Thread: -Dive Light Project-

1. -Dive Light Project-

Hi guys. Recently found this forum and quickly found it to be a wealth of useful information and helpful people.

My friend recently shelled out over \$1k for a dive light and being a mechanical engineer major who works with optics and a diver it occurred to me I can make a flashlight twice as awesome for a quarter of the price, and have the ability to say I designed and made it when people ask me where I bought it.

I am going to document the project here and annoy you guys for occasional feedback/scrutiny; no substitute for experience.

Lets make a flashlight.

-------------------------------------

Question1(sealing):

I rather not engage in finite element analysis, even though I have a parametric model going, since I think the answers can be found much faster by probing the experience pool around here.

The question is do the concentric side sealing o-rings really do the job well and seal the light down to ~300psi (600ft)? I have seen that feature reading though the threads but I wanted to make sure I wasn't misinterpreting the design. The o-rings I am referring to are the three seals on on the maroon body in the image of the model.

And if so what the typical % height compression of the circular cross section of the thickness diameter; ie:the part of the o-ring in the image that resides inside the purple volume.

Considering everything is lubed well with silicone.

thanks!

*yes there are a few overlapping areas on the model but the geometries are changing my the minute as I play with different lenses and other components; this is just a snapshot of the creative flow process*

2. Re: -Dive Light Project-

Hi Sonodelirii, and welcome to the forum!

Yes those seals can work very well. My light has a double at the back (where batteries are changed) and a single at the front (where the lens fits in, which never moves). It's been tested on dozens of dives and to 500 ft in a pressure chamber. Never leaked.

The secret is proper constructon of the gland (the area where the o-ring sits), machining good surfaces, proper o-ring selection, and proper maintenance. Okay, that's several secrets. And they're not really secret.

There are a couple of good instructions on working with o-rings, from the manufacturers. I have a hard copy, so have never used one on the web, but I think you can find one at www.applerubber.com.

D

3. Re: -Dive Light Project-

Hi,

seems like a nice project. subscribed
I really have to get windows on my PC in a dual boot so I can also use mechanical design software to make those pretty graphics.

as for the question, I've seen some tutorials on o-rings, but since I don't have access to lathes and cnc things, I didn't bother to study the details. the triple o-ring setup works well as long as you use good quality o-rings. If I remember correctly, there is a simple formula for the depth and width of the groove based on the diameter and thickness of the o-ring.

on the side of the lens, is that also an o-ring in front of the glass and around/on the fesnel lens? The design looks a bit similar to the greenforce heads, except that they put an extra o-ring around the (in your graph) blue part and the orange part would screw just over the o-ring for additional sealing.

Is that fresnel lens in the design an existing lens of just something you are toying with? if it's an existing lens, I'm interested in the details and where to get it

already decided on the LED you will be using?

4. Re: -Dive Light Project-

A good O-ring information page...

http://www.rlhudson.com/O-Ring%20Boo...ing-tables.htm

5. Re: -Dive Light Project-

?I know the likes of greenforce use that sort of arrangement, but ive never been able to make it work due to the very exact machining tolerances needed.

All my lights now have O'rings that are heavily compressed by the lenses that have a step in, the same way a battery cannister seals.

Its a bit dated but you might like to find a book called "dive light compaion" by airspeed press, it covers many housing designs.

6. Re: -Dive Light Project-

jspeybro,

The seals in the front are seals but are not o-ring type. Laser cut to spec from a flat sheet of material; flat rings as opposed to o-rings.

The LED will most likely be a Luminus SST-90. Hopefully by the time I am done someone will make a driver for that amount of current, I understand it's "in the works" around here

It's an existing aspherical fresnel lens, we have a few at work.

Here are some excellent companies/sites for you guys that we use for purchasing optical and photonic products at work:

http://www.eksmaoptics.com/en/main/products
http://www.edmundoptics.com/products/

------------------------------------

Packhorse,

Huge thanks! That link was like Christmas when you are 10 all over again!

