The Frankenfresnel

What do you call it when you place three fresnel lenses side by side and use three light sources? I think this huge three eyes monster should be called the…

FRANKENFRESNEL

The goal hear is to use two or three fresnel lenses attached to each other to allow the use of two or three light souces. One thread was started on this and I have been waiting to test the idea for month’s. However my fresnels are on a slow boat from china!

Lucky for me Staples has a cheep page magnifier with a 330mm fl made out of some kind of acrylic for $8.69 so I bought three. [edit] five!

The two main problems for a lot of the PJ is to short throw and not enough brightness. I am going to be building an outside only PJ so my requirements are going to be a little different than most builders.

There is many ways to make the PJ brighter more efficient reflectors or bigger lamps a totally new curved fresnel and soon to be for everyone the pleasure to skin your LCD and replace or modify the LCD. In my case I don’t care about heat if I can cool it or power consumption or lamp life. It’s no big deal if the PJ only on for two or three hours a week. The lamps I used are Smith Victor 250 watt halogen lamps I happen to have three to use for testing and if things go well I may use then in the final PJ.

For a really large screen the main concern above all is brightness. One 400 watt lamp is not going to cut it. We may soon see how a 1500 watt works on the Photon cannon. This will be interesting.

A lot of people have a 400 watt system right now. If this works well and they want to double the brightness of their PJ only one lamp system and
one new fresnel would be required.

This is the brute force method. The use of multiple fresnel allows multiple lamps with this idea an 800 watt systen is possable.

[edit] This the only way I have seen where you can increase wattage without increassing arc image size or the rear fl. The two or three arc images add on top of each other. Thus not requiring a larger triplet! Infact a smaller lamp could be used making a smaler arc image letting some people use smaller triplets.

The first test will be using two fresnels and two lamps. Soon I will have the fresnels joined togather at the plastis shop for a more perfect fit.

[edit] It looks like the joining of the fresnels needs more work.

The second test will be using three fresnels and three lamps and have them solvented togather[edit not going to do three]

The first test ,which is already done, is using two fresnel lenses cut and lightly sanded then taped together.

Lamp adjustment was rather interesting. I set my make shift screen at the triplet focal point. You could clearly see two lamp images on the screen. The image on the left was the right lamp and right was the left.. Even brightness was easy to get just adjust the lamp images on top of each other. by moving the lamps togather slightly the line vanished!

This works so good that everone that has $20 to blow on two fresnels and a couple of lamps has got to try it!

Thanks Dave.

you have them side by side? not both focused on the "lcd" ? do you have a diagram?

A simple drawing of the test setup.Click to view attachment

Hear is some first pictures. The 1" wide tape can be seen bit the crack between is almost invisible.Click to view attachmentClick to view attachmentClick to view attachment

Some more.Click to view attachmentClick to view attachmentClick to view attachment

one more bigerClick to view attachment

I might do some pictures without the tape and just the crack. The heat makes the fresnels warp so I put tape on the whole fresnel. it looked better the first time with no tape

it looks like only 1/2 the light from each of the collectors will hit the 3rd fesnel..how does that increase light?

no tapeClick to view attachment

Interesting. The joint seems to look ok. I did not expect that.

But I think the issue is we need shorter focal length fresnels otherwise theirs no improvement.

The crack is rather large you could put a nickel in it!Click to view attachment

But it looks like you don't see it in the projection.

There is shorter fresnels out there. Most will now fit becouse you can place them sideway. Using two fresnels cuts the distance from arc center to edge in half. This is a big deal. A shorter fl could be used without dim edges. This system is already more efficient because the angle the light hits the fesnell has ben reduced and distance to the edge has ben cut in half. A blind guess would be 25% inprovment,[edit] but the area that each lamp sees is half so this will hurt the efficiency some mabey a wash?] mabey ROX could help here.
I can get a light meter and I do have a 500 watt lamp for the same fixture so I could do a test two 250 watters V one 500 watt.
I think we could get a double frenell made. that should be a lot easier than the curved fresnel idea if they made the joint with a slight amount of dithering the seam could be invisible.
If you were going to have is made it would be a good time to shorten the fl too.

how big is the pj?

50" diag. one 500 watt china lamp on lots of windows

ah... the old pythagorean thingie ;-) hmmm...

Jjust the light.[edit] had plastic shop solvent two fresnels together

.Click to view attachmentClick to view attachmentClick to view attachment

One reason I wanted to use three fresnels is the center image would be perfect no mater what. There might be a small flaws on the sides but is would be out of the main viewing area.

A few things that I learned today about the joining the fresnels. First the Plastic shop had no idea what to to with the cut [prism] side of the fresnel. The solvent ran into some of the grooves and distorted the prisms. They say they use a special cutting blade to cut the plastic and there cuts were real smooth.

