The 1974 Telescope Secondary Experiment
I wanted the most from my new 14.25 inch [36cm] Horseshoe mounted
reflector. Conventional wisdom of the era was that a reasonably
sized secondary of 1/5 to 1/4 the diameter of the primary mirror was
fine, that there was little benefit to going to a smaller secondary.
Some people though were agitating that secondary size mattered a great
deal, that I would only be getting the very best images if my secondary
was small. Maybe there were onto something, maybe convention wisdom was
Worried that I might not achieve the very best images, I designed a
nifty self-collimating quick-mount secondary holder so that I could
swap back and forth between two diagonals: the smallest possible
diagonal for planetary and lunar imaging and a larger diagonal for
deep-sky use. What resulted was a surprise.
Here is the 1.83 inch diagonal in its self-collimating quick-mount
holder. The large diagonal was a 2.60 inch. I built a motorized slide
focuser with a single straight stalk diagonal holder. The 1.83 inch
diagonal was 1/8 the diameter of the 14.25 inch primary and the 2.60
diagonal was 1/5 the diameter of the primary.
In the 1970’s, besides doing a great deal of visual observing, I was
into eyepiece projection photography of the planets and cold-camera
long exposure astrophotography.
I compared the two diagonals visually over many nights observing the
planets with colored filters. I compared hundreds of highly magnified
eyepiece projection negatives and slides. For deep-sky I visually
counted stars in a small region of M13, the great Hercules Globular
After a number of months I was forced to conclude that visually there
was no noticeable difference on planetary images and no difference in
star counts on M13. To my utter surprise, after months of photographing
Jupiter and Saturn, I was forced to concede a very slight advantage to
the 2.60 inch diagonal, the opposite of my expectation. How could this
Could the 2.60 inch diagonal be covering a bad center zone of the
primary mirror? No, the mirror is without zones.
Could there be alignment or vignetting issues? No, I checked again and
Could there be a difference in optical quality of the diagonals?
Perhaps the 1.83 diagonal had a slight edge problem. After all, only
when a diagonal is absolutely minimally sized will its edge be used to
form the image at the center of the field of view. I was forced to
conclude that this was the most likely reason.
The real message though is that closely sized diagonals do not make a
discernable difference in visual performance. Intrigued I set about
making larger masks of 1/4, 1/3 and 1/2 the size of the primary. Over
several nights I saw a consistent result: the 1/4 mask give nearly
almost identical performance to the existing diagonals, but the 1/3
caused a modest degradation and the 1/2 sized mask gave sad images.
I also realized that the best mask for star testing spherical
aberration or overall correction of a telescope is a 1/3 mask.
Well, how about them apples, as my father used to say.
I learned to avoid worrying about secondary size, and became aware of
Parkinson's Law of Triviality. I also learned to ignore fear and
concentrate on getting the most from all the factors involved.