a couple of us over in the pj section are wondering if a CNC machine can cut 60 degree prism array in acrylic, and is acrylic polishable after cutting? think of this as a straight grooved fresnel lens... I was thinkking that you might be able to use a cutting bit with 60 degree sides(if there is such a thing) and just cutting a series of grooves in a piece of acrylic, if the grooves were spaced properly you would end up with 60 degree prism array, no? we found a patent thru google patent search where it appears you could aim two collimated beams at this "straight fresnel" at a 60 degree angle and the beams would be deflected back to 90 degrees and thence into the field fresnel....if it works this would make doubling the pj output a fairly easy thing to do.... but this is theory... the smallest 60 degree prisms we can find(to hand build an array) is 3/8" which might show whether it will work , but the size of them also might introduce enough other effects that the test might not be valid
if polishing the acrylic is not possible(or it would not be smooth enough on a small scale) then can the grooves/prisms be cut into metal, which could be polised pretty finely and then we could cast an acrylic "lens" off of that? if the idea panned out we could probably get a group together and have a mould made to order by a fresnel comapny.... ideas or thoughts guys?

Can you put together a drawing showing your diminsions, i use a 60deg. bit alit and can try to cut some acrylic.

joe

well, heres a couple of really bad pics , basically 17x11" rectangle(to fit my 19" widescreen)
Click to view attachment

and a side view of the prism layer:
Click to view attachment
and yes, feel free to laugh.... my kids took the crayons away years ago when they surpassed y drawing skills(they were about 5)

anyway... as for the size of the preisms..the smaller the better, the smallest we can find is 1/2" wide triangles.... I'mm thinking if possible, 1/8" but I have no idea how small you can go
it really is like a fresnel lens except straight grooves(axis of groove parallel to short side) with a traingular cross section to the ridges...
any ideas or help is greatly appreciated....

so its just a bunch of 60 Degree slots in parellel with each other? I am pretty sure acrylic can be polished. I have polished alittle bit of it with steel wool i believe.

exactly, it just depends on how you look at it , its either a bunch of parallel grooves or parallel prisms !

I tool like this would do the job.

http://www.harveytool.com/products/product...graving+Cutters

So this is pretty much you want to make?

it would be a simple part to make even on a regular mill

yep thats it....... I actually was looking at hand tools to do it, but, well, lets just say my eye-hand coordination is not the best plus it sounded perfect for a cnc machine..IF acrylic can be polished really well ......
this is the original patent if you're curious:
http://www.freepatentsonline.com/5504544.html

probably, if I had any idea what a mill was ...the only reason I know what cnc machines are is because they're on lumenlab ..

now the question is...how small can the grooves/prisms be made?

How small do you think the groves need to be? id think it would be possable to get them as small as 1/32" or maybe smaller.

well, assuming the polishing could be done really well(because a 1/32 inch scratch on a 1/8" prism is no good but not bad, but the same on on a 1/32" prism would defeat the purpose..I think...) the smaller the better..again... I think..... this is all theory right now, the original patent is talking about microgrooves.... I think micromillimeter range, but obviously thats not feasible....personally I think on the range of fresnels would be plenty good(which are 1/2mm I think)..ummm trying to convert in my head.....standard fresnels are 2 grooves per mm, 1/32" is ...ummm.... anyway, the whole point here(asking you guys) is to see if it works reasonably well at any size we can test it, so 1/32 is(got the calcer out about 50% bigger than fresnel spacing so that would really give us a great idea of the whole ideas efficency..if it worked out and was worth the effort in making a dual lamp system feasible, then we could have a mould mad eby a fresnel manufacturer or something..... so again, as small as possible, but anything less than 1/8" would give us a good idea of its efficacy...

Have you done a little research on Lenticular Fresnels that already exist?

Joe

A lenticular fresnel is similar but it consists of a series of cylindrical lenses. What we need is a series of 60° prisms.

DJ

fluorescent prismatic diffusers come to mind

what sizes do they come in ...not sure ?

so... any of you guys want to take this on? of course I will pay for any materials or time involved..... it just seems like it has a good chance of being the easiest solution for using 2 lamps in a pj..... I looked up polishing acrylic and it looks like a high speed stitched felt wheel(the kind composed of many layers, not the solid kind) will do it.... still unsure about the size of the prisms though...Dazzz? I'm thinking for polishing purposes anyway that 1/16th would be the smallest I'd want to go and still give us a good idea of the patents feasibility or efficency... and even 1/8th inch might be good....

wouldn't it be better to machine a mold and then cast an acrylic piece?

injection moulding springs to mind...expensive

Im just wondering if someone like 3dlens who may have the equipment could do this...I think the smallest groove pitch is 0.2mm

Is this sorta nearly what we are after..kinda

You can see the prism angles are not what you want.. but makes you think perhaps it can be done


http://www.bhlens.com/linear_fresnel_lens.aspx

I will pick up some 1/8" acrylic next week and cut some to test.