The process on making the line between the fresnels dissapear is becoming more clear. If you move the fresneles all the way foward and adjust the lamps for max brightnes there will be a nice dark line. Slide the fresnels backward slowly there will be a point when the line is gone move father and the seam becomes a lighter line instead of a dark line.
I think that the "pro solventing" of the fresnels didn't help. The lenes now has solvent in a lot of the grooves near the cut. there may be a way to seal the grooved side with sealer then apply the solvent to the other side then peel off the sealer.
I think the dark crack will still be there even if a perfect seem could be made. Making it blend in is the real trick. If the rear of the fresnel where the seam is could accept more light this might help so next I am going to make fresh cuts then angle the back side to hopefully make the blending effect better. The fresnels then can be placed between two pieces of glass or plastic.

[edit]
Gota get my heat shield in. The tape around the triplet was smoking last night!!
Heat is going to be a real problem so I might spend some time on heat blockers and cooling.

One new idea about using two Fresnels that it would allow the use of two mirrors and one lamp this has some good benefits. First is a shorter box second is the use of only one lamp. The light pattern should be vary even and there would be no bulb matching.
The system would only offer slightly greater efficiency due to less distance from the far points of the Fresnel to the lamp arc. But using a mirror instead my yield greater light output than using a spherical reflector.
I have some front surface mirrors to use so I will set up a test after Christmas when I get a new light meter.


Click to view attachment

Any updates on your testing?

I'm waiting until I get two heat shields so I stop melting the Fresnels. I now have 1 real 330mm front Fresnel but I melted one more rear Fresnel
I have gone through 6 of the cheap ones I keep breaking them.
I still hope someone will try it our on a real LCD with a mh light system but it seems no one cares.
In a week or so I will get A light meter then I will do more testing.

Thanks for the interest.
Dave

I like your idea Dave, if I wasn't doing what I was doing I'd give it a shot. I think the idea has plenty of potential, but sadly I also think if you don't develop it no one will. Just do it, post pics and if it works out you are a hero because everyone will copy you. If you fail... well then your just a waste of good bandwidth, hehe.

I like it, keep going.

Dave, I like your idea as well, and have a subscription to your thread

The process had crossed my mind, but since my knowledge on projectors is SO limited, I didnt' know where to start.

As a matter of factly... I still don't! I'll be paying attention - so keep up the good work!

I have been following th is thread since I signed up. I loved the idea of using 2 250w MH bulbs instead of a single 400W or higher for that matter. I think you can get better light concentration with the smaller arc lengths and closer to that elusive point light source. If we can properly achieve this w/o the line/seam causing a problem then I see this being a great way to achieve a brighter image.

It seems (no pun intended) to me that the line in the center would not be a problem if the two fresnels where angled in towards eachother just slighty. I personally do not beleive the light needs to be perfectly collimated, just roughly. Angle them in towards eachother and the seam should disappear.

Or so it seems.

You have to offset the lamps to make the line go away and yes I said in one of the post that if you angle the fresnels it makes the line go away just fine and it lets you place the lamps in the center brightest spot.

This is one way to make the light both more even and brighter. Two 250 watters would be brighter than a 400 watter.

I have nothing to contribute except:

Bravo!

This is cool!

Had a thought......

If you stick a precondensor between each bulb the fresnel its pointing light at, then you will actually atain double the brightness with your "double bulb" setup.

It will get round the fact that you have halved the field of view from fresnel to bulb, by halfing the size of the fresnel......

Adding the pre condensors keeps the amount of light hitting the fresnel constant, regardless of its size!!

Any updates?

Thanks for brining the thread back from the DEAD.

And yes I have a fix for the line in the middle problem its simple. I was trying to seam two Fresnels together with little success. The seam was never that good and the solvent always got in the groves and mess thing up.

What we need is a perfect end of one fresnels light pattern and a perfect start of the next light pattern. The answer is not to make them seam up evenly [can't be done] but to make the light stop at the end of one fresnel and start at the next one in a perfectly straight line so the two light patterns can then be matched up to each other. The line only needs to be truly straight.

The answer is just to use a small piece of metal maybe with a T cross-section to connect the two fresnels if the metal is straight the two light patterns will be nice and straight too. then you can either angle the fresnels or move the light to make the light patterns line up yep it's that simple no you don't even have to cut the fresnels vary straight and you will have a perfect light pattern. Cool

Of course I have not tested this yet. I have only one good small Fresnel right now I will be getting to this in a few weeks but please feel free to go to office max and try this first if you wish.

Before I though of the fix for this I decided it would be easer to just run one large lamp for my PJ. Two 400 watt lamps just wasn’t enough!

When the new LL lamp’s gets here there will be lots of extra lamps and ballast for sale cheep so this idea may make for a vary cheep twin power PJ.

I have a slightly different plan for the two mirror setup that will be even more efficient with the new lamp but I am working on testing the super light engine and my soon to be 'the megawatt pj"

Thanks for asking

Dave

This is what I've really been waiting for, cheap ballets! Bring on the babes in tights!

