May be worth looking at .

I was considering XP-Gs also but only because I thought there were more 3up lenses for them. I'll check DX for 3up XM-L lenses now. Some people say that lenses made for XP-Gs wont work nicely with XM-Ls?

As long as you can get the heat from the LED to the body in an efficient manor you dont need a lot of surface area exposed to the water. If you intend to use the light out of the water its a totally different story.

Yup, I am with you on that one. Perhaps machining some wings and using it with thermo control of the Taskled driver could give me a few mins on low power out of water. That would be pretty smart on night dives. Oh, I guess if I was making a night dive torch, a red emitter would be good to put in for on land usage such as fumbling around with gear without losing night vision. But I am getting ahead of myself here...

Thanks for all your comments!

 

[Diving Gecko]

taskled. as for wich one, if energy of 3 cell is enough for your needs\runtime, use lflex, if you'll be using 4-5 cells use b3flex. i used many different drivers, taskled, are the best, imo, der witchel driver is also very good one, but taskled is easier to heatsink, even thou i have derwitchel driver in my 3xp7 mag, without any heatsinking, and it works great for years.
p.s. you can wire 6 cells 3s2p and use lflex. of course your leds must be wired in series.

Hi alpg88,
Thanks for your thoughts. Taskled it is. I might need the thermo control, so will check if the iflex has that.

 

[Diving Gecko]

I'll bet Diving Gecko is running windoze 7 and his enter key doesn't work on CPF. Mine doesn't either.

I have been lurking on CPF for a while - def long enough to know that you (and packhorse) are the guys in the know. Heck, so many of you are, but I can see that you guys are always there to help and that's much appreciated but... strangely enough and very surprising, you are incorrect on this one!;-)
I am on a Mac running Firefox and it looked fined when sent to authorization. And this post looks good when previewing it also. And the first one wich I edited looked fine after editing. Weird.
Let's see if the comments I am making on you three guys' answers will end up without line breaks again as these still have to go through authorization...

As far as 3A vs 9A, it depends on how you wire the LEDs. If wired in series, they will want 3A at around 10V. If wired in parallel, they'll want 9A at 3.35V. Either way it's still around 30 W.

Can I ask how you would wire them...? I guess it would be series as the Taskled drivers don't put out more than around 6A tops, rigth?

The light I'm building now uses a single XM-L and jspeybro's 25mm aspheric Fresnel lens. This can be focused any way you like it from quite tight (but not quite as good as Ahorton's aspheric, which is larger to quite wide. Three of these would fit in a reasonable sized head.

Hmm, now I do like the idea of a focusable beam. Maybe even on the go. But I also want the light to be eh, light - are jspeybro's lenses heavy? This might be for V2 of the torch

Thanks again,
D.

 

[Packhorse]

I agree 100% on spearing being the best fishing method. No by catch, no undersized fish caught by mistake. Some say its not sporting. I say no fishing is.

I dont do a lot ( if any) freediving. But those I know that do spend more time in the water than bubble blowers. I would think it would be a pain in the arse to turn your light on and off for every dive. Better to just leave it on all the time.


Since its your first build then I recommend the KISS approach. Keep It Super Simple. Perhaps consider modifying an existing light.
The ultrafire W300 is a great light ( just under powered). Swap the LED out and a new driver would make it an awesome light. Change the battery arrangement for more run time.
I have 2 W300 based can lights. The head is a good size, the reflector is fantastic and it has a nice thick lens.

 

[Diving Gecko]

I agree 100% on spearing being the best fishing method. No by catch, no undersized fish caught by mistake. Some say its not sporting. I say no fishing is.

I dont do a lot ( if any) freediving. But those I know that do spend more time in the water than bubble blowers. I would think it would be a pain in the arse to turn your light on and off for every dive. Better to just leave it on all the time.


Since its your first build then I recommend the KISS approach. Keep It Super Simple. Perhaps consider modifying an existing light.
The ultrafire W300 is a great light ( just under powered). Swap the LED out and a new driver would make it an awesome light. Change the battery arrangement for more run time.
I have 2 W300 based can lights. The head is a good size, the reflector is fantastic and it has a nice thick lens.

