Custom DIY parabolic metal reflector?

Nubo · #1 01-12-2005, 05:12 PM

I’ve been lurking for a few weeks and have some ideas about a light I’d like to build. An incandescent that probably would be best with a metal reflector. Doesn’t look easy to find. Especially if one has specicific dimensions in mind, it seems like the availability of reflectors becomes a design constraint, forcing people to design their lights to parts with off-the-shelf dimensions of commercial lights.

This being my “mark I” light, I’m looking for ways to use hardware-store materials, which don’t necessarily match the dimensions of a Maglite. I’m not up for the expense of milling equipment at this point, and certainly nothing on the order needed to machine a parabolic reflector to optical dimensions. But, I have an idea and would like to submit it to see if anyone’s tried it or would like to. It may be a few weeks before I have a go myself.

If one takes a cylindrical container of liquid and spins it, the liquid will take the form of a parabola. The faster it spins, the steeper the curve. This phenomenon has bee used to construct rather large telescope “mirrors” out of spinning liquid mercury! (http://www.abc.net.au/science/k2/moments/s64751.htm)

For a flashlight reflector we don’t need anything approaching 20 billionths of a meter accuracy. So, I submit this as a possible DIY method:

You would need some type of rotating wheel with variable control of rpm. Something like a potters’ wheel or turntable.

Partially fill a cylinder of the desired diameter with a slow-setting epoxy or some casting material that doesn’t shrink much when it sets.

Place the cylinder on the “turntable”, with it’s center aligned with the table’s center of rotation.

Spin the turntable at a speed which yields the depth of parabola that you want. Maintain that speed until the epoxy has set. The more constant the speed the better. Weight added to the rotating mass would help cancel out any variations.

You now have a parabolic mold, hopefully with a nice smooth surface.

Make a plaster cast of this mold. It will need to have “sides” so that you can cast your reflector.

With a suitable metal (Silver solder or silver brazing rod?), melt and pour into the mold and let set.

Voila – a metal parabola, hopefully with decent reflectivity when polished. Perhaps needing a clear coating to prevent oxidation?

Well, a lot of assumptions here but worth a try, I think. Would seem nice to be able to design the dimensions of a light around considerations other than what size reflectors you can find.

Regards.

pbarrette · #2 01-12-2005, 06:08 PM

Hi Nubo,

And welcome to CPF! You will quickly discover that it'll eat up more of your time and money than you ever expected. [img]/ubbthreads/images/graemlins/smile.gif[/img]

I think the main difficulties with the home-brew spinning method is going to be properly centering your spinning mold and reducing vibrations.

Without an accurate centering method, the parabola itself won't be centered and thus won't produce a quality beam.

Without reducing vibrations, the liquid parabola will not have smooth sides and will likely solidify without smooth sides.

It's a really interesting idea though. Instead of slow-set epoxy you could use a clear, UV curing epoxy. When the parabola looks good, blast it with a high-power UV curing light (UV dental luxeon comes to mind) and you have a solid almost instantly.

pb

gadget_lover · #3 01-12-2005, 10:42 PM

Accurate centering can be done with a test indicator mounted on a stand.

One thought that comes to mind is to use a cylinder in the chuck of a mini mill or drill press with suitibly small runout. That might work.

Daniel

Nubo · #4 01-13-2005, 06:00 PM

FYI, this pdf contains the formulas that can be used to determine what speed of rotation is required for a desired focal length. For example, a focal length of 1/2 inch would require about 188 RPM.

http://homepage.mac.com/astronomytea...tingmirror.pdf

And here's another page describing a large-scale facility for creating large telescope mirrors by spinning the molten glass until it cools. This represents phenomenal savings, as "figuring" such a large mirror from a flat slab requires a tremendous amount of work. If I recall correctly, the 200-inch mirror at Mt. Palomar took over a decade to go from a "blank" to its final shape.

http://www.science-spirit.org/articl...article_ID=361

carnal · #5 01-13-2005, 06:36 PM

There might be some absolutely outstanding reflectors around, being thrown out by the thousands. I worked for a city gov traffic signal dept. I've installed hundreds of led displays, and thrown out all the socket/reflector assemblies. Some refl's were aluminum/tin but many were flawless 8 and 12 inch glass. Wish I'd saved them all.

Dont know what the focal point was but they had a hole in them for the socket, and you'd screw a bulb through the reflector into the socket. So what ever the distance the element is from the base is where the hot spot throws from.

I've thought about a 3D mag light with a 12 inch reflector.
You'd have to put the led on an aluminum pedistal a few inches high to go through the back of the refl to reach that hot spot. Then there is the whole enclosure consideration. I'm sure it would be bright. I'll have to talk to my traffic signal friends to see if glass refl are still around. Lots of this old technology is in landfills now.

