Latest Update:  August 23, 1992

My First Telescope by Donald Rosenfield

   The background to this essay is a message written by Tim Pyle and
addressed to me. Tim said that he wanted to make an intelligent
choice of a first telescope for himself and his family. He asked
several serious questions, which you will see answered for him as you
read the essay. One of his questions was about the comparison between
a Meade (r) 2" reflector, which I don't think is made, and a Questar
(r). Also you might want to take a look at MFT2 and 4 and my RT
essays, Rambling Thoughts, which also are stored in [U] File Folder.

   About the Questar:  At the AAVSO observing site for the total solar
eclipse a few miles from Orono, Maine in 1963, several fellow AAVSO
members had Questars. I had heard some poor reports about them but
was impressed that those AAVSO folks had them. I guess I had heard
that, about 1954, when the company started, there had been some that
didn't have perfectly matched meniscus lens/primary mirror
combinations. Questar apparently changed their manufacturing so that
by 1957 or so, they made three lens/mirror sets for each unit they
sold, throwing away two of them. The surfaces are accurate to 1/64th
wave. As the wave front must be considered--with light hitting a
mirror and getting distorted to the limit of the accuracy level, and
then being reflected with that same distortion, the wave front of a
1/64 mirror is 1/32. That isn't all to consider though, in a
catadioptric system like the Questar, Celestron (r), or Meade designs.

   For the Questar, the light travels through the meniscus lens, so
that means it  goes in and out or another 1/32, hits the mirror for
that 1/32, is reflected to an aluminized spot at the center of and at
the front of the meniscus lens, so it's into the lens, forth and back
<grin> and then out, so that's 1/16 wave and then through a hole in
the primary mirror. If you use the Questar head on, without a
diagonal, then, the system accuracy is 1/32 + 1/32 + 1/16 or 1/8 wave,
which is certainly a tasty number. If you use a camera, that's the
accuracy you have. The light that enters the lens is sent to the
focal plane precisely focusing all the entering light rays, without a
trace of achromatic or spherical aberration. Operating at f13.5,
there isn't any coma.

   Yeah, I know Tim, it won't be that good with a diagonal, and who,
nowadays wants to get a crick in his/her neck? Okay, if you use a
right angle prism, then into the prism, to and fro the reflecting
secondary surface, and then out the prism. At 1/64 per surface,
that's an additional 1/16 wave; the resultant of 2/16 (the accuracy
before entering the prism {1/8 being 2/16}) and the 1/16 of the prism
gives a system accuracy of 3/16 for visual observations. An Amici
prism, which reverses the reversed image of the Questar, giving an
image correct up and down as well as left and right is available, but
only with the cavil of lesser sharpness--which is the opposite of what
we are striving for.

   You might compare those results with the Schmidt-Cassegrains
offered through astronomy magazines and local stores. The advertise a
1/10 wave surface accuracy. Okay, let's analyze that. Into and out
the thin lens corrector plate, (1/5 wave) to and fro the primary
mirror, (1/5) to and fro the secondary mirror, (1/5) giving 3/5 wave
for photography. Add the right angle prism (2/5) and you get a
resultant system accuracy of 1 wave. If instead of the prism, you use
a mirror diagonal, the system accuracy is a little better, at 4/5 wave
. . . but not much better.

   If you want to use your instrument visually, a 1/4 wave is recom-
mended. If you want to photograph, a 1/8 wave is recommended. Those
recommendations were made before the development of the Schmidt
-Cassegrain; they referred to the system accuracy of a Newtonian. I
never have felt that those levels of accuracy were sufficient for any
telescope. A 1/4 wave mirror has a 1/2 wave wave front. Two such
mirrors have a resultant system error of 1 wave. And that is what the
books recommend? No, I can't accept that. Not enough rays of focused
light come to a point in a 1/4 wave figured mirror. In this writing
and in my book, I suggest that 1/4 wave SYSTEM ACCURACY is what we
should strive for in a visual telescope. 1/8 wave system accuracy
will afford superb photographs.

   It's a lot easier to find commercially made 1/20 wave surfaces in
Newtonians than to find that in one of the Schmidt designs. A
Newtonian made to 1/20 surface accuracy having two mirrors . . .
Lessee now, each mirror's wave front is 1/10, so the system accuracy
is 1/5. You don't use a star diagonal on a Newtonian, so either for
photography or visual work, the accuracy is that 1/5.

