Baffling the Cassegrain Telescope
I recently upgraded the baffling on my 12.5 inch f/20 Cassegrain and thought I would
prepare some practical notes for your consideration. There is no new science here. The
technology for baffling a Cassegrain is quite mature. But, the improvement to the
image from good baffling, v.s. ill thought out or no baffling I found to be impressive,
so I thought you might be interested. Although the telescope described in this article
is a classical Cassegrain, the thoughts here apply to any type of Cassegrain
When I first made my 12.5 inch Cassegrain (read the story here),
the last detail in the construction was the baffle. I made the mirrors, made the OTA, was in a
hurry, and out of budget. Sound familiar? So, I found a tapered shop-vac nozzle at my local
home improvement store, fashioned it to fit into the central hole in my primary, trimmed the
front so I could see only the secondary from the focal plane on axis. Eureka! I had an instant
baffle. (See image below.) That approach has served me quite well for 8 years now and I have had
very favorable comments from people who looked through the telescope. But I always knew I needed
better baffling. It took me 8 years because I underestimated just how much improvement there
was to a good baffle system. I bet you do too.
I also discovered that visual observing of deep sky objects were breath taking at f/20.
Who’d a thought it? Sure surprised me. With a simple 40mm eyepiece, globular clusters
filled the eyepiece with fine points of light right to the edge of the field of view.
Galaxies were nice and big so you could savor them. Eyepieces perform their best at long
focal ratios. And eye relief is very comfortable. The problem was that extended objects,
like galaxies and nebulae, got dimmer with increasing distance from the center toward the
edge of the FOV, and the sky brightened also with increasing distance toward the edge of
the FOV. Not a lot. You would not notice this looking at planets or even bright globular
clusters, but with low surface brightness objects, it was a noticeable issue. Stephen’s
Quintet, for example disappeared when not in the center of the field.
The original "Shop Vac" baffle
The Baffle Design
The limitations to my primitive baffle system were that the shop-vac nozzle was too short,
and I did not incorporate a secondary baffle. Let’s just discuss for a moment what
baffling a Cassegrain is conceptually. In an optimized system, there are two baffles.
There is a primary baffle which protrudes forward from the central hole in the primary
mirror. This baffle blocks stray outside light, that comes past the secondary, as seen
from the focal plane on axis. And there is a secondary baffle which is a cup or disk
around the secondary that is larger than the secondary and blocks light that comes past
the secondary as seen from the focal plane off axis. So, with both baffles, the light that
is coming to the eyepiece past the secondary is blocked both on and off axis.
On axis, the front aperture of the primary baffle is centered on the secondary and the
front aperture of the baffle should be sized such that as seen from the focus point, the
diameter of the front of the baffle exactly matches the projected diameter of the secondary.
This is how the primary baffle blocks light on axis. However, as you view off axis, the
front aperture of the baffle and the secondary no longer stay lined up due to parallax.
Since they no longer line up, the baffle will vignette image light from the secondary on
one side and open a path for light to come past the secondary on the other side. The amount
of vignetting and the amount of light admitted past the secondary, due to this parallax, is
related to the distance forward the end of the primary baffle is. Parallax should be
minimized by insuring that the front aperture of the primary baffle is as far toward the
secondary as possible. But there is a limit as to how far the front of the baffle can
Determining where the front of the baffle should be is really easy. I lay out a sketch of
the optical elements in full scale. Usually, only one side of the optical axis is necessary
to draw. Draw rays from the edge of the secondary shadow in the primary, which is also
likely the size of the perforation, to the edge of the secondary reflection on the
secondary. Draw another ray from the system focus point on axis to the edge of the physical
secondary (which is probably slightly oversized). Where those two rays intersect defines the
location and diameter of the front of the primary baffle. That is all there is to it. Very
simple, conceptually. That done, you can dimension a tapered baffle scaled from this layout.
The taper provides the strongest structure, and the baffle walls will be edge on to all the
image rays in the vicinity, so you will not vignette and loose light to the image from the
baffle structure itself. See the sketch below.
Once you have the primary baffle laid out, it is real easy to conceptualize the geometry for
the secondary baffle which will block light from coming past the secondary off axis. Draw a
ray from the edge of the field of view (FOV) at the image plane diagonally across past the
front lip of the primary baffle. Draw this ray past the secondary. The simplest secondary
baffle would be a disk around the secondary of a diameter equal to where this ray intersects
the plane of the secondary. The central obstruction can be further minimized, however, by
drawing another ray from the edge of the primary to the edge of the secondary on the same
side of the optical axis. Where those two rays intersect defines the diameter and axial
location of the aperture of the secondary baffle. This diameter is a bit smaller than the
diameter of a disk at the secondary mirror plane. Usually a cylinder or tapered cup can be
made to match this geometry. See sketch below.
My point in describing all this is that it really does make a difference that can be seen
visually. With the optimized baffle system, the background sky is a uniform shade of black
across the entire FOV and the light retained in dimmer extended objects is much greater. My
recommendation is not to get too hung up on increasing the size of the central obstruction
due to the secondary baffle. I was, until it became very apparent that the reduction in
contrast due to light getting past the secondary overwhelmed the loss in contrast by the
relatively slight increase in central obstruction.
