How much light does a prism diagonal lose?

[DaveM]

I have a prism diagonal that I am using with a Coronado Helios 1 solar telescope. I was wondering how much light is lost with the prisms? The solar image is pretty bright, but I was wondering if a mirror diagonal would show fainter bits like prominences better?

[gbeal]

I have two (high quality) prism diagonals, and an elcheapo mirror. If I need the best I use the prisms, other than that I use the mirror. I seriously doubt you will notice any difference.

[g__day]

Surely you have to evaluate two criteria - light loss (less than 100% reflectivity) and light scattering (less than 100% perfectly flat surface). I'd imagine the second factor comes more into play at the high end range where you can buy mirrors or prisms tested to 1/15 of a wavelength of light vs those only managing 1/8 of a wavelength?

[DaveM]

Thanks for the comments. I hadn't thought about scattering.

I guess I'll just have to borrow a mirror diagonal and see the difference for myself.

David

[janoskiss]

AFAIK it is much harder to make a ~99% reflectivity mirror than prism. A high quality prism diag is probably your best bet. Chromatic aberration is the only downside of prisms vs mirror. I believe that in high quality diags, crystalline materials like quartz are used instead of glass to minimise CA. (But for a Ha solar scope CA is a non-issue, as the image essentially contains just one colour.)

Another plus for the prism diag is that it does not degrade with time as mirror coatings eventually do.

[DaveM]

My prism diagonal is a cheap and cheerful GSO ($40) or something, so I am not exactly top of the range. I'll do some comparisons with other that I can get my hands on.

David

[Don Pensack]

Your prism reflects 1.5% from each outer surface, and 1% from the interior, making transmission 96%, about the same as an enhanced aluminum mirror.
But, there are two refractive and one reflective surface to add wavefront anomalies. They could add up or cancel--it's the luck of the draw. However, it's unlikely your unit has optical surfaces better than 1/8 wave, making the possible error 1/8 wave + 1.44(1/8) + 1/8 wave, or 0.43 wave. A mirror with the same accuracy would yield a 0.18 wave inaccuracy in the wavefront.
If the light cone entering the prism is from a scope with an f/ratio less than f/12, there is also some prismatic dsipersion taking place. There is no such dispersion from a mirror diagonal. The same mirror accuracy will generally yield a better image. Prisms (even well made, high-end ones) belong to refractors of yore, IMHO.

[janoskiss]

GSO 2" dielectric (i.e. prism) diagonals are listed as 1/16 wave, 99% transmission at Andrews.

[g__day]

Great post Don, I hadn't thought the interference patterns at all for a prism, very nicely explained!

[DaveM]

Just a follow up on this thread. I now have a Williams Optics dielectric diagonal (1/10 wave, 99% refl) for the Helios and tested it against the prism diagonal. The image with the WO 'feels' clearer if that makes any sense, but the resolution, brightness and contrast of fine solar disk features seems the same. Perhaps if I was a more experienced observer I could tell the difference, but I suspect the difference is only incremental. Still, the WO eyepiece lets me focus without having to pull the diagonal out slightly, so that's a plus.

David