   Another problem with the Schmidt-Cassegrain design is that the
comparatively heavy secondary mirror cell hangs from the center of the
thin corrector plate. Especially in the larger sizes, that causes
flexure when you direct your telescope to another part of the sky.

   In a discussion with the optical technician at Celestron, he said I
was absolutely right. There's no getting around that problem with the
Schmidt-Cassegrain (SC) design. The larger the telescope diameter,
the thinner is the correcting plate and the heavier is the secondary
mounting and thus the worse is the problem.

   Another possible problem with commercial telescope makers is that
of the testing devices they use. Meade and Celestron, in their
catadioptric SC design use a spherical primary, I'm not sure of the
figure on the secondary mirror, but I'd guess it were spherical, and a
thin correcting plate that corrects the primary's sphere to bring all
entering light rays to the same focal point.

   The problem is to figure that correcting plate. Celestron's
patented solution (I don't know if their patent ran out, if Meade pays
a royalty, or if they do this differently) is to have a properly
shaped master die. They put the disc of glass that will be the
correcting plate against the die and remove the air between them.
That causes the required flexure of the plate, which is then ground
and polished on its reverse side--that is, on the side not facing the
die. When the vacuum is removed, the corrector unwinds from the plate
and Voila! It has the necessary curvature. That was an impressive
invention, making the SC into a design that could be repeated in
production. I have a little problem with that procedure, though. I
got the impression that the flat surface of the correcting plate isn't
figured. Any deviation from the optimum surface has to be figured
into mathematics, above. I was assuming that the rear surface of the
correcting plate was 1/10 wave surface accuracy.

   From those manufacturers, as well as many others around the world,
other telescope designs are available. The following discussion is
concerned with telescopes made using a Foucault Tester (FT) by
amateurs and professionals. So, for example, if you wanted to buy a
used telescope, you should be familiar with this argument.

   This applies only to a FT, where you cut a beam of light returning
from a mirror after reflection from a pinhole light source. Usually a
razor blade or something similar is used for the knife edge with which
you cut that beam. At least one book on telescope making, Berry's
_Build Your Own Telescope_, has plans for building a FT with a fixed
knife edge and a moving pinhole light source. Good Grief! If you use
the kind of FT with a permanent light source and a moving knife edge,
with which you measure the activity while figuring your mirror, you
can use the formula "r squared divided by R," where r is the radius of
your mirror and R is twice the focal length. (R would equal the focal
length if you used a parallel light source. The light from a pinhole
is not parallel.)

   If you use Berry's method, you must use "r squared divided by 2R,"
which is half as accurate. So if a manufacturer says a given mirror
is accurate to 1/20 wave and on questioning (he'll NEVER say this in
an ad) you find he's using the Berry method, the real level of
accuracy you'd find using the moving knife edge tester, is 1/10 wave.
Caveat emptor. On top of that, with either kind of tester, you have to
factor in the ability of the maker to properly read the results . . .
and then the moral 'suasion of that maker. Does he, as in a recent
description in one or the other of _Sky & Telescope_ or _Astronomy_,
send out a mirror with a turned down edge? Sure, why not? How many
mail order buyers of mirrors have their own testing devices? And if
they don't, how would they know that their mirror isn't one of high
quality?

   Tim, this message is getting long winded. I hope you don't mind.
One reason to answer you painstakingly is that these messages will be
stored in the file area for astronomy and then, when the next person
asks a question like you did, I won't have to repeat all this.

   I figurer my own mirror, a 10"f4 to a surface accuracy of 1/38
wave. The FT I was using could be read to half a ten thousandth of an
inch. That's another thing to ask the person who figured that mirror
for sale. What kind of micrometer pushes the knife edge back and
forth? Obviously, if the micrometer used reads to 1/1000" and thus can
measure to 1/20 wave, it couldn't read as accurately as the one I
used, which could measure to 1/40 wave. I stopped correcting the
mirror at 1/38 wave to leave a bit for temperature correction as the
night air cools. (Some mirror makers use a FT that is hand divided to
read to 1/100 inch.)
z
   Lessee now. The wave front on my mirror is 1/19 wave and the
diagonal has a surface accuracy of 1/20 so a wave front of 1/10. So
the worst system accuracy there is about 3/20 wave, which isn't
shabby. One night at the OCA's Anza observing site, Minor White had
his telescope, a 10" Celestron, I think, set up a few feet from my
10". He would look at deep sky objects in his and then in mine. He
kept saying, over and over, that he couldn't get over the crispness of
my images in comparison with his.