A very interesting corollary conclusion from this exercise is that if your secondary is
slightly oversized, image light from the edge of the secondary never gets to the image. The
parallax of the primary baffle covers the edge before the image gets there. Since it is
difficult to make a hyperbolic classical Cassegrain without edge defects, the answer is to
make the secondary slightly larger and those defects will not affect your image.
The sketch below shows a comparison of the image light and light past the secondary due to
parallax of the old short baffle and the new longer baffle. This sketch therefore gives a
graphic demonstration of the improvement of the new optimized longer baffle. Also shown is
how much smaller the secondary baffle, assumed as a disk at the secondary plane, becomes. I
(for now) have used the disk at the secondary plane because it is easy to remove and replace
if I want smaller central obstruction for collimation and planetary viewing at high
magnification. Note that a cup shaped secondary baffle, as shown in the above figure, would
result in a slightly smaller central obstruction.
Now that I have discussed what is needed conceptually regarding Cassegrain baffles, I will
show you what I made and how I made it for successful baffling.
An important practical consideration, that should be mentioned at this time, is that it is
an advantage to mount the primary baffle separately from the primary mirror and that a
provision be incorporated to aim the baffle. The baffle I made is made out of fiberglass,
and glued to a flange made out of PVC pipe fittings. The baffle is light, flocked on the
inside, flat black painted on the outside, mounted on the OTA back plate with push pull
screws to aim the baffle from the eyepiece location.
The first step is to draw a full scale, half axis layout to define the geometry and
dimensions of the primary baffle. There are excellent computer programs that you can
download that will work these details out for you without drawing it.
Next was to make a mandrel out of wood to form the fiberglass layup on. I used a machine
lathe because I had one, but a simpler wood lathe or fixtured router can do the same job.
Then I made a fixture to hold the mandrel and laid up the fiberglass/resin. Shown below
is the lay up on the mandrel curing. The top layer is Dacron peel ply. Dacron does not
stick to fiberglass resin and is used to absorb surplus resin from the layup. Once cure,
the peel ply is removed and the outside surface of the piece is smooth, but slightly
textured making it an ideal surface for further sanding and/or painting. Proper techniques
for fiberglass layup and protecting the mandrel are a separate subject not discussed here.
After the peel ply is removed, while the matrix is still not real hard, the ends of the
fiberglass tapered tube can be trimmed. It is easy to use masking tape and apply it to the
fiberglass while turning it on the fixture to get nice perpendicular cut lines.
The pictures show the flange construction. It can be made out of very inexpensive, easy to
work, PVC pipe fittings. Shown on the left is a PVC reducer glued to a large diameter pipe
plug. The square wrench plug has already been machined off and a nice fitting hole made for
the reducer. I used typical PVC cement. The picture on the right shows the finished machined
flange on the OTA back plate. Holes in the flange are to secure it to the back plate and the
holes in the cylinder part are there so the glue used to hold the fiberglass tube to the
flange will fill these holes and act as rivets to help hold the two pieces together. The glue
is to be JB weld, which should hold pretty well, but epoxies do not bond well to PVC.
The back plate (below right) needed to have a recess machined into it to accept the flange. The
back plate for this telescope is plywood. The recess is necessary for this telescope because
incorporating this type baffle attachment scheme is an upgrade and there was not space in the
original design for the flange thickness.
The fiberglass tube (below left) was bonded to the flange using JB Weld. A fixture was used to
make sure the flange was perpendicular to the baffle axis.
The picture below left shows the primary mirror with the shop-vac nozzle stuck into the
central hole. Also shown to the left is the new baffle. This illustrates the difference in
length between the old and the new baffle. The new baffle has not been flocked nor painted yet.
The picture on the right below shows the new baffle mounted on the back plate and painted flat
black. The inside is flocked with flocking paper, but that can not be seen too well.
For a simplified secondary baffle (below left), I made a ring out of artists “Foamie” which
fit over the secondary as shown. This is a secondary baffle which is a disk at the secondary
plane. The next upgrade will be to make a cup type baffle which will reduce the central
obstruction a little bit more.
The picture below right shows the OTA with the back plate installed containing the baffle.
Shown are the three pairs of push pull screws that are used to collimate the baffle. These
screws are hidden once the focuser is in place. My OTA incorporates removable focusers. The
rectangular brackets around the drawtube hole secure the focuser in place.
The picture below show the OTA complete with both baffles in place.
The baffle tube for this 12.5 inch Cassegrain is relatively small in diameter, therefore, the
ID was flocked. For even better light suppression, if your baffle is large enough light stops
can be installed similar to light stops used in refractor tubes. Light stops are much more
effective than flocking. For my 16 inch Cassegrain the primary baffle was larger in diameter
so I was able to make light stops out of CD’s, machined to size, and stringers. The picture
below shows the light stops on the stringers before painting flat black. The light stop and
stringer assembly simply slips as a unit into the tapered tube.
Finally, the picture below shows a view from the back plate of the 16 inch Cassegrain showing
the light stops in place. Note the focuser is not installed. This view also shows the three
pairs of push pull screws that are used to collimate the baffle.
I hope you have learned from this article that proper baffling of a Cassegrain type telescope
does indeed make a difference visually and that incorporating proper baffles is not a baffling
Thanks for reading