------------------------------------

Thanks for the useful info guys. I am sure the design will change about a thousand time by the time I am done and might not go too fast since I am drowning in school and work but here is snapshot from last night for the pixel hungry subscribers. These kind of threads are totally worthless without pictures

diameter ~2in
length ~3in

*some things like knurling are absent from the model
*overlapping parts imply threads right now; they are displayed as holes before tap
(until I figure out how to edit the hole wizard database and add new standards in solidworks' toolbox)

7. Re: -Dive Light Project-

- add water cooling under led (you won't have problems with rising power
of leds / heating )

- I have build CC driver last week finaly (still need some trimm) but it is
much bigger than your space (just main coil have 25mm x 25mm x
15mm) , whole driver have 30mm x 50mm x 20mm just for your
orientation

8. Re: -Dive Light Project-

Originally Posted by lucca brassi
- add water cooling under led (you won't have problems with rising power
of leds / heating )

- I have build CC driver last week finaly (still need some trimm) but it is
much bigger than your space (just main coil have 25mm x 25mm x
15mm) , whole driver have 30mm x 50mm x 20mm just for your
orientation
Thanks for the tips. I haven't forgotten about cooling or space for components not modeled currently. I am taking design in prioritized steps:

Optics
Enclosure designed around optics
Enclosure modified to focus beam
Enclosure modified to seal
Enclosure modified to house driver
Enclosure modified with cooling features
Enclosure modified for wiring

Thanks for the driver dimensions. Do you plan to sell it here? Do you expect it to change in dimensions or is it a final revision? Is there a thread for it?

9. Re: -Dive Light Project-

here you can get a single mode driver for the sst-90 or sst-50:
Der Wichtel Driver.
Multimode is on the way but may take a while due to lead times of components.

also taskled.com has a 6.7A driver (H6CC) if you don't need 9A or more.
I think DIWdiver also had a 10A linear regulator if you can work with that. not sure if he's still selling those.

I know about those optics companies, but didn't really find anything that I liked for a reasonable price or with the right specifications. Which one do you plan to use? if the diameter is 2" (what's wrong with metric units ), I suppose the diameter is somewhere around 45-48mm for the lens?

You'll want to place the led slightly further than the focal length, measured from the front of the lens (grooved side).

10. Re: -Dive Light Project-

Hi,

Maybe I'm wron, but it won't work. The external pressure will push the lens and the hood towards the LED.
On 30-50m you can't even pull it out (depending on the diameter of the lens)
You must use some spiral-like zoom mechanism.

11. Re: -Dive Light Project-

I don't think he's trying to make a focussable light. note that the overlapping parts are threads, which should keep it from moving under pressure.

12. Re: -Dive Light Project-

Originally Posted by jspeybro
here you can get a single mode driver for the sst-90 or sst-50:
Der Wichtel Driver.
Multimode is on the way but may take a while due to lead times of components.

also taskled.com has a 6.7A driver (H6CC) if you don't need 9A or more.
I think DIWdiver also had a 10A linear regulator if you can work with that. not sure if he's still selling those.

I know about those optics companies, but didn't really find anything that I liked for a reasonable price or with the right specifications. Which one do you plan to use? if the diameter is 2" (what's wrong with metric units ), I suppose the diameter is somewhere around 45-48mm for the lens?

You'll want to place the led slightly further than the focal length, measured from the front of the lens (grooved side).
Thanks for the driver info that's good to know.

There are plenty of reasonably priced products on those sites. In fact just about everything at edmund optics is extremely reasonably priced and this is coming form someone who works with optic and photonic equipment.

Yes I am looking at the 2" aspherical fresnel (for now), and for the quality and type of product its very well priced at only \$25. If that's too much for you then I don't know what to say; \$25 is half a tank of gas in my RSX which will be gone in half a week or a nice dinner vs a lens that is something that will last indefinitely. Unless you are making thousands of lights \$10, \$20, or \$30 is all the same, and if you are, you would get bulk discount anyway. Also considering the alternative is a \$2000 brand name light I think you win in the end game by a long shot.

Nothing wrong with metric units; in fact my model is in ANSI Metric. 2" is 50.8mm even.

I know exactly where the led should be with respect to the lens, however that's irrelevant as the light will be focusable.

Originally Posted by kubi
Hi,

Maybe I'm wron, but it won't work. The external pressure will push the lens and the hood towards the LED.
On 30-50m you can't even pull it out (depending on the diameter of the lens)
You must use some spiral-like zoom mechanism.
Originally Posted by jspeybro
I don't think he's trying to make a focussable light. note that the overlapping parts are threads, which should keep it from moving under pressure.
The light will be focusable and it should work fine.

The lens assembly (orange cap, gaskets, window, lens) remains stationary attached to the purple shell.

The purple shell however rotates around the threaded maroon body and is limited as to how far it goes by retaining steps in both directions.

The cutaways around the circumference of the purple shell are not cosmetic, they provide grip when wearing thick gloves.

The grip handle will be attached to the maroon body.

----------------

This is a work in progress and everything you see in those shots is just brainstorming. Its possible that nothing you see will actually make it to the final revision. I just posted those so curious people can see the evolution/design progression because threads without pictures are boring.