   Once I had the 10" at an observing party in Ohio. Except for my
soon-to-be-wife and a friend from the Chicago Astronomical Society,
everybody else there was from the general public. We couldn't stop
one couple, who were so impressed with the naked-eye Milky Way, from
taking flash pictures of it.

   I took my Questar to Brazil for our 3-year stint there. I used the
8" Clark refractor for my variable star observing and the Questar for
star parties. On one trip my wife and I made, 1500 miles south of Sao
Paulo where we lived, I ran a star party in a small city, Rio Grande
de Sul. One of the people who took a look was a minister. He talked me
into coming back about two months later and giving a star party and
talk about the Star of Bethlehem to his congregation, in Portuguese,
of course.

   Note that I don't have the standard Questar. I have the Field
Model, which costs considerably less. Mostly I use it on a Davis &
Sanford heavy duty tripod. Total weight of tripod and Questar is
eleven pounds. A major advantage of the Questar is that you can use it
as a long distance microscope. Most telescopes can't focus closer
than 100 feet. Some can focus at 30 feet. The Questar can focus at 10
feet. A cousin of mine who lives on Cape Cod wrote me a long letter
asking what kind of telescope to get. His major use was for looking
at birds. He wanted my opinion of a small Celestron. I pointed out
the advantages of a field model Questar, which he bought. I got rave
notices on that. He says than on bird watching trips, there's always
a long line of fellow bird watchers waiting to use his Questar.

   Tim, one reason to cover so much about the accuracy when making a
mirror or a mirror telescope is not so that a person would go ahead
and make one but to make that person aware of which questions to ask
when buying one. Of course, if you want to make one, I refer you to
the OCA mirror making interest group. I go into all this much more
exhaustively in my astronomy book. I've had two amateur telescope
makers, each of whom put 8" diameter mirrors into operation years
ago--when we all were younger--with turned down edges, knowing they
shouldn't, but not having anyone around to help them, read my appendix
on mirror making. Both readers said that for the first time they
realized what had been wrong with their procedures and how to correct
those turned down edges.

   Back to your questions, Tim. You say you want a family telescope
and aren't sure you will become an active amateur astronomer. Okay,
don't get a telescope that is so heavy that it spends most of your
life sitting in your closet. I have friends with eight inch SCs that
haven't been used in years. My 10 inch weighs 150 pounds; my Questar
weighs 11 pounds with that tripod. Guess which one gets set up more
often! Though in my design of the mounting of the 10", no piece
weighs more than 35 pounds.

   You said you read _Astronomy_ magazine. Let's start skimming from
the back. (I'm using the May 1991 issue.)   Pages 102-3 has the Parks
ad. You should consider the 4-1/2" f5 at $399. But not without
finding out what accuracy it has and what flavor of FT was used to
figure it. You should ask them to send you an analysis of the figure
before you send money for their telescope. Another thing you should
do, with all mail order telescopes, or those back ordered by a store
is be prepared to wait. That telescope you order in November for an
Xmas present still hasn't arrived the following October. Telescopes
take their own time to finish. My 10" took me 2 years. During those
two years I was in post-graduate school and working full time. I
managed to sleep 3 hours a night.

   Forward a few pages is the focus camera ad. The Meade company's
3.1" refractor (lens telescope) at $437 may be good. But like all
achromatic refractors, it probably has color fringes around planets
and stars. That's one reason to use a reflector, to get away from
chromatic aberration. An apochromat has 3 lenses instead of the 2 of
an achromat, but for the same quality can cost 2 to 3 times what an
achromat does. Even apochromats don't get rid of all the color
fringes. And, of course, you get to add the in and out distortion of
that third objective lens....

