Cold Ellipsoidal Reflector and 15" LCD

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
and a spherical reflector

That last was an elliptical reflector.

This one shows a large spherical reflector, a smaller spherical reflector (pick either), a T15 lamp, and a pre-condensor lens. Note how the rays reflected back from the edges of the reflector have to match the diameter of the pre-condensor lens.
 

Attachments

  • size.jpg
    size.jpg
    6.1 KB · Views: 284
jaws2421 said:
Ok Guy,
I'm not sure if your still saying no to this one http://www.surplusshed.com/pages/item/l3329.html. If i did go with this one, i would need a pre-condenser with say a diameter of about 152.4mm (about 6") right? I am starting to understand FL stuff, but could use some clarifycation.


That lens is a flat mirror that is cut in the shape of an elipse, it is not an eliptical mirror which would look like the top half of an egg i.e. a 3D shape.
 
O, ty :). I just had a random thought that im sure someone has brought up in the past. What about creating a mirror box? I guess no one does it because it would jsut blast one hella lot of random light everywhere without a dictated pathway.
 

Attachments

  • mirror.jpg
    mirror.jpg
    18.7 KB · Views: 254
mirror box

If you were using a giant CRT projection lens (with a focal length around 110 mm) that expects diffuse light from an object just a centimeter or two away, then a mirror box could direct light to a diffuser panel just before the LCD. This would work like a very bright backlight. But that only works with an LCD panel no wider than the CRT lens.

If you use a smaller projection triplet (with a focal length around 300 mm) then the light that gets through the lens and on to the screen has to be focussed by the fresnels right into the center of the lens. So it has to reach the field fresnel as a mostly parallel bundle of rays. You can get that from a spherical reflector and condensor fresnel, or a parabolic reflector that's as wide as the LCD with no condensor fresnel, or an elliptical reflector that makes the light look like a point source to a condensor fresnel.

But a square-ish mirror box will send light off in directions that won't get through the lens.
 
bowl reflector

On the topic of reflectors, how do you know how big a reflector you need for whatever size light source you have? I am probably going to get this bulb since i have a M59 ANSI ballast.
http://www.plantlightinghydroponics.com/product_info.php?cPath=375_437&products_id=93 It has a CRI of 90. Anyone have input on why or why not i should buy it? So, if anyone has an idea of how to find out what type of size reflector i need (i am just going to go with a stainless steel bowl probably because they are pretty decent and alot cheaper and easier to find. I have been looking at this bowl from Ikea. http://www.ikea.com/webapp/wcs/stor...Id=10101&storeId=3&langId=-15&productId=12493

There are also bowls that are 36cm x 16cm, 28cm x 13cm or 12cm x 6cm. Should i go this way? If yes, what size is best.
 
Re: bowl reflector

juice_e said:
On the topic of reflectors, how do you know how big a reflector you need for whatever size light source you have? I am probably going to get this bulb since i have a M59 ANSI ballast.
http://www.plantlightinghydroponics.com/product_info.php?cPath=375_437&products_id=93 It has a CRI of 90. Anyone have input on why or why not i should buy it? So, if anyone has an idea of how to find out what type of size reflector i need (i am just going to go with a stainless steel bowl probably because they are pretty decent and alot cheaper and easier to find. I have been looking at this bowl from Ikea. http://www.ikea.com/webapp/wcs/stor...Id=10101&storeId=3&langId=-15&productId=12493

There are also bowls that are 36cm x 16cm, 28cm x 13cm or 12cm x 6cm. Should i go this way? If yes, what size is best.

I was looking at the this bulb before but its not as bright as the Ushio S400DD (MH that runs on a HPS S51 ballast) and has a long arc length of 37mm as opposed to the S400DD with a 27mm arc length. Its available for sale with a compatible ballast at lumenlabs.com or Atlantalightbulbs.com. The CRI and color are better with this bulb but the S400DD is brighter, shorter arc and is a smaller bulb with pretty good CRI and color.
 