When I do finish thinking about it I will post the final design for some feedback, but that wont be for a couple of weeks

progress

14. Re: -Dive Light Project-

Id love to be able to design stuff like that but I just dont have the patience to learn the software.

Good job and nice handle.

15. Re: -Dive Light Project-

Originally Posted by Packhorse
Id love to be able to design stuff like that but I just dont have the patience to learn the software.

Good job and nice handle.
Thanks Packhorse.

I don't know what you do for a living but I am majoring in mechanical engineering and I work in a start up that revolves around optics. If I couldn't design and model a flashlight I would be a pretty huge failure as an engineer.

It does take literally forever to learn. Partially because there is so much content and partially because it grows faster then any one human being can keep up.

One of my cad professors told me he was considered a master expert in Pro-E back in 1996 when he knew 4% of the software. Since then he doubled or quadrupled his knowledge every year. In 2009 he officially knew 2% ...

Now add to that the fact that one company uses solidworks, next one pro-e, next one catia, next one autodesk, etc. Then add to that the fact that they all have finite element analysis packages, add to that the stand alone meshers and stand alone FEA packages.... oh then computational fluid dynamics FEA software packages..... makes your head spin when you realize how much you need to know but don't. Or at least makes my head spin and I consider myself pretty competent when it comes to modeling

-----------------

WITH KNURLING

COOLING VEINS

16. Re: -Dive Light Project-

Originally Posted by sonodelirii
Yes I am looking at the 2" aspherical fresnel (for now), and for the quality and type of product its very well priced at only \$25. If that's too much for you then I don't know what to say; \$25 is half a tank of gas in my RSX which will be gone in half a week or a nice dinner vs a lens that is something that will last indefinitely. Unless you are making thousands of lights \$10, \$20, or \$30 is all the same, and if you are, you would get bulk discount anyway. Also considering the alternative is a \$2000 brand name light I think you win in the end game by a long shot.
That 2" would be nice, if it were round. I think with these lenses the lens itself is round but it is molded in a square for 'easy mounting' (so you need the 2.5" to get a 2" lens). But if you have access to laser cutters, that's probably not a problem. 25\$ is indeed very reasonable for a quality lens like the ones from edmund. (lenses I usually work with costs 5 times that amount or more, I work with laser optics and machine vision).
On the other hand, this is an acrylic lens which cost them probably only a few dollar.
I wish I had access to all these engineering tools...

I was more talking about the specifications of the lenses. The focal lengths of the edmund lenses are too big for the project that I had in mind.

nice progress with the design.
be carefull not to post too much mechanical details, some big name company might copy it and earn money on your back

johan

17. Re: -Dive Light Project-

@sonodelirii

you will prepair plans / fabricated /sell this lamphead like kit ?

18. Re: -Dive Light Project-

I have seen people cutting up and gluing maglites with epoxy for divelights but this is the real stuff. It looks good.

All tips I can give to you are:
A driver might be the best underwater but if you want to try without a driver then you could use 3 NiMH D-cells in direct drive for about 2 hours runtime.
I do not know if the cooling grooves are necessary for cooling underwater but I would surely put attention to making the metal behind the LED thick enough to aid in transporting the heat to the outer surface efficiently enough.
If the lens does not work you could concider using a aspheric lens behind a flat lens (so the aspheric is surrounded by air).
If you really want to go wild the manufacturer of the SST-90 (phatlight) also makes a led that concists of 4 SST-90 leds combined into one LED. Gives about 4 times the light output of a SST-90 and cooling should be no problem underwater.
I wish you success.

If you decide to make some extra there might be interest on this flasaholic forum.

19. Re: -Dive Light Project-

Very interesting project.

I’m not totally sure if I know what I’m talking about but I think you need it shorter – emitter closer to lens. SST-90 has 100 degree viewing angle. Assuming 50mm lens you need emitter closer than ~21.3mm from lens, otherwise lot of light will hit walls instead of lens. Seems like Aspheric Fresnel Lens 2.0" x 2.0", 0.85" FL from Edmund optics would be close enough (assuming you can cut in nicely to 50mm circle).

20. Re: -Dive Light Project-

Originally Posted by jspeybro
That 2" would be nice, if it were round. I think with these lenses the lens itself is round but it is molded in a square for 'easy mounting' (so you need the 2.5" to get a 2" lens). But if you have access to laser cutters, that's probably not a problem. 25\$ is indeed very reasonable for a quality lens like the ones from edmund.