   The Obsession Telescopes ad on page 97 shows a light bucket
Dobsonian reflector. Several OCA members have Dobsonians. John
Dobson, a former monk is a "sidewalk astronomer" in the San Francisco
Sidewalk Astronomers. Dobson took the Couder reflector mount and
somehow attached his own name to it. There are two problems with the
"light buckets." The first is that most of them move on a Teflon(R)
bearing surface. Teflon is a wonderful product. It's so smooth you can
spend your time moving the bearing back and forth. It's like listening
to a sonata or swimming in fresh water. There's only one problem with
it:  It's no good for a telescope bearing.

   "Boy, what a letdown," you say. "What's wrong with it."

   Well, after you start moving the telescope to another part of the
sky, it's motion is smooth as silk; it's like going down the water
slide at an amusement park; it's like rubbing your child's smooth
cheek.... The problem is in getting it going. This is another thing I
get into in my book. When you start to move that bearing it let's you
know right then who's the boss. IT doesn't want to go where YOU want
it to go. Here's a direct quote from my book, _Astronomy For The Love
Of It_, (All Rights Reserved.):

     Mike Bennett, a combination astronomy teacher and engineer-
     ing consultant, commented about the Teflon(R) bearings used
     in the Dobsonian design. He said that the transfer of
     mechanical energy involved in moving the Dobsonian's tube
     from one position to another causes the bearing to start to
     move jerkily, instead of smoothly. Teflon moves smoothly
     enough once started, but starting is a problem.

     I was using a friend's 17" Dobsonian at an observing party
     of the Orange County Astronomers. You couldn't follow
     anything in the sky, because when you tried to move the
     mounting in the direction needed to follow the object, the
     jerkiness threw it out of the field of view. Many Dobson-
     ian users, across the country, frustrated with that jerky
     motion, remove the Teflon bearing. They replace it with
     used roller skate wheels. Used wheels are smoother than
     new wheels.

   Another problem with the Dobsonians is their accuracy. If you look
at the ads for them, you see they advertise 1/4 wave. "Ok," you say,
"I can go along with that . . . but wait a minute. The wave front
will be 1/2 wave and with two mirrors the system accuracy is 1 wave.
That's not a good system either."  I'm proud of you. You're learning.
Another company sells an 18"f5.5 light bucket Dobsonian. The optics,
which are from Galaxy(r) are advertised as having a 1/10 wave
accuracy. What would you question if you were interested in this
telescope?

   "Ok, Don . . . hmm . . .1) What are the telescope's bearings made
of? 2) Is that 1/10 wave the surface accuracy? [By now I know enough
to not ask if it is system accuracy, a question which would only be
answered by a Yes!]  3)  What kind of Foucault Tester was used in the
figuring? That is, is the knife edge or the light source movable? And
to what part of an inch does it measure . . . 1/100", 1/1000"? Of
course, if the mirrors each are 1/10 surface accuracy, giving a wave
front of 1/5 or a system accuracy of 2/5, that isn't spectacular, but
at least it is better than the light buckets with 1/4 wave surface
accuracy. Even that's not a fair comparison, as we still don't know
if they measure with a moving or fixed knife edge or to what accuracy
that tester can achieve."

   Another company advertises a 14.5"f4.5 instrument that looks at
first glance like a Dobsonian alt-azimuth, but in reality is an
equatorial with a clock drive. Its optics are advertised as hand-
figured to 1/20 wave. This time I want you to tell me what are the
possible system accuracies of that telescope.

   "Well, the simplest answer is two wave front mirrors at 1/10 wave
so the system accuracy is 1/5 wave. While a far cry from the
advertised 1/20, it is better than the others we've been discussing.
If the Foucault Tester only could read to 1/100" we couldn't rely on
that 1/20 wave. If the knife edge was fixed, but they used a 1/1000"
micrometer to measure the primary's surface then we would have to say
that compared with a similar micrometer and a moving knife edge, the
system accuracy was 2/5 wave, even though it is legal for the maker to
(effectively) advertise a 1/5 wave mirror, which by the way THE MAKER
THINKS IS A 1/20 wave system. Of course, we don't use the Foucault
Tester on the flat diagonal mirror. We might ask the company to what
part of a wave flatness is their secondary test flat. I never heard
of the things we're discussing here. I suspect that the commercial
telescope makers also are unfamiliar with these parameters."

   Great. You have that down pat. By now you could write this essay.
One point I didn't mention is that a manufacturer or reseller might
advertise that a particular mirror's optics are good to +/- 1/8 wave.
Go figure. I mean it, Tim. What is the system accuracy of a Newtonian
reflector with that surface accuracy?