But hows that for an idea? 8 or 12 in premaid glass reflectors!

Trader55 · #6 01-14-2005, 04:29 AM

Have you tried Edmund Scientific?..for the parabolic mirror.

Trader55

jpeg · #7 01-14-2005, 10:35 AM

carnal: Ditto with train crossing lights. 13-15 in. glass. I put a 300w halogen bulb in one and accidentally bubbled the paint on my ceiling. I sure would love to get my hands on another one. [img]/ubbthreads/images/graemlins/smile.gif[/img]

LightSward · #8 03-19-2009, 09:39 PM

This issue is an old resolved one by now for sure, but just in case:...I did successfully design and build, rather easily, a 24" parabolic reflector for about $35.00 plus time...and I used a slowly spinning 'screed' shaped like a deep parabola to apply the final surface of Dry Wall Joint Compound on fiberglass and Styrofoam framed reflector. I used this design to harness as much light from a standard HID Home Depot Metal Halide Lamp 100 watts, as I can. I get a very well focused brilliant beam that goes out several miles and the beam can be seen for about one and a half to two miles especially when the light beam is moved across the sky.

I took more beam shots of the 35 watt HID modified Cyclops, simply because it is much easier to pose the beam for photos. The 24" homemade reflector LightSward is still temporarily sitting on a box.

I am very satisfied with the results of both lights. I got many of the ideas from "CandlePowerForum" Once I finish the 24" assembly and can manuver the light more easily, I will get better beam shots. I'm excited to see what happens when Iinstall a 400 watt HID.

LightSward

The totally from scratch, homemade 24" parabolic reflector on temporary stand, next to the modified, HID added, Cyclops

Searchlights at 45 degree angle from horizon and aimed overhead towards camera. The New brighter 24" homemade reflector, with 100 watt metal halide short pulse arc tube, is on our (viewers) left and modified 35 watt HID Cyclops on our right.

Tree is 200 feet away

The Hot Spot from the modified 35 watt HID Cyclops is actually smaller. Camera overexposes. I need a better camera

The Modified 35 watt HID Cyclops beam competes with street light

Beam Shot of Modified 35 watt HID Cyclops on the left versus the much brighter New Homemade 24" LightSward beam

Cyclops HID warming up to full brightness

Beam from 35 watt HID modified Cyclops paints the night sky

Benson · #9 03-19-2009, 11:41 PM

You seem to be posting links to images in an email. Since we're not logged in to your Yahoo mail account, we can't see them...

LightSward · #10 03-20-2009, 11:34 PM

If someone can walk me through step by step how to enclose photos. I am usually quite computer literate, but sometimes...

dom · #11 03-21-2009, 05:15 AM

Hi LS
You have to host your pix at PhotoBucket or similar place -then link them to your thread with the "insert image" icon.

Search Google for "image hosting sites".
I use flickr myself.

Cheers
Dom

LightSward · #12 03-22-2009, 01:14 AM

Thanks
LightSward

LightSward · #13 03-22-2009, 02:35 AM

I have gone through and fixed the photos and found I had put duplicates up. I will get better beam shots as budget allows a better night camera.

Beam Shot of Modified 35 watt HID Cyclops on the left versus the much brighter New Homemade 24" LightSward beam

Cyclops HID warming up to full brightness

Beam from 35 watt HID modified Cyclops paints the night sky

Beam is easily seen for a mile or more in the night sky. When lights are moved or rotated back and forth, the beams can easily be seen almost two miles. Once I get a better camera, I can photograph the beams from a mile or so. I plan to upgrade the 24" LightSward from a 100 watt metal Halide to a 400 watt metal halide (pulse ignition with 1/2" arc discharge tube). The partial purpose of the large 24" reflector was to allow experimentation of using ordinary household light bulbs etc., for fun!

gadget_lover · #14 03-22-2009, 10:05 AM

That was nicely done. I was disapointed with my cyclops the first time I compared it with a real search light. It looks like yours compare well.

It was not clear from your narrative which beam was which, especially in the 4th and 8th pictures.

Ironically, large reflectors are easier to fashion than small ones like pocket lights. A small mistake in a small area is proportionately bigger than a big mistake in a huge reflector.

Daniel

LightSward · #15 03-22-2009, 10:14 PM

Thanks, There seems to be some 'threshold' that is achieved on at some point, when the beam becomes bright enough to be seen easily. The original Cyclops is on the verge. The 35 watt HID modification seemed to have done the trick...perhaps the bluer light color and the much more focused beam crosses the line into an easily seen beam. The 24" reflector allows me to make some mistakes that a smaller one would be less forgiving of as mentioned.

