In most reflecting or catadioptric telescope designs, a secondary mirror is
required in the optical path to send the light from the main mirror to the
correct position for viewing.  This mirror and its holder obstructs some of
the light entering the telescope, and for larger secondaries, this has the
potential to cause some problems.
   A central obstruction affects the image in two ways:
1. Causes a light loss due to the blocking of light entering the telescope.
2. Introduces diffraction effects which can cause a slight loss of both light
and contrast for high power images if the secondary is too large.
   As far as light loss is concerned, the best human eye can just detect a
difference of 0.1 magnitudes, so with obstructions less than 30 percent of
the aperture diameter (9 percent of the area), the actual light loss caused
by just the blockage is essentually undetectable visually.  Indeed, many
people have trouble seeing a magnitude difference of 0.2 magnitudes, so for
obstructions of less than 41 percent of the aperture (16.8 percent of the
area), the light loss due to the obstruction is not all that noticable.
     Of these two effects, the second (diffraction) is more significant.  The
obstruction from the secondary and its cell slightly alters the disk and ring
diffraction pattern of stars, taking a little light out of the central Airy
disk and putting it into the rings (mostly the first ring out from the Airy
disk for common-sized obstructions).  If the secondary is large enough, this
energy redistribution can result in a slight reduction in the contrast of
fine detail for high power images of the moon and planets.  How much a
problem this is for the observer depends on how big the obstruction caused by
the secondary is.  In practice, if the secondary obstruction is less than 20
percent of the main mirror's diameter (1/5th of the mirror is obstructed),
the effect on the image is negligable.  For example, a six inch with a 1 inch
secondary mirror (16.7% obstruction) would perform about as well as if the
obstruction wasn't there at all.  Indeed, the secondary's obstruction can be
somewhat larger without hurting high power images all that much.
     As you use larger and larger secondary sizes, eventually, high power
images will tend to gradually aquire a slight "softness" to them, which may
make fine low-contrast detail a bit harder to see.  A certain limit comes
when the secondary obstruction does become somewhat more noticable, and that
point is when the obstruction reaches about 25 percent of the main mirror's
diameter.  This amounts to a six percent light loss, and becomes significant,
especially for daytime use, when the shadow of the secondary mirror may
become visible in the eyepiece at low power.  Only Rich-field instruments or
those requiring large fully illuminated fields should have secondaries which
obstruct more than 25 percent of the aperture.  One frequent claim by some
authors is that a larger secondary can help increase the apparent resolving
power of a telescope.  This is somewhat of an exageration.  While the
diffraction caused by the secondary obstruction does cause a reduction in the
diameter of the Airy disk, the actual amount of reduction for common central
obstruction sizes is slight, and would not significantly improve the ability
of the telescope to resolve close double stars.  Indeed, the diffraction disk
of a telescope with a 20 percent central obstruction is only about four
percent smaller than that of an unobstructed instrument.  Even a 33 percent
central obstruction would only yield a 10 percent reduction in the Airy disk
size, so for common central obstruction sizes, the "improvement" in effective
resolution is negligable.  The amount of energy put into the first ring by
the obstruction would negate any alleged resolution increase on extended
objects, so it is still best to keep the secondary obstruction at 25 percent
or less if possible.
     As long as you keep the secondary's minor axis size below 1/4 of your
primary mirror diameter, the telescope should yield good images.  However,
even at a 25 percent level, the image degredation is far from fatal, and the
telescope will still function.  As for resorting to ultra-small secondaries
to improve the telescope's high-power performance, this can backfire.  You
don't really get significant image quality improvement by using a secondary
size much under 20 percent of the main mirror's diameter, and you may
actually lose light with a secondary mirror which is too small to catch all
the light from the primary mirror.  In addition, low profile focusers used
with such small secondaries may allow some external scattered light to get to
the eyepiece without being blocked by the diagonal.  Small secondaries of
high optical quality can also be hard to find or to make.  You need to
balance the desire for high power contrast with the need for proper overall
design, and for this, the 20% to 25% obstruction guideline is often a good
rule to follow.
    Overall optical quality is more important in the long run than is how
small a secondary mirror your telescope uses.  A quarter wave of spherical
aberration has about as much effect on the overall energy in the Airy disk of
an imaged star as a nearly 33 percent central obstruction does.  The effects
of a quarter-wave (peak-to-valley wavefront error) mirror, while just meeting
the Rayleigh Limit, would basically overshadow the benefits gained by the
use of a small secondary mirror.  Resorting to a tiny secondary mirror while
tolerating a lower-quality mirror makes little sense to those who are trying
to get the best planetary images from their scopes.  Indeed, many of the
often talked about performance differences between various telescope designs
are due more to lack of mirror quality than to the problems caused by the
presence of a secondary mirror.
    The "modified" Schmidt-Cassegrain telescope needs a much larger secondary,
often obstructing 33 to 35 percent of the primary mirror's diameter.  This
does cause a visible loss in contrast for high power images and a slight
reduction in limiting magnitudes for stars, but overall, the telescope still
performs adequately.  Indeed, many planetary observers do successfully use
SCTs.  The tradeoff is in contrast verses telescope compactness.  The SCT
does offer a very convenient package for people who want portability or ease
of use for photography, so the secondary obstruction isn't the only factor to
consider.  Once again, optical quality is the most important thing to have
when it comes to an astronomical telescope.

David Knisely, Prairie Astronomy Club, Inc.