3D printed centering rings

Recently I built a pair of torches for which I could find no led centering ring (an E21A S2+ and a Luxeon V Jaxman M8).
I thought about 3D printing them, and after a bit of research I found a website that offered laser prints using nylon (https://www.vectorealism.com/), with reasonably low tolerances and a reasonable price for small orders. So I modelled a bit of stuff and had it printed:

- 3 Luxeon MZ rings for 7mm reflector hole

- 3 Luxeon V 7mm

- 3 nichia E21A 7mm

- 2 Quad E21A 7mm

  • 2 21mm rings to center a 20mm optic in the S2+ head
    here is a pic of the file used:

I got them today, they look pretty good

So I tried them out, first the Luxeon V one, it fits perfectly

Pretty satisfied with that one, it’s basically perfect

Then I tried the one for the E21A, I found two problems:

- minimum print thickness for the process is 0.8mm, which is far too tall for the E21A. This is easily solved by sanding the bottom of the spacer a bit, structural integrity didn’t seem to suffer.

  • when tightening the reflector, the spacer completely destroyed the ceramic casing / phospor of the led, so now I have a royal blue chip directly soldered to an high performance pcb (this might turn out fun seeing as in Clemence’s test the led without phospor can take a lot of current). I think with how fragile the E21A is, a round spacer hole is necessary. This might not be a problem with quad E21A’s because the leverage should be less than half, but I’d still rather modify my design to have a round hole.

I expect the Luxeon MZ spacer to work just as well as the one for the luxeon V, and I’ll probably use it for a future build. The S2+ triple centering ring hasn’t been tested because my only S2+ triple has a soldered pcb so the optics can’t really move.

If anybody is interested in the files, I can upload them in various formats,but you need to find some place (i think the website I used is Italy-only) that can laser print, I think the common filament technique is too imprecise. Also make sure the material used can whitstand the temperature inside a flashlight close to the led. There was a much more accurate printing option that I avoided because of price and low heat tolerance, but I’m thinking it might be usable for single emitter low power E21A builds.

Thanks for the report 1C3. :beer:

How do they handle the heat? I thought 3d printing filament melted pretty low.

This was my thought also but I think nylon melts at quite high tempererature but it is harder to print it with regular (cheap) 3D printer.

This is not filament printing, it’s done with a different process where the shape is made from a liquid compound using a laser.
That is why it can be so accurate, I don’t think you could do small parts like these using filament, regardless of heat resistance (disclaimer: I’m not an expert on this argument).
Manufacturer specifications for this material say that it can resist heat up to 80 degrees, and the light where I used it is an LD-B4 setup with pcb sensor stepdown set at 65C, so I don’t think it should have any problems

Minimum 0.8 mm? Surprising, using fdm and a small nozzle you can print really small detailed parts.
I have one with a 0.4mm nozzle and it prints at 0.2mm layer height. Can even be lower, some are printing at 0.05 layer height. That might require a little bit expensive printers but it’s doable.

The filament (pla) melts around 200 degrees Celsius to print it, a printed part will weaken at temperatures around 100 degrees Celsius iirc. There are some tests on YouTube where they pour boiling water on top of the printed object and it gets pretty soft.

Minimum thickness of 0.8mm is just for the base of the object for structural reasons, the actual layer thickness is 0.1mm. The stuff I printed has some 0.2mm details and they look pretty accurate.
The high precision printing they offer has 0.05mm layer thickness and allows for 0.014mm details but the plastic softens at 50C which is too low for stuff that must stay close to an led, but might be ok for single E21A builds