THE YOLO REFLECTOR

Arthur S. Leonard
-Figure 1-

The Yolo Reflector is an unobstructed reflecting telescope consisting of two concave

mirrors, both tilted relative to the approaching light path, and having the general
configuration shown in Figure 1. Many variations on this basis design are possible,
depending on aperture, focal length, and optical excellence desired.

The design of the Yolo optical systems are based on the following general principles:

• 1. When a mirror or lens in an otherwise perfect optical system is tilted, a

wave-front aberration, or error, is introduced into each ray traversing the system.
The magnitude of this error, in terms of path-length difference, w, may be
expressed as follows:(1)*
 • where the C's are constants whose values depend on the amount of convergance
or divergance of the cone of rays striking the surface, the index of refraction (if a
lens surface is involved), and the radius of curvature of the surface; is the angle
of tilt of the optical axis of the surface relative to the approaching central ray; and
z and designate the point (in polar coordinates) where the ray in question
strikes the optical surface.

____________
*
References are listed at the end of the paper.

-2-

• 2. Because of the cyclic nature of the individual terms (proportional to ,

, etc.), it is possible, by tilting N elements (mirrors or lenses having
curved surfaces), and each tilted through the correct angle and having its tilt-axis
oriented in the proper direction, to make N-1 of the terms cancel out completely.

• 3. By applying a properly designed warping harness to one of the mirrors of the

system, acceptable control or cancellation of the residual astigmatism (
terms) can be affected.

• 4. If the radii of curvature of the two mirrors are made great enough in
comparison with the aperture of the instrument, only the first two terms in the
infinite series indicated by Eq. (1) will be large enough to require cancellation.

The first term in the series (the term) is coma and the second, astigmatism. In
the Yolo designs the tilt angles are adjusted to make the coma term of the secondary
mirror just cancel the coma of the primary, and the astigmatism terms from the two
mirrors are taken care of by some additional stratagem. If a third tilted element, such as a
lens of long focal length and which itself may introduce a small coma term, is employed,
the tilts of the two mirrors may be altered slightly so as to cancel this small additional
amount of coma.

With small apertures and relatively long radii of curvature, the residual
uncompensated tilt-aberrations (mainly astigmatism, in this case) may be so small as to
be tolerable. In this case no additional optical or mechanical device is necessary. With
slightly larger apertures and higher standards of optical excellence, the addition of a small
cylindrical lens or a tilted spherical lens of relatively long focal length, located in the
light path a
-3-

short distance ahead of the focal plane, gives a very satisfactory instrument. Another
variation employs a tilted simple lens of very long focal length located in the light path
ahead of the primary mirror. Where reasonably good optical performance is required, all
of these designs are limited to rather small apertures and rather long focal lengths.

For instruments of large apertures and top optical performance, one or the other of two
additional stratagems must be employed. Either the surface curves of one or both mirrors
must have predetermined amounts of astigmatism figured into them, or a properly
designed and adjusted warping harness must be applied to one of the mirrors.

Although the equations and comments contained herein may be applied to more than
one of the variations just described, the bulk of the material of this paper is concerned
with the harness-warped-secondary Yolo reflector.

General Equations
The effective focal length, F, is given as follows (see Figure 1):

where R1 and R2 are the radii of curvature of the primary and secondary mirrors,
respectively, and S is separation distance (slant-distance) from the center of the primary
to the center of the secondary.

Equation (3) gives the diameter of the illuminated area of the secondary, D2:

where D1 is the illuminated aperture of the primary.

-4-

The slant distance, T, from the center of the secondary to focal point of the system is
given by Eq. (4):

Finally, the relationship between the angles of tilt of the two mirrors is given by Eq. (5):

where A and B are the angles of deflection of the central ray by the primary and
secondary mirrors respectively (see Figure 1).