   "Let's start with the +/- 1/8 wave, which, of course, is 1/4 wave.
Wave front on each of the mirrors is 1/2 wave so the system accuracy
is 1 wave. You are right; I have this down pat now, Don."

   The telescope I learned on was a 3"f11.5 Skyscope with no finder on
it. People at the OCA star parties were bemused that my 10" has no
finder. I learned with that little 3" scope how to find things you
can't see by aiming along the telescope where they should be. I
think, though, that the Telrad(R) finder is a good idea. It projects
a red spot onto the sky without magnification. Many finders are
barely useful, with too small a field of view, and upside down to
boot. Those makers who supply a right angle finder with their
telescopes should be sentenced to use them for a night.

   Tim, I'm not sure Meade makes a 2" reflector as you assert in your
message 608. That's a 50 mm mirror. I know they make a 60 mm lens
telescope (refractor). For a specified suggestion (which, after all
is what you asked me for) I think you would be happy with either a new
or used Questar. The Field model would be adequate. If, later you
want an equatorial driven mounting, Questar sells a Duplex Mounting
for the Field Model. If you get a new one, you should specify the
Zerodur (r) ceramic mirror and the multi-coated lens surfaces. An
Amici prism is available as I noted above, but I don't recommend it.
By the way, you can of course, start with a Duplex Mounting and take
the Questar out of that mounting when you want to use it terrestrially.

   I would be remiss if I left you with the idea that the 1/4 wave
that J. Allyn Thompson and others suggest for a minimum surface
accuracy had no bearing  on modern telescope production techniques. I
don't remember seeing my argument as stated above in any book on the
subject. I wrote this in my manuscript in 1984. Now I am writing
this in late April 1991. I know that there is a committee of telescope
manufacturers  trying to come up with an agreement on whether wave
front accuracy is what should be advertised. I think it should go
farther than that:  What should be advertised is system accuracy. It
won't happen though, as the SC manufacturers will find themselves out
in left field in comparison with Newtonian manufacturers.

   Taking the example of an advertised 1/4 wave for a Newtonian, the
system accuracy would be 1 wave. Nobody--other than some authors--
suggest that is adequate. You might make such a telescope; certainly
many amateurs have done so. When you get disgusted with it--in
comparison with well-made optics you can put it up for sale and you
get or make a better mirror. I better make a point here:  Any
telescope is better than no telescope. The difference between no
telescope and that little 3" $29.50 Skyscope (I kid you not; that was
its price in 1951.) is amazing. As I said I used it for a semester and
was happy with it. It showed rings of Saturn, 4 satellites of Jupiter
and cloud belts on Jupiter, galaxies, star clusters, nebulae . . . all
things you would have trouble seeing naked-eye. (Unless I'm talking to
Clark Kent.)  You see, it is a difference in kind, going from no
telescope to any telescope. In fact, going from naked-eye to a good
pair of prism binoculars is a revelation on a clear moonless night
away from the cities. All other size telescopes are a difference in
quantity.

   Another option if what you have is a 1/4 wave mirror is to get a
1/20 wave diagonal. That way, you have 1/2 wave wave front on the
primary and 1/10 wave front on the diagonal or a system accuracy of
6/10 (3/5). That's a LOT better than 5/5.

   At the OCA star parties I attended, I was the only one with a
Questar. People I questioned were unaware of any member that had one.
The reason I got a Field Model in the first place is that I wanted to
use it as a telephoto lens for terrestrial subjects in Brazil and on
the drive home from Brazil. (We flew to Sao Paulo. Our Land Rover
Camper didn't arrive for a year; like telescopes, fine vehicles take a
long time to make.)

   Also, I had a J.W.Fecker 4"f8 Newtonian, with a clock driven fork
mounting. The Questar was so far and away better than that 1/8 wave
advertised scope that I dismounted the Fecker tube and made a
convertible telescope. That is, when I want to use my Questar for
looking at birds or at spiders in their webs, or whatever, I take it
out of the fork and use the Davis and Sanford tripod. For celestial
objects, I use the fork.