On the other hand, this is an acrylic lens which cost them probably only a few dollar.
I wish I had access to all these engineering tools...

I was more talking about the specifications of the lenses. The focal lengths of the edmund lenses are too big for the project that I had in mind.
They are square, but I know some laser cutting places that I used to get some things cut for work. I was going to get them to trim the lens if I can't locate a round one.

I am starting to work out the specifics at this point. Until now I was just getting mechanics out of the way. Drawings for the machine shop have to get done one way or another sooner or later. That or g-code if you machine your self. Either way you need a model.

Originally Posted by jspeybro
(lenses I usually work with costs 5 times that amount or more, I work with laser optics and machine vision).
right on, machine vision is cool. Its amazing what you can acomplish when you give a computer eyes and a nice algorithm. What exactly do you do?

Originally Posted by jspeybro
nice progress with the design.
be carefull not to post too much mechanical details, some big name company might copy it and earn money on your back
You know I had that thought but then I decided I am not exactly inventing the wheel here. I don't think I have or will come up with anything revolutionary on this project. But thanks, I guess I'll try to curb my trigger happy schematic posting impulses.

Originally Posted by lucca brassi
@sonodelirii

you will prepair plans / fabricated /sell this lamphead like kit ?
-Will I prepare plans?

Off coarse, how else will the shop machine parts

although, I might go down to the shop I used to take classes at and use their haas cnc machines. Would certainly save me some money on the prototype. No blueprints then, but I would have to write some g-code with mastercam.

-Will I make it?

Off coarse, I don't try to design things I have no interest in making. Plus I am a diver and I need a dive light and I refuse to spend thousands of dollars on a nice one when I can make a better one.

-Will I sell it?

If people are interested I would be more then happy to. I don't think anyone should be paying the kind of money for a light companies out there charge

Originally Posted by Roland
I have seen people cutting up and gluing maglites with epoxy for divelights but this is the real stuff. It looks good.

All tips I can give to you are:
A driver might be the best underwater but if you want to try without a driver then you could use 3 NiMH D-cells in direct drive for about 2 hours runtime.
I do not know if the cooling grooves are necessary for cooling underwater but I would surely put attention to making the metal behind the LED thick enough to aid in transporting the heat to the outer surface efficiently enough.
If the lens does not work you could concider using a aspheric lens behind a flat lens (so the aspheric is surrounded by air).
If you really want to go wild the manufacturer of the SST-90 (phatlight) also makes a led that concists of 4 SST-90 leds combined into one LED. Gives about 4 times the light output of a SST-90 and cooling should be no problem underwater.
I wish you success.

If you decide to make some extra there might be interest on this flasaholic forum.
I made the metal thicker behind the led, just didn't post a section view after that. I used to build systems that cool computers to -100C for fun so I know a thing or two about moving heat

It WILL work or I will be damned!! I set my mind on a fresnel lens. I think the only thing that could talk me out of it is something more extreme or cool. Which just may yet happen... we'll have to see...

If this works out I'll be sure to let you guys know. I like what I do at work but I get a lot more satisfaction out of designing things I can actually use. If I can also make some money doing it while making divers happy its time well spent.

Originally Posted by arek98
Very interesting project.

I’m not totally sure if I know what I’m talking about but I think you need it shorter – emitter closer to lens. SST-90 has 100 degree viewing angle. Assuming 50mm lens you need emitter closer than ~21.3mm from lens, otherwise lot of light will hit walls instead of lens. Seems like Aspheric Fresnel Lens 2.0" x 2.0", 0.85" FL from Edmund optics would be close enough (assuming you can cut in nicely to 50mm circle).
150 degrees according to their graphs, or that's where the intensity falls off to 0% anyway; about 50% at 100 degrees.

Yeah, the shell was designed as a guide not so much as a final revision; any dimension can be changed with a click of a mouse. Once I know what optics I want I'll adjust for that, then find led and adjust for that, then driver, then batteries.

21. Re: -Dive Light Project-

I have thought cooling LED in that way.....and I think it is something real different and effective

22. Re: -Dive Light Project-

Originally Posted by lucca brassi
I have thought cooling LED in that way.....and I think it is something real different and effective

Yeah that looks like it would work very well. The only thing is as long as your led is on a piece of metal that is in contact with water it will do the job. Most metals conduct heat very well, water is substantially colder (delta T is large) and the ocean has infinite heat capacity. A bad design and a good design will both work equally well. I put the cooling veins on mine because they work and are very easy to machine. But they aren't really necessary they just look cool

-----------------

picture time, no reason, just because

23. Re: -Dive Light Project-

Hi, here's another idea :-)

the mounting of the lamp to the goodman handle seems solid, but I see room for improvement on the handle itself.

