Star Projector

[Steve Hamley]

It's a cold, gray morning about to rain here in Tennessee, and I'm not inclined to get out until the rain washes the salt off the roads which means I have too much time to think...

Bob Salomon has mentioned the Linhof and Sinar use a $500k Siemens star projector to test lenses, and I want one. Note that I did not say that I wanted to BUY one! Although when I Googled "Siemen's Star Projector" I did get a lot of hits to compare prices and write reviews for one.

So how does this thing work? I assume a lot of the high cost is producing an easy-to-use, rugged, high volume machine for Linhof and Sinar, so maybe I could make something that remotely resembles one. It seems to be something like projecting a pattern and looking for changes as the lens is rotated.

Questions:

How does this thing work in theory? It sounds sort of like an enlarger projecting a "resolution chart" with the lens to be tested used as a projector lens.

What's a "star"? Is it a 5-pointed thing like we're accustomed to, or is it something like a fine pattern of 3 or more intersecting lines, etc.

Are all the "stars" the same size or are there large ones and small ones? I assume they are very precisely cut as in laser cut?

What's the light source requirements?

Is there a "reference lens" that candidates are compared to, or do people just use their judgment about what's good enough?

So could I use an 8x10 enlarger head to project an image of a laser-cut pattern through a test lens mounted in a "lens chuck"?

Well, that should be enough to incite some comments.

Cheers, Steve

[Arne Croell]

This is a Siemens star, a specific pattern: http://en.wikipedia.org/wiki/Siemens_star

I have seen one of those machines about 19 years ago during a workshop at Linhof. Yes it is like a slide projector with several Siemens stars or other patterns as slide. What I seem to remember is that the lens to be tested is mounted on a rotating mount, driven by a remotely controlled motor - that way one could stand at the projection screen (actually a large wall) and check for centering problems driving the rotational lens position back and forth. Of course there was also a means to move the lens axially to focus.

[Paul Fitzgerald]

"So could I use an 8x10 enlarger head to project an image of a laser-cut pattern through a test lens mounted in a "lens chuck"?"

If you could find a decent 4x6" microfiche film and glass mount it, it should work very well for this. The printing is exactly even across the whole film and would show any defects of the lens. They were extremely sharp and high contrast with different sized print across the film.

Good luck with it.

[Arne Croell]

Not a Siemens star, but 2 or 3 of these babies mounted on a larger piece of glass would be perfect as "slides": http://www.edmundoptics.com/onlineca...productid=1790

[Bill_1856]

It's a cold, gray morning about to rain here in Tennessee,
Cheers, Steve

Welcome to East Tennessee between Christmas and Easter (roughly speaking), Steve. I spent my first 30 years in that part of the country, and until moving away I had no idea how F***ing depressing on the human spirit those miserable dark months were!
(Incidentally, I didn't start using this kind of language until after I'd moved away -- it just wasn't done in those days.)

[Peter K]

So could I use an 8x10 enlarger head to project an image of a laser-cut pattern through a test lens mounted in a "lens chuck"?

No, the test projector is a collimator, so the test-pattern is for "normal" lenses optically located at infinity and for e. g. enlarging lenses in the distance the lens is made for.

Peter

[Steve Hamley]

Peter,

That's useful - so the projector has to be collimated light?

Cheers, Steve

[Peter K]

That's useful - so the projector has to be collimated light?

Yes, but over the whole image circle of a lens. This can be 500mm for a LF-lens. So this needs a highly corrected and focussable image not only in the center but also in the outer aereas.

Peter

[Arne Croell]

No, the test projector is a collimator, so the test-pattern is for "normal" lenses optically located at infinity and for e. g. enlarging lenses in the distance the lens is made for.

Peter

Why would the test pattern be at infinity? It would be where the film normally is, whereas the projected image would be at "infinity" (in reality at the right distance for 20x or 50x or similar enlargement on the screen). A collimated light source to illuminate the pattern is a different story, though.
You use a pattern (slit or edge) at infinity for MTF measurements, but that is a different setup.

[Steve Hamley]

Peter,

But "collimated" light is basically light coming in from an object at optical infinity, right? So in theory I could use the sun as a collimated light source - direct sunlight that is, or light from a uniform sky? And the light coming out of the lens that's projected on a wall or screen is not collimated, so why would it make a difference if the source is collimated provided it didn't diverge highly over the star?

Cheers, Steve