   Having just reread your original message 608 (which was lost along
with everything else in that message base) I turn to the Edmund
Scientific (tm) catalog. Another option to consider is a Rich-Field
Telescope (RFT). My 3"f5, my 5"f5, and my 10"f4 are RFTs. Edmund has a
4-1/4"f4.2 RFT that they call an Astroscan(R). The catalog gives the
surface accuracy as 1/8 wave. However, they have a thin window "to
keep the dust out." So the system accuracy is what? Surely you can
figure that by now.

   "Sure teach, that's easy. You take the wave fronts of the primary
(1/4), the diagonal (1/4), and the front optical window that carries
the diagonal (1/4); the system accuracy therefore is 3/4 wave. Now I
see why you figured your 10" to a surface accuracy of 1/38 wave."

   And why the Hubble Space Telescope is figured to 1/65 wave
accuracy. The problem there was accurate to what standard? They
thought they were doing it to 1/65, but the primary surface turned out
to be accurate to 1/2 wave, giving 1 wave wave front before you even
get to the secondary. The May 1991 _Sky & Telescope_ has the Chicago
Optical ad on page 562. They advertise RFTs, but not their size,
price, or accuracy. You would have to get their catalog.

   "Don, what's a rich-field telescope? I think you are assuming I
know what you are talking about. . . .

   Sorry Tim. A rich-field telescope is a short focal length
reflector (mirror telescope) or refractor (lens telescope). The short
focal length telescopes give wider fields of view than the same
diameter telescope but with a longer focal length. Most telescopes
have a field of view smaller than the half degree diameter of the
Moon. My 10" has a 2 degree field. My 3"f5 refractor is a richest-
rich-field (RRFT) telescope; with a seven degree field of view, you
can see the whole constellation of Lyra in the field at the same time.
The RRFT is the design of a telescope which will show more stars in
its field at the same time than ANY OTHER TELESCOPE, REGARDLESS OF ITS
SIZE!!!

   But, you ask, "How can that be done with a 3" telescope which has
to compete with say a 17"f4 light bucket, or for that matter, the 200"
f3.3?"

   There is an anomaly in the magnitude system. That is caused by our
solar system's position on the inside of an outer spiral arm of our
Milky Way galaxy. There are 2.512 times (Sounds like a log number,
doesn't it?) more 6th magnitude stars than 5th, 7th than 6th, 8th than
7th, and so on, until you hit 11th magnitude. Then there is a bit
less than 2.512 times more 12th than 11, 13th than 12th, and so on.

   "I still don't see why that would affect the RRFT."

   Ok, look at it this way. The 3"f5 with a 7 degree field sees just
to 11th magnitude. That includes mostly stars inside our spiral arm.
To see the 12th magnitude and fainter, which are mostly galactic
central and halo stars (Arcturus is a halo star.) you need a larger
telescope as you have to jump across the largely empty space between
the arm we're in and the next one in, and then the central bulge of
the galaxy. The larger telescope has a smaller field of view.

   Granted, sometimes a larger telescope is needed to be "the" RRFT on
a given night--in case the Moon is visible or city lights brighten the
neighborhood. The more background sky fog, the larger is the RRFT on
that night.

   Rereading your message (Again!) I note that while you say your time
to do astronomy is limited, you don't say that about your children.
Children are always welcomed (Well, nicely behaved children.) at star
parties. We were all kids once. Who showed us the wonders of the
heavens? I wasn't exposed to astronomy until a course I took in
post-graduate school. The astronomy professor, Morris Goran (This was
at Roosevelt University in Chicago.) loaned me that 3" Skyscope for
the semester and loaned me a copy of _Making Your Telescope_, by
Thompson. I read that book seven times through and jumped into this
marvelous hobby. There were about 25 students in that class. Dr.
Goran announced to the class that whatever I got on the final, an A
would be 10 points less. I got an 85 (It was a hard test.), so an A
would be 75. No one else in that class got above 50. I got some hard
looks from my classmates.

   What I mean to say, Tim, is that your children could use your tele-
scope to work on various observing goals--all useful as extra credit
in science classes--such as drawing or photographing the craters and
mountains on our neighbor, the planet Moon; variable star observing;
keeping track of Jupiter's satellites; comet searching; etc. Help is
available in any of astronomy's myriad fields.

   One question, Tim:  Did I answer your question?

   Donald Rosenfield, Asst. General Sysop/Cosmogonist