Experience with the design of the light-baffle system for the Yolo reflector has shown
that by designing the optical system to bring the focal point well back of the plane of the
primary mirror, the "muffle" (that part of the light-baffle system which is located forward
of the secondary mirror) can be shortened very appreciably. It is recommended, therefore
that T be made about 1.2 times S. When this condition is combined with Eq. (4), the
following is obtained:

Although R1 and R2 need not be made equal, in most practical designs they will turn
out to be not greatly different; and there is no good reason why

-5-

they should not be made the same. Furthermore, when the warping harness is placed on
the secondary mirror (see "The Mirror Warping Problem," ahead) other conditions dictate
that the secondary mirror blank should be nearly the same size as the primary. If we make
the two mirrors the same diameter and the same radius of curvature, then one grinding
tool and one pitch lap will suffice for both. This represents an appreciable saving in time
and expense in constructing the telescope and is, therefore, recommended. If R1 and R2
are replaced by the single quantity, R, and T is made equal to 1.28S, Eqs. (2) through
(6) reduce to the following:
Parabolization of the Primary Mirror
Theoretically, the primary mirror should be a paraboloid, and the secondary, an
hyperboloid. Both mirrors have such long radii of curvature, however, that neither one
deviates very much from a perfect sphere. In Yolo reflectors of small aperture and
moderately long focal length, this deviation is practically undetectable, and instruments
employing spherical mirrors will show no need for correction. However, instruments of
eight-inches aperture or larger, made with spherical mirrors, will be found to be under-
corrected by an appreciable amount. This defect can be eliminated by a small amount of
parabolization applied to the primary mirror only.

-6-

The amount of parabolization, Y, which when applied to the primary mirror, will
correct for both mirrors, is given by the following expression:

When R1 is equal to R2 and T is 1.2 times S, the following is true:

This means that the primary mirror should be parabolized by 416 percent. Although
this might sound like a lot of parabolization, in actual amount it is quite small and will be
difficult to measure. To test for this amount of parabolization, we recommend either a
caustic tester or a null test such as that proposed by Dall(2). Another alternative is to have
the secondary mirror aluminized and then mount both mirrors in the telescope tube and
test on a bright star with a high-power eyepiece. This method will require good seeing
and may be much less convenient than a shop test, but it is very effective in obtaining the
right amount of overall correction for the system.

Design Procedure
Although the design procedure to be described here may seem somewhat unorthodox,
it has the effect of forgiving the telescope maker for some of his mistakes in meeting
tolerances or allowing him to just about ignore them. The first step, after the aperture has
been decided upon, is to select a good value for the effective focal length, F. The few
Yolo reflectors which have been

-7-

built (or started) to date have been designed as f/15 and f/17 instruments. Throughout this
paper reference will be made to two of these, an 8-inch f/15 design described very briefly
in Sky and Telescope(3), and a 12-inch f/17 design currently under construction for the
Sacramento Valley Astronomical Society Observatory at Colfax, California. Although
there seems to be no inherent reason why fairly short focal lengths should not be
satisfactory, the prospective telescope maker is cautioned to not try anything much
shorter than about f/12 or f/10 until some Yolo designs of these focal ratios have been
built and thoroughly tested.

With values for D, and F decided upon, values for R, D2, S, and T are calculated,
using Eqs. (7), (8), (9), and (10). Next, a scale drawing is made showing the two mirrors
and the light-path, much like Figure 1. The secondary mirror blank is located so that it
clears the parallel bundle of rays going to the primary by about one-twentieth of its
diameter. This, then, permits angle, A, to be measured or calculated. With this known
value for A, a value for angle, B, is calculated, using Eq. (11). This permits the focal
point to be located and the outline of the light-path to be drawn to scale. Following this,
the system of light-baffles is designed (see "Light-Baffle System," ahead) and drawn to
scale.

At this point we are ready to start work on the construction of the telescope itself. The
first step here is to rough-grind both blanks to the desired radius of curvature, thin-down
the secondary blank (if this has been decided upon) and smooth up the backs of both
blanks until they are reasonably flat and parallel to the front surfaces. Next, the mirror
cells and warping
-8-

harness are constructed and fitted to the glass blanks. After this has been completed, the
mirror blanks are fine-ground and polished.

After the mirrors have been tested and found to have the desired surface curves, the
warping harness is tested with the secondary mirror in the test set-up shown in Figure 4.
After the warping harness has passed its test, the radii of curvature of the two mirrors are
measured carefully, and a second design is made, based on the measured values of the
radii of curvature of the two mirrors as they have actually turned out, Eq. (6). Usually, the
new value of mirror-separation, S, will turn out to be not greatly different from the first
calculated value, and the plans as first laid out can be altered fairly easily to conform to
the new values.

The telescope tube and baffle system is constructed next, based on this second design.
Finally, after the tube has been built, the actual separation distance, S, is measured, and
the angle B, is calculated from measurements of the primary mirror relative to the line
running from the center of the eyepiece tube to the center of the secondary mirror. With
these final values for S and B, a third design is calculated (mainly the angle