Have a look at the design of the goodman handle of Big Blue:
http://www.bigblue.com.hk/tl-1x30w-led.html
Not sure if it can be seen very well in the pics, but the plate on which the lamp is mounted, can only move up and down, not twist around the bolts because there is some metal in front and behind the part that you hold in your hand. The slot is on that part and not in the part that is connected to the light head.

You can then also use the upper part of the slot to attach some bungee cord and/or a boltsnap for easy attachment to your BCD.

24. Re: -Dive Light Project-

Originally Posted by jspeybro
Hi, here's another idea :-)

the mounting of the lamp to the goodman handle seems solid, but I see room for improvement on the handle itself.

Have a look at the design of the goodman handle of Big Blue:
http://www.bigblue.com.hk/tl-1x30w-led.html
Not sure if it can be seen very well in the pics, but the plate on which the lamp is mounted, can only move up and down, not twist around the bolts because there is some metal in front and behind the part that you hold in your hand. The slot is on that part and not in the part that is connected to the light head.

You can then also use the upper part of the slot to attach some bungee cord and/or a boltsnap for easy attachment to your BCD.
Well there is a groove that prevents the handle from rotating around the screws, its on the bottom part of the handle if you look closely. But if you want a slot to clip things too or clip the light to the bcd through there you go. I was going to remove some material from there anyway to make it lighter. May not matter in the water much but gear is a pain to haul, grams add up fast... also holes/slots look cool for some silly reason

25. Re: -Dive Light Project-

The only thing is as long as your led is on a piece of metal that is in contact with water it will do the job.
True ... but difference is in heat transfer cross-section of profile and length which brings us to volume and accumulation of heat .. and leds loves to work at very low temperatures .

( If don't work in these waters it is little difficult to explain and understand problem) But if you have ProE you can simulate heat transfer .
We use at our company I think Moldflow for tool construction of moulding plastics which is very similar at this problem (heatsource / sinks/ actual temperature)

26. Re: -Dive Light Project-

oh, I hadn't seen the little edge. Should work fine I guess.

right on, machine vision is cool. Its amazing what you can acomplish when you give a computer eyes and a nice algorithm. What exactly do you do?
I do research for sorting machinery that uses laser, LED and camera technology for example to get the mice out of your frozen vegetable mixes

27. Re: -Dive Light Project-

Very cool design

Will

28. Re: -Dive Light Project-

Originally Posted by lucca brassi
True ... but difference is in heat transfer cross-section of profile and length which brings us to volume and accumulation of heat .. and leds loves to work at very low temperatures .

( If don't work in these waters it is little difficult to explain and understand problem) But if you have ProE you can simulate heat transfer .
We use at our company I think Moldflow for tool construction of moulding plastics which is very similar at this problem (heatsource / sinks/ actual temperature)
I have pro-e but I like ansys better for cfd analysis. I can run a sim later but I am pretty sure the temperature difference will be minimal and the led performance won't be effected a relevant amount. But I do understand what you mean.

Originally Posted by jspeybro
oh, I hadn't seen the little edge. Should work fine I guess.

I do research for sorting machinery that uses laser, LED and camera technology for example to get the mice out of your frozen vegetable mixes
good, i hate mice in my stir fry

------------------------------------------------------

Question2(wire),

How do you guys (divers) feel about spiral self retracting wire (like phone cords used to be) between the head and battery canister vs straight wire?

I noticed some people use those in conjunction with a buckle clip to secure things to bcd like lights and cameras. Buckle cliped when item not in use keeping item close, buckle uncliped when item in use proving a longer leash that always has a little tension on it so it doesn't tangle. I don't want to use the buckle, just using it as an example of a wire like that being use by divers already.

Lanyard wire like this, item 9:

29. Re: -Dive Light Project-

Design is awesome, makes me want to break out my Alphacam cad/cam software again!

IMHO spiral cable is a definite no no!

A light like this would definitely be of interest to many who'd dive in confined spaces like wrecks and caves etc. Cave line, old cables in wrecks etc would slide off a single smooth cable but you having a chance of catching those items in the curly telephone cord like stuff is just one of many reasons why not do change the standard cable.

Really look forward to seeing this progress

30. Re: -Dive Light Project-

+1

IMHO spiral cable is a definite NO NO NO !
Even telephone cord somtime made such mess ... think about fishing thread

Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•