The making of a reflector - The nCluster

While thinkering what to do next, and trying to be original, a task that is increasingly difficult given the resident talents we have here, I decided to do a reflector, but since Don already made single reflectors, thus rendering my idea non-original, I decided to make a cluster reflector.

The project is not finished yet, I still need to metalize, chrome or surface the reflectors to have a better idea of their abilities.

So far I made 2 designs, a cluster of 4 reflectors that resemble the IMS20 shape, and a deep reflector, thats intended to be a thrower.

The nice thing about this project is that the reflectors were made on a machine I built myself, I posted about it here

And since every project needs a name, I called it the nCluster (neoca-cluster)

The cluster diameter in both cases fits within 40mm and the height for the shallow is 17mm and the deeper one is 29mm.


If an image is a thousand words, this post is worth more 100000

Concept


Raytracing




I always start my prototypes in a soft material:


Then I increase the hardness a little:



And then aluminium:


This was the first proto and it had many defects, the most visible one is the lines, these are mostly my fault because I didnt had at the time a ballnose mill and I used a flat one

At this point I was happy, so I made a video (15mb, windows media)

But then I continued, this one wasnt made on my machine but on a Haas VF0 professional Vertical Machining Center (VMC) as you can see there are still visible artifacts there.


Haas VMC on the left, my machine on the right

After some analisys, reading, and more trying I figured that the problem wasnt the machine but the CAM (computer assisted modeling) to put it shortly I was making routes based on a low quality aproximation of my CAD, thus the facets and lines, instead of nice curves.

Machining the deep reflector



Family photo:



Using a dremel I left an irregular surface on two reflectors and the third one was left mirror finish.

This is what I have now, these will be chromed to test them better than bare alu, and Im still looking for a shop to do the aluminization process, which is the correct way of coating a reflector.

I will try with more shapes in the future.

I hope you like my project.


Pablo

Pablo,

That is a realy cool project!

I especially like the raytracing shots. Did you write the code yourself for the software, or is it an off the shelf app?

Can't wait for the real life beam shots!!

Cheers,

Matthew

Nice work! How long does it take the machine to churn out one of these?

And, will you be selling these in the future?

Nitroz said:

Nice work! How long does it take the machine to churn out one of these?

And, will you be selling these in the future?

Click to expand...

My machine takes around and hour, the Haas around 30min. The most time consuming task is the finishing.

Regarding sales of these... I never put that idea away when I start a project for the time being its just a project...


Pablo

http://optics.mellesgriot.com/opguide/oc_5_1.htm
http://www.lambda.cc/PAGE69.htm

http://www.kruschwitz.com/HR's.htm

When looking at the charts, the human eye mainly works from 400nm to 700nm.

Remember, Aluminum will quickly oxidize during exposure to air. Adding a little car wax to your final polishing stage can delay this.

:bow:
Outstanding! :twothumbs

I'm for the final results.

:thinking: would that square area in the middle have a negative effect on the beam pattern?

That is totally cool !!!

:goodjob:

Will

wow, :bow:

Great project.
Can you talk at all about the curve you used? Is it parabolic? Also, thats a pretty cool raytracing - how did you model the emitter - as a point, or a more accurate approximation?

Great quad-flector
Will your finished reflector go into a Mag mod? Imagine the thermal properties of an aluminum block sinking to the top of four K2's rapping out at an amp each? Better start saving some money just in case "one falls off the truck".

For the raytracing and parabola I used an optics developing software, and for the emitter I simulated it using an emmiting rectangle with a dome, not sure about the accuracy of it

One thing I can be sure, after I polished one of the reflectors I inserted a white lux3 in one of the holes and I see yellow all over, so if Im wrong Im not far from being right

These shallow reflectors used 1.3mm die height and 0.7mm for the deep one.

Do we have a resident optics guru?


Pablo

woah... that is cool...

Pablo, you amaze still amaze me...
Great efforts !!

I was just thinking, if this same configuration can be made,
to be just as small as regular reflectors, the reflector *might* line up for a 4die led, say the LuxV ?
Like I said, I was only thinking....

Cheers !!

Cool

Now ... my feeble technical mind starts thinking ... what if a reflector like that wouldn't have 4 LEDs but just one in the center ... a LuxV ... to remove the donut and focus this thing better?

bk

Kiessling said:

Cool
Now ... my feeble technical mind starts thinking ... what if a reflector like that wouldn't have 4 LEDs but just one in the center ... a LuxV ... to remove the donut and focus this thing better?
bk

Click to expand...

Just like I said in post #12 , no ? :laughing: :laughing: :laughing: :laughing:

Paraphrased :lolsign:

When I started testing the shallow cluster reflector I noticed that when the reflector shape is interrupted with the adjacent one (am I being clear:thinking the beam becomes part of the corona, so the longer the distance between the led and this opening, the more spot/throw you gain.

Doing what bombelman & bernie suggest would make the shape of the resulting reflector not 360deg & parabolic, thus, in my limited knowledge of optics, making it almost flood only.

Thats why I machined the deeper cluster, so more throw can be expected.

I cant wait either to see real beamshots, I hope a chromed surface will aproximate the results of an aluminized one.

I did some early testings, but took no photos, mostly because I was using a single luxeon and testing the individual reflectors.


Pablo

bombelman said:

Just like I said in post #12 , no ? :laughing: :laughing: :laughing: :laughing:

Paraphrased :lolsign:

Click to expand...

Oooops ... :huh:
Guess I am getting old here :sick2:

Wasn't there a rant sometimes back about people not reading the replies before they post?


PEU ... looking forward to the real show, too!

bernie

:wow: Great project. :goodjob: :twothumbs
Can't wait to see some beamshots.

OMG! Way to go Pablo!



Dibs on first run!

I hope this isn't already patented.

well... in a ideal world if I publish a design before patenting it it becomes previous art if someone tries to claim for himself the invention.

So, since I don't have time and money to file a patent, I published it here, maybe is a hit in the future and I end up patenting it

Pablo