Yeah, keeping it simple is not my strongest side. What does make me believe in this though is that I'll have professional help with the machining and I am good at tinkering myself. So, I'll give it a go. If I have to capitulate, I will go down the modding path.
You are right about wanting as little to think about as possible when freediving/spearing. No more gear or buttons than absolutely necessary. I have been diving with a Chinese single XML light and just the fact that it is a twisty is frustrating as I actually do want to turn it off after each dive - talking freediving here.
Also, when I am out spearing I might be in the water 4-6 hours so just keeping it running is not really viable.
But from diving with the twisty I know that I just need one simple push button. Again, the programming of the Taskleds sounds pretty perfect. As in one click turns light back on to the last used level and one click turns it off.
I appreciate the concern though and if I was not such a stickler for trying to get it right the first time, I know it is good advice. And I know my approach is what kills many newbies off.
But let's see if I can pull it off - I def wanna try.

 

[Packhorse]

How about some sort of pressure activated switch. Turns on whenever you descend!

Its a "Feature" of many twist switch type lights. Turn the light on then twist it just enough to turn off. Then when you descend the pressure is enough to push the head on to the body closing the contacts and the light comes on! Of course you may be better off designing a more reliable method, but could be worth a look.

 

[Codiak]

I'll bet Diving Gecko is running windoze 7 and his enter key doesn't work on CPF. Mine doesn't either. .

Try shift-enter

 

[alpg88]

i got win 8 same story, enter doesn't work, neither tab or anything else, but notepad ctrl V trick worked

 

[DIWdiver]

Shift-enter doesn't work.

 

[Diving Gecko]

How about some sort of pressure activated switch. Turns on whenever you descend!

Its a "Feature" of many twist switch type lights. Turn the light on then twist it just enough to turn off. Then when you descend the pressure is enough to push the head on to the body closing the contacts and the light comes on! Of course you may be better off designing a more reliable method, but could be worth a look.

That's actually an amazing idea!
I would still like to be able to flick through the different light levels and turn it on at the surface.
So, without having had time to give it a lot of thought it sounds like this might have to be an extra switch or sensor. Perhaps even with some extra programming on the driver chip.

But yes, it's a great idea and I'll see what's possible. If not for the first light, then for the second. Or third... Oops, I got bitten

 

[DIWdiver]

I would definitely wire the LEDs in series, unless I had a 9A driver that would be fairly efficient. Hey! I make a driver like that! Unfortunately, it is designed for an on-off switch, not a momentary one. And to get two modes, you need an on-off-on switch. For what you want, Taskled is probably better, which means batteries and LEDs wired in series.

Building a focusable dive light is far more difficult than it appears. Or maybe I just suck. Design is pretty simple, flat lens in front, aspheric behind, mounted in the head, which screws on and off the body to focus or disassemble. Body has the LED mounted on front, switch on back, everything sealed even if you take the head completely off. Double o-ring seal between head and body, on the water side of the threads. Flat lens has o-ring seal in the head, so nothing should get in anywhere.


Worked great on land, and for the first dive. Halfway through the second dive, I couldn't move the head. Fortunately it's locked in a tight focus, which is what I like, so it's not a big deal. But I don't think I'll try focusable again soon.

My next light is a mod rather than a full custom build, again with a single XM-L and single 26650, but fixed focus (very tight), using jspeybro lens instead of Ahorton, because the former fits and the latter doesn't.

jspeybro's lens is very light, a plastic disk 25mm diameter by 2mm thick. It can't focus quite as tight as Ahorton's, and probably doesn't gather quite as much light from the LED, but given the dimensions, it's a good compromise. If you don't want a really tight beam, it probably compares even better.

 

[Diving Gecko]

Batteries
I am a bit in the woods about the batteries.

I did the simplest of mock-ups yesterday and I mean simple! Just the inner of a toilet roll over the small Chinese XML dive torch that I have. Just to give me an idea about the size and looks.
I am also stuck up on the idea of making this as light and as little negatively buoyant as possible. I made a few simple calculations and settled on a cylindical shape of about Ø50mm x 120mm.