That is a linear fresnel. Think of it as a cylinder lens that has been fresnelized (is that a word?) The same way as a PCX lens is made into a fresnel.
I’ve contacted this company and finding it difficult to communicate exactly what is needed. I emailed them a sketch to try to clear up the confusion but so far I’m in limbo trying to word my English so it is more Chinese friendly. Here’s the picClick to view attachment
There response was “It's the linear fresnel lens ,right?” Sigh…

I’ve also emailed a number of different fresnel manufactures but only received two other responses. Fresneltech do have a 60° fresnel prism. It is only 3.7”*2.7” @ 0.63mm pitch so 9 of these would be required to cover a 17” LCD. Their price is $25 each, $20 each for 5 or more, $16.50 each for 10 or more and $15 each for 25 or more. $165 is a bit expensive for a full test, buying one isn’t out of the question for a small test of principal though.

Here’s who I have emailed so far:
sales@fresneltech.com see above
bohailens@yahoo.com.cn still trying to communicate with them
info@knightoptical.co.uk no response
rhk@ntkj.co.jp no response
info@awi-industries.com not a stock item but can be manufactured



The DIY CNC approach:
A possible problem I see with polishing the facets using a tool that is not 60° would be uneven pressure on the peaks. Meaning the peaks of the individual prisms may be polished off and become rounded.
AWA mentioned that they could manufacture the fresnel prism on there in house diamond CNC mill which suggests that they may have a 60° polishing tool that they use in their CNC or that a single tool is used that is sharp enough and has enough RPM to cut the acrylic without any further processing. From some simple experiments I have done using a drill press and a normal drill bit, I could mill the acrylic but the surface is left white (diffusive).

DJ

Dazz as ever your in tune with my thoughts and as usual a step ahead ...did you try 3dlens

Its a pity we cant use the shop linear fresnel...suppose positioning wouldnt give us the angles we need..(Im not thinking to hard tonight)

Have found the use of linear fresnel in a Rear Tv Projection Screen Patent ...of course did they ever get into production..

The use of the sample would be proof of concept and not over the top expensive to try..but home made (or locally engineered on a mill or lathe..would be bestest.

yeah, I wonder if that might not be better, although that depends on if when casting the acrylic(dazz mentioned earlier that you can get acrylic beads meant to be melted in the oven, but I'm not sure if they would leave ummm.....streaks? not sure how to describe it...well..if you look at reall cheap plastic magnifying glasses you can see like "waves" in the internal lens....

the sample might be good to try(you could use a couple of well collimated flashlights)..but in a pm dazz mentioned that there might be some odd arc issues(I think because if you imagine the fresnel being 1/2 the size of the prism array, yet the output from the fresnel is "stretched" across the entire length of the array)) so I'd like to try it witha full size lamp just to make sure, the arc on flashlights being incredibly small....

cool! thanks a lot joe....

the polishing away of the peaks is what I was alluding to in an earlier reply, when I said I thought maybe 1/32" might be too small(in replation to the thickness of the layers of cotton of the polishing wheel).....whereas maybe 1/8 or 1/16th might be easier to polish with less deformation of the prisms...also why this one site I was reading said to use loosely stitched layered cotton or felt to polish it, the idea being that as you put light pressure on the polishing wheel, the layers would uhh...split at the top of the prism and go down the sides, whereas a solid polishing wheel would impact the top and abrade it away.... and it depends I guess on how much polishing the surface needs to..
and if we could show that this prismatic array would make a great dual lamp combiner and get enough interest, perhaps the easiest thing to do(other than find someone who speaks chinese really well would be to ship them a piece of the test sample and say make the pyramids smaller?
maybe I'll look and see if I can fin some of the meltable acrylic dazz was talking about...if it didn't leave internal "waves" or flaws, a metal mold would work....I'm sure the bottom(the flat side of the array) would need to be polished, but being flat, I think that would be fairly straightforward
I would guess dvds are made out of a much harder plastic but this thread is interesting on polishing dvds..
http://www.lumenlab.com/forums/index.php?s...=17806&hl=#

If you want me to try and machine something i can do it. just let me know.

rhk japan doesn't have them as stock.

DJ

well, I guess for starters we'd need a couple of pieces mostly just to test polishing on...dazz....I'm thinking that (if possible smackre or joe) 1/16th" prisms on a 1/8" thinkness piece of acrylic would be a pretty good test? as far as size goes, dazz, you have any suggestions? if I was asking for the moon, I'd say a few(4-5?) smaller pieces to practice polishing on(and possibly test a couple of nicely collimated flashlights on), maybe 6 - 10" square ? I could ship dazz a couple and keep a couple, that way we'd have 2 differnt mindsets working on it... once we saw how the polishing went , then I'd like to try a full size test in a pj, to see if there is any artifacting , shadowing etc...