Yes! The last Cheap Ballet I saw was "Mallard Lake", a low budget, off broadway production. But I have to say it was pretty good. I thought it was a lot better than their previous production of "Drake Quixote" and even better than "The NutQuacker".

The plumage was beautiful.

Fixed. that's ballast. I'ts not like this thread is named after some cheap sci fi movies or something.

Thats fowl humour!

Well, I was thinking I pretty much laid an egg on this and I should probably take flight. But I thought a bit about it and decided I would just wing it. Quite frankly its a wonder nobody flipped me the bird...

If you place a condenser lens on each lamp this system could be twice as bright as a single lamp system. A duel core PJ.

Hi- This might be a stupid question to most of you.. but why cant we just do something like this:



or use mirrors to direct light from all the lamps straight

It's because the bulb has to be at the focal point of the fresnel. Each fresnel will only have one focal point, so multiple bulbs with 1 fresnel would be very dificult.

The straight line from the bulbs in your drawings make it look like it might work, but you have to consider the whole "light cone."

arizonavideo, I'm glad you started this topic! I wish I had seen it sooner, because I've been wondering about this for a long time now.

Personally I think the 1 bulb, 2 fresnels, 2 mirrors looks like a great idea! I think you said that it wont offer much of an advantage over just using a spherical reflector. That doesn't seem right to me, the way I understand it, the reflector only adds ~10% more light, not the 100% that you'd get using this setup (less the light lost due to ineffencies of the mirrors). Another big advantage is you'd have 2 hotspots, which of-course would make the light look more even. Also, there would be less (I don't know what it's called) lost light from the light hitting the fresnel at too great of an angle. Not to mention the shorter box length.

I wonder if you could work in enough mirrors to have 4 fresnels, 4 hotspots, maybe about 175% more light?

Oh, actually after making some 3D models, that using 2 mirrors with 2 fresnels, but with only a single bulb would be about 16.5% brighter than a single fresnel without a spherical reflector. Using 4 mirrors with 4 fresnels would be about 26.4% brighter than a single fresnel without a spherical reflector.
If a spherical reflector only increases the brightness by 10% it woul probably be better using the 2 mirror method, and would definately be better using the 4 mirror method. Seems to me that the spherical reflector MUST increase the light more than 10%, it's gotta be closer to 30% if not more. In which case it would not be worth the hassle of adding a bunch of added cost and complication of putting in mirrors, and cutting more fresnels and stuff.
This is only talking about using multiple fresnels, but with only 1 bulb.

OK, I figured I'd explain how I came up with those numbers.

More or less, I modeled how the light would have to travel from the bulb to an accurate model of a WUXGA LCD size fresnel in each of the 3 senarios. I modeled it making the total distance from the bulb to the LCD 220mm taking into account the mirrors.
I then made the same size sphere around each of the focal points and cut off all of the spheres that the light did not go through. After that I could measure the area of the sphere that remains. Here's pretty much what I ended up with...
On the left there is the 4 fresnel model, on the right the 2 fresnel model, and in the foreground the single fresnel. Each of the yellow lines is 220mm long. In the center you can see the remains of the 3 spheres.

Here are the areas:
Single: ~121.33 square milimeters
Double: ~141.36 square milimeters
Quadrupal: ~153.39 square milimeters

I never looked at the areas so closely. If you are at work and they are paying to play around you might also look at a 3 Fresnel system with a center Fresnel of 220mm fl and both side reflectors 330mm but with a condenser lens of 400mm. for each side reflector.

After the computer meltdown trying to figure all the distances I did have other another idea.

The main fault of the two Fresnel systems is that the collection area is halfed so if you have two lamps the light output would not be doubled.

If you have two lamps and add a condenser lens then the size of the collector Fresnel is no longer dominates the condenser lens is. Two lamps would now make twice the LUX with out the lamp arc image being too large to fit in the triplet. So two sd400 lamps and two condensers lens with two 330mm Fresnels will be twice as bright as a single sd400 setup.

You could now have high brightness and long life.

There is still the lamp matching problem and having two of everything.

I'm still not exactly clear how to calculate... anything when it comes to precondencers. If there's a thread that explains it clearly; I'd be glad to do what you're talking about. I'm at home, but I've got the time anyway...

Any one want to try this out again?

I was looking for this topic... glad you brought it back... looks like elken has some sort of double lamp thing going, but it's not clear if he's using two fresnels or not... it kinda sounded like only 1.

I want to try this, but it will be some time before I have time to do a project such as this...

gs

I am glad I happened upon it, very interesting, but I am not totally understanding how there isn't double images of everything? Is it because we are still only using 1 front fresnel?

Yea only one front (field) fresnel.

gs

Well, my thought would be that you would need to maybe have a wall that seperates the two light sources, since if one crosses into the others fresnel, the light would infact give off a multiple image. Just thinking.

Yes you would need a divider. I did not use one because the lamps I used had a fairly focused beam.

Two collector Fresnels (or three) and one field fresnel.