37 mm arc

Most of the light output from a MH arc lamp is at each end of the arc (unless it is a DC-powered lamp, then it is at one end). You have to use the reflector & fresnels to get most of that light through the LCD and focussed back down into the projection lens. The image of the lamp arc at the projection lens is a product of the distances between lamp & condensor fresnel, and field fresnel and projection lens.

For example, if you use a 220 mm fl fresnel and a 550 mm fl fresnel, then a 37 mm lamp arc will give you a (550/220) * 37 mm = 92.5 mm arc image. You won't be bale to get light from both ends through any available projection lens with an arc image that size. I think you would get a brighter image if you position the lamp so one end of the arc image hits the center of the projection lens! Then you would at least get half the light output in there. But you would still only get half the light through the lens.
 
Re: 37 mm arc

Guy Grotke said:
Most of the light output from a MH arc lamp is at each end of the arc (unless it is a DC-powered lamp, then it is at one end).


Is there more information about this somewhere? How the light is distributed along the arc and at what angles it radiates?
I'm interested in modeling the light distribution of the arc and how much actually makes it out the triplet to the screen.
 
Pagercam:

There is an OSRAM paper titled "Metal Halide lamps Photo Optics", that discusses this stuff in section 4.2. Huge amount of technical info on MH lamps and their lamp types. See if you can find it for download from an OSRAM website. The copy I have is a badly scanned pdf of 6.5 Mbytes. If you can't find it out there, I can email it to you. (Hopefully, you have broadband!)

juice_e:

If you are making a short-throw design, then the 220 & 330 fresnels are fine. (That means you are using a projection lens with a focal length around 300 mm.) Then your lamp arc image will be more like (330/220) * 37 mm = 55.5 mm long. You can probably get that through an 80 mm diameter triplet.

You can do what I would: Look for online stores that sell M59 ballasts. Some of their websites list the lamps that work with each ballast, and vice versa. Then check out each lamp. Really good small arc lamps tend to require a more expensive electronic ballast, because they are designed for best performance. Big parking lot lamps all use cheap magnetic core ballasts, because they are competing on price.
 
Thanks for the reply guy. The reason i like that lamp is because the CRI is really good, most of the lamps for a M59 ballast have CRI of 65 or 75 max. It sucks living in Canada because there is little places to get the type of equipment for these DIY projects. And most of the places with great deals are in the US so after the exchange over to Canadian $$$ and the extra cost of shipping, it ends up costing more . So than i have to rely on ebay for my stuff. But ebay tends to have almost mostly M59 ballasts available to Canada at least. Maybe ill move to the US, buy all the parts i need, build it than move back to Canada and enjoy it ... :)


I also stil need input on the reflector size, how would i know the size of the ref;ector bowl i will need. And yes, i am using a short throw design with the 220/330 fresnels and 80mm pj lens, 15" lcd. unsplit, if that makes a difference. I have everything expect for the lamp and reflector which i am taking my time to find a decent one at a decent price.
 
Guy Grotke said:
Pagercam:

There is an OSRAM paper titled "Metal Halide lamps Photo Optics", that discusses this stuff in section 4.2. Huge amount of technical info on MH lamps and their lamp types. See if you can find it for download from an OSRAM website. The copy I have is a badly scanned pdf of 6.5 Mbytes. If you can't find it out there, I can email it to you. (Hopefully, you have broadband!)

Thanks Guy, I actually remebered that I had a copy and read it a about an hour after this post. So I have that information and an trying to think how I might turn this into some model/simulation of exactly how much light gets into the triplet based on a few parmeters like arc length, fresnel FL and triplet diameter. The objective being a way to figure out what bulbs will work well and which don't which also needds to consider color temp and CRI. For example if you compare the most common MH spectrums the major spectral line is in the only spectral band that LCD filters don't pass. The problem is that some bulb companies give spectrum but not arc length, or arc length but not CRI but atleast I hope to have a analytical method to evaluate bulbs.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.