The point is that I need some buoyancy and 50mm diameter is still comfortable to hold and there will be room for both 2,3 or 4 x 18650s.
If I manage to make the rest of the torch light enough, 2 x cells would likely make it neutral or slightly positive and 4 would be negative.

So, I guess the smartest thing would be to forget about the 3 x cell option and the wire the batt. holders so they are either 2s or 2s2p so that the driver sees 7.4V no matter how many cells are in.
I have a feeling that wouldn't be feasible with three cells. With three, I could wire two in series and one in parallel but I think it's unhealthy for the cells, right?

 

[Diving Gecko]

I would definitely wire the LEDs in series, unless I had a 9A driver that would be fairly efficient. Hey! I make a driver like that! Unfortunately, it is designed for an on-off switch, not a momentary one. And to get two modes, you need an on-off-on switch. For what you want, Taskled is probably better, which means batteries and LEDs wired in series.

Taskled it is.
Set to around 2.8 - 3A. And to go with my battery questions above. Is it OK to drive triple LEDs from 7.4V or would they like triple cells better?

As I am based in China, it might take a while for the Taskled to arrive so I will try out a few locally made drivers as well. Just to get started with the lenses I have sourced so far (single plastics). I do not, as many people still do, believe that Made In China automatically equals crap as living here proves to you beyond a doubt that you get what you pay for. Granted, there is A LOT of crap and most of it is dirt cheap, but there is good quality to be had if you can find it and want to pay.

Building a focusable dive light is far more difficult than it appears. Or maybe I just suck. Design is pretty simple, flat lens in front, aspheric behind, mounted in the head, which screws on and off the body to focus or disassemble. Body has the LED mounted on front, switch on back, everything sealed even if you take the head completely off. Double o-ring seal between head and body, on the water side of the threads. Flat lens has o-ring seal in the head, so nothing should get in anywhere.

As I have been reading your posts, I know you don't suck. So, I will refrain from trying to make my first attempt focussable.
I am still hoping to find a nice 3up lens for simplicity and actually, I now think a medium beam is better than a super narrow for my needs.
As I will be very close to what I need to light up (within 1-2 meters often), a medium beam makes a lot of sense.
I need the potential power of the triple XM-Ls as I will often be in bright daylight trying to light up harsh shadows.

 

[Diving Gecko]

Oops, double post.
Deleted now.

 

[Diving Gecko]

Batteries
I am a bit in the woods about the batteries.

I did the simplest of mock-ups yesterday and I mean simple! Just the inner of a toilet roll over the small Chinese XML dive torch that I have. Just to give me an idea about the size and looks.
I am also stuck up on the idea of making this as light and as little negatively buoyant as possible. I made a few simple calculations and settled on a cylindical shape of about Ø50mm x 120mm.

The point is that I need some buoyancy and 50mm diameter is still comfortable to hold and there will be room for both 2,3 or 4 x 18650s.
If I manage to make the rest of the torch light enough, 2 x cells would likely make it neutral or slightly positive and 4 would be negative.

So, I guess the smartest thing would be to forget about the 3 x cell option and the wire the batt. holders so they are either 2s or 2s2p so that the driver sees 7.4V no matter how many cells are in.
I have a feeling that wouldn't be feasible with three cells. With three, I could wire two in series and one in parallel but I think it's unhealthy for the cells, right?

I could still do with some help on the batts.
Basically, I want to buy as many parts ready made as possible. Eg. I am now revisiting 35mm lenses over 50mm lens as it seems the LEDIL 35mm Cutes are still the best matched for XMLs.
And for the batts, I have come across some 18650 holders in 3 and 4 cell with different series and parallel configs. So, I need to make a choice on those.

If we forget for a sec that 4 cells obviously will last longer, then what would be the best for triple XMLs in terms of effiiciency?
I read somewhere (perhaps in a post by DIWdiver) that bucks are more efficient than boost.
So, would 3 cells in series be best (I guess they would be matched to LEDs better)?
And if I go with 4 cells, would 2s2p or 4s be best?

All best,
David

 

[jspeybro]

And for the batts, I have come across some 18650 holder in 3 and 4 cell with different series and parallel configs. So, I need to make a choice on those.

hmm, interested... any info?