1/16 on 1/8th acrylic sounds like a good starting point I don’t hold much confidence in polishing it though. I’ve tried too polish an acrylic test piece that I cut on my make shift mill with a layered cloth polishing mop and it did as I thought it would, rounded the peaks. We need some sort of polisher that has a mating 60° surface so that it polishes without causing pressure to the peaks. Maybe Joe will have some luck and the cut doesn’t need to be polished.

DJ

I emailed bhlens.com again, this is what I wrote:
“No not a linear fresnel.
A linear fresnel has different angles to each facet. A fresnel prism has the same angle (60°) for each facet.”

And their response:
“The linear has the same angle too.”

Big sigh…

actually, even if we can't polish it, a few small samples would do wonders to cross the language barrier nothing like seeing something in person...well, as long as we can make them understand it needs to be polished
well, we could try 2 pieces together and some really fine grit.... fine enough and probably you coule buff it ithout impacting the peaks.... or..... a metal mould and we see how well acrylic casts...

for polishing...

what if you took a piece of acrylic that has been CNC'ed and strapped on pieces of buffer cloth on each facet, then use that piece to buffer other acrylic pieces. You wouldn't have to make the buffer piece a whole fresnel either, 10 rows might be enough.

Don't know if that would work, but thought I would throw in my 2 cents anyway.

I would try polishing it with steel wool. It worked vary well when i did it b4. But as for taking the peaks off i have no solution for that. Maybe if a diamond router bit was used it would leave it nice and polished. I am not sure if a solid carbide router bit will make it polished. Diamond maybe though. I will look into a diamond bit.

cool ! I think the finer the surface is left,even if not polished totally) the less wear on the peaks..hey...could we spray some kind of shellac or some kind of coating on the surface if polishing won't work?

hmmm that would fill in the holes.light dusting coats...

but will it change the way / angle the light travels thru and exits the sheet..?
only one way to find out..

Acyrlic -urethane spray springs to mind..gloss..semi-gloss or matte I think

I think this is what has been used in some AVS forum DIY PJ Screens. so I guess it's a possible

sounds good, remember , this is for testing purposes mostly at this point. basically just to see how even the light spread is(any artifacting, banding,how the arc image is affected etc), and how efficient the "turning " of the light path is

I was going to respond like this: Unless the coating had the same refractive index as the acrylic I don't think it would work. The inside surface where TIR occurs would still be diffusive. But now I'm questioning my own logic. If it looks clear to your eye then why shouldn't it be clear for visible light? Hmm... perhaps I'm missing something.

DJ

You could be right..but it is sorta acrylic..I guess only one possible way to know for sure..thinking of a test we could do to prove either way..hmm gloss semi or matte...no matte is more diffuse

I was thinking along the same lines....the index of refraction might be close enough that, while not ideal, it may only detract a few percent.....lets put it this way...if we try it and get a decent result, a "real"(ie fresnel manufactured) one could only be better!

I've got a single acrylic prism that I might test to see if TIR still occurs after a touch up with some paint. I’m a bit hesitant, it’s the only one I have.

I’ve dissected one of by sons flip cards, I had to trade it for a bottle of soft drink . The card I’m talking about is one of those cheap cards you get in a cereal packet or the blinking eye glasses, look at it from a different angle and see the image flip to another. I thought these things might be the ideal fresnel prism we were looking for but I’m afraid they aren’t. After I dissected it I shone a laser though it I hoped to get the beam to bend but all I saw was the dot become stretched to a line. I new exactly what it was, a lenticular lens.

I’ve also scratched up a corner of my single prism to see if I can re-surface it using paint or some other coating, not looking good though. TIR doesn’t seem to worry too much about the different refractive index as I first thought. Although the coating fills up the diffusive scratches and makes it clear again, the light passes through to the outside surface of the new coating were it is TIR there instead. But getting the new coatings surface flat and even is a problem.

how "fine" were the scratches? I think the key is how smoothly the bits in the cnc will leave the surface of the acrylic.... and on that I know nothing

1200 grit paper, enough for the acrylic to turn white

I also looked into those acrylic beads I was talking about for kids sun catches. I couldn't find one that had clear beads, only white which we don’t want. That doesn’t mean they don’t exist, just not in the toy store I looked in. Of coarse pouring reson still could be a possibility.

damn.. Iwas hoping you'd say 150 grit .... well, lets see what joe and smackre try .... I'll pick up some acryllic next week and use a couple of different grits on it and see hoe it polishes up maybe..maybe (lightly)sand a couple of corners and polish them up..... I'll go google casting acrylic too, see what pops up

to cast acrylic resin do you have to use an autoclave? i've been looking for acrylic resin but it doesn't seem to be commonly available like epoxy and urethane casting stuff.
pouring with a mold would probably be best since its how they make cell cast shapes and sheets. but an autoclave requirement would throw a wrench in that method.

yeah, it looks like you do need an autoclave for acrylic.... heres a good link on casting resins:
http://users.lmi.net/drewid/resin_faq.html
it looks like getting somehitng useable(and keeping the prisms nice and even) will be difficult by casting....