I clarified the captions on the photos.

For me, the best way to find the focal point, before I attempt to either build a "Delta Focusing Plate", or just using spacers and washers, is, I first carefully hold the operating bulb base with plyers in one hand and steady the reflector with the other hand, and gently move the the bulb in and out and back and forth a little, to find the sweet spot on a wall a hundred feet or so away. Then I memorize the exact spot of the bulb base relative to the bottom mounting area of the reflector. I have someone photograph this if needed. I can usually measure and confirm the exact focus spot using a laser as well. From this I can then design very good bulb positioning hardware, sometimes with the aid of AutoCAD.

I need a better night camera.

I took more beam shots of the 35 watt HID modified Cyclops, simply because it is much easier to pose the beam for photos. The 24" homemade reflector LightSward is still temporarily sitting on a box.

I am very satisfied with the results of both lights. I got many of the ideas from "CandlePowerForum" Once I finish the 24" assembly and can position the light more easily, I will get better beam shots. I'm excited to see what happens when Iinstall a 400 watt HID.

The totally from scratch, homemade 24" parabolic reflector.

24 inch reflector is chrome tape finished and ready for first test. 2bc

First test run of 24 inch parabolic reflector. 2bc

The lower spot is the modified 35 watt HID Cyclops, (Cyclops was originally 100 watt incandescent), compared with the upper Homemade 24 inch reflector (LightSward) light beam spot. Both lights are about 30 feet away from wall. The picture frame around carpet is 36 inches tall!

Searchlights at 45 degree angle from horizon and aimed overhead towards camera. The New brighter 24" homemade reflector, with 100 watt metal halide short pulse arc tube, is on our (viewers) left and modified 35 watt HID Cyclops on our right.

Beam Shot of Modified 35 watt HID Cyclops on the left versus the much brighter New Homemade 24" LightSward beam

Recently:
I've just purchased a 400 watt metal halide lamp and ballast kit. I'll be hooking it up and installing it into the 24" reflector, replacing the successful 100 watt HID lamp. I'll do this the next couple days! I'm excited to see the results! The issue is the arc area of the 400 watt HID is twice the size as the 100 watt HID arc. The 400 watt arc size is similar to the size of the carbon arc on the big 60 inch searchlights. I don't expect as focused a beam, but it should still be a respectable two mile beam. I'll let everyone know with in the week.

LightSward

bluepilgrim · #16 05-20-2009, 01:44 AM

Catenaries!

Two interesting sites:
http://mathforum.org/library/drmath/view/65729.html look at the graph

http://thoxbui.com/catenaryreflector/ and the video at http://www.youtube.com/watch?v=ALl4y-MtgNw (there is a part two also)

The vid shows Tho X. Bui making a caternary reflector for a solar concentrator, but it could be used in reverse as a reflector. The flat one, as shown would work with a tube source of light -- for a point source like a bulb you would have to figure out how to do that (maybe some sort of netting?), but you would at least be able to make up a template easily -- I don't think you would see much difference between a catenary and parabolic reflector in practice.

LightSward · #17 05-21-2009, 01:06 AM

Quote:

Originally Posted by bluepilgrim

Catenaries!

Two interesting sites:
http://mathforum.org/library/drmath/view/65729.html look at the graph

http://thoxbui.com/catenaryreflector/ and the video at http://www.youtube.com/watch?v=ALl4y-MtgNw (there is a part two also)

The vid shows Tho X. Bui making a caternary reflector for a solar concentrator, but it could be used in reverse as a reflector. The flat one, as shown would work with a tube source of light -- for a point source like a bulb you would have to figure out how to do that (maybe some sort of netting?), but you would at least be able to make up a template easily -- I don't think you would see much difference between a catenary and parabolic reflector in practice.

Thank You for showing us this. I will investigate the use myself. I learned at an early age how to quickly make a parabolic reflector. The catenary looks pretty easy.

zeropoint · #18 08-15-2009, 10:31 AM

I'm new to this site and would very much like to find out how to build the 24" reflector that is posted here by LightSward. Would you consider sharing the process with others?

Can E-Mail me at bill@wyrinc.com

Thanks..

Bill

LightSward · #19 08-16-2009, 12:17 AM

Sure, I can show the process. A lot of it depends on the materials you are comfortable working with. I'll e-mail you. Check out the other photos and videos on the 24 inch homemade searchlight thread, (try Google search at the top of the page...It is also listed with the BIG LIGHTS)....There, I have some information to help you get started.