Pip's 9th Annual Old Lumens Challenge - Modified Light Category

Giveaways and Contests
1

So I decided to enter this year, never done it before but I had an idea that I’ve wanted to try for a while.

My plan is to modify a flashlight to shine through a fiber optic for inspecting hard to reach places.

Like this:

Only lets be honest here, no offence to the person who designed that, but that’s a terrible design. It probably only gets about 2 percent of the light to the fiber. What you really need is a high intensity flat top LED and the fiber needs to be directly over the die, not at the end of the reflector.

So I ordered a couple convoy S2+ flashlights with 1mm osrams in them, and a roll of pmma fiber optic. With any luck, they should arrive sometime before the contest ends.

This will probably get off to a slow start while I wait for the parts.

2

Good luck.
Sounds like very useful modification

3

My roll of fiber arrived, that was quick. still waiting on the convoy lights but I decided to throw together a prototype from a singfire SF-348 that I’ve had sitting around for a while with no good use for it.

So I threw together a quick sketch in freeCAD:

Revolved it around the vertical axis:

Exported the file to an STL and sliced it for the printer:

It printed in under 10 minutes and then I had to chase the holes with drill bits to clean them out and get the size to fit the fiber perfectly.

4

Here is the assembly:

And a beamshot (well sort of):

This is all just a prototype to keep me busy while I wait on the convoy order. I mangled the silicone dome on the LED when I assembled it, but that shouldn’t be an issue with the flat LED.

5

Great idea! I still have one of those wands…actually found it quite useful a number of times even with the anemic output of a mini-mag. Got me thinking about mods to utilize it again with better lights.

6

My convoy lights are in, so here goes. Spent the time getting everything together and taking pictures so I should be able to get it all into one post.

The first step was to tear down one of the Convoy S2+ heads and get the measurements from the reflector. I didn’t realize that the S2+ didn’t have a bezel that could be removed. I had planned to do everything from the front without removing the pill, but since I had never had an S2+ apart I didn’t know how they were built. It was easy enough to take apart with a pair of snap ring pliers though.

I measured up everything and drew up a quick sketch in FreeCAD. I wanted to make the holder for the fiber out of PETG because it can take more heat than PLA, and I wanted to make a strain relief from TPU so it would be flexible to keep the fiber from getting a sharp bend where it exits the light.

The easiest way to design something shaped like this is to draw a profile of one side, then revolve it around the axis to make the 3d model.

Here is the sketch of the fiber adapter:

Everything is sized exact in the drawing, generally the holes will run a little tight off the printer but I planned to run a drill through them anyway to get it perfect.

Here is the 3d model after I revolved the sketch around the axis:

I exported the model as an STL and loaded it into the slicer software:

And then printed it:

The next step was to design the strain relief. This follows the same steps as the adapter, only difference being that I printed it in TPU instead of PETG. TPU is a flexible material, like a hard rubber.

Sketch:

Model:

Slicer:

Print:

Everything turned out pretty much how I wanted it to, the holes were a little tight so I chased them with the number drill bits that were the right size for the holes, but other than that it was ready to assemble.

The LED centering ring fit perfectly into the adapter, which was important because alignment is pretty critical when you’re trying to drop a 2mm fiber in front of an LED. I stripped back the black pvc cover over the fiber to keep it away from the LED, otherwise it would probably have gotten too hot and melted.

Here is the assembled fiber, adapter, strain relief, and centering ring:

And here is the final light in operation:

I don’t have a way to measure lumens, but from comparing it to other lights I would guess it to be about 300 out the end of the fiber. It’s a pretty even circle of light coming out the end.

Here is a beamshot comparing it to the same light unmodified:

And of course it wouldn’t really be complete without trying to burn some black paper…

Yep, it passes the one true flashlight test, it can start a fire. Of course you have to be really close to the fiber, and it would probably melt the end if you left it there long enough to actually start a fire. But I’ll call it a pass.

And that’s pretty much it!

7

If you put a convex lens in front of a regular flashlight you can also concentrate the light into the fiber end, so it gets a lot more than 2% of the light.

8

To a certain extent yes, but I see no practical way of doing it on a small light at a budget price. I actually have a lens from a laser scanner that I thought of trying to use with it but it’s hard to set it up without being able to look at the focus with the light turned on.

9

Wow, nice work! That is a gigantic improvement over the original concept/implementation. What current is on that osram in the photos? I guess it would depend on the need, but for how I’ve used these in the past, warmer color temps are much more useful than what we usually use in emitters these days.

And a question….I know almost nothing about 3D printing, so what are those large rings of material on the bed outside the print? Waste material at the start or finish of the job?

10

Thanks! The convoy s2+ has the stock driver for the osram, so probably 8 amps or a little less. I wanted the led as small as possible so I could keep the fiber small, the idea being that I can snake it through a small hole to see things, like for trying to fish wires through a wall or something. I ran some electrical wire through a wall from the basement to the ceiling and it would have been really nice to have it then, there was no way to see what I was doing and I just had to keep poking it around until it finally chanced upon the hole.

The material around the print is called a skirt in the slicer software. It’s basically a purge to get fresh material in the nozzle, it isn’t really necessary most of the time and if I print something that’s almost at the limit for the bed size I disable it. It’s on by default so unless I uncheck that box it’ll print. And I usually forget. :slight_smile:

11

Great for that. They’ve got some LED tips for fish rod these days, too, although they aren’t so useful when you have insulation packing the space. I’ve mostly used these for automotive and mechanical stuff.

Interesting about the print skirt…thanks. I think 3D printing is one of the neatest home-accessible inventions to come along in quite awhile - sure has opened doors for creativity and repairs! Anyway, great job, you made something useful and improved on a design!

12

Interesting light you are making, never seen anything like it before and it seems to be working out for you as planned (which is always a relief!)
Love you are utilising a 3d printer too!

13

This looks very useful. You really improved the design to let the maximum amount of light pass through. :+1:

Is the inner part of the optical fiber heat resistant?

14

I REALLY look forward to seeing how this comes out. I’ve thought about trying adapt this to a headlamp- I use one often when I’m working on vehicles and sometimes I just don’t have space to get the beam from the headlamp to line up with my eyes. This would be a perfect way to always have light shining where you need it.

15

The inner part is PMMA, which I think is the same as acrylic. It isn’t exactly heat resistant, but has a fairly high melting point at around 170 degrees Celsius. The service temperature is only up to around 90 degrees Celsius but with no physical load on it I imagine you could run up quite a bit higher than that without harm.

I had to make a slight change from my original, I tried running it for a while to see if it would hold up, and the fiber melted, but not at the point I expected it to. I would have expected the failure at the closest point to the LED, but that’s not what happened. It actually melted at the point where the PVC jacket started.

I’m guessing one of two things happened, either I damaged the fiber when I stripped back the jacket, and the light was coming out at the damaged point and overheated the black PVC, or the PVC was too close to the entrance point and the angle of the light was still under the TIR angle so too much light hit the black PVC and overheated it.

I reprinted the adapter to keep the PVC jacket 3mm further away from the end, and re-terminated the fiber and reassembled it and it works fine now, though I haven’t run it longer than a couple minutes at the highest setting.

You can hit the white plastic adapter with a lot of light without hurting it because it diffuses through the plastic and spreads out the heat. The black plastic can’t handle that much light because it instantly turns to heat wherever it hits.

16

I don’t think it would work too well for a headlight, the fiber has a pretty large minimum bend radius so you wouldn’t be able to get it aimed very well, though you could possibly use a mirror on the end to reflect it at a 90 degree angle. That would allow you to keep it really close to your head. The other issue would be the efficiency, it seems to only get about a third of the light out the end. The incoherent light from the led only transmits through the fiber if the angle coming in is greater than the TIR angle for PMMA. I couldn’t find the angle specified anywhere but I would guess it to be about 45 degrees, so anything coming from the led with an angle greater that doesn’t transmit through the fiber.

17

Nice project Pip! Interesting to read. I don’t know anything about 3d printing or optical fiber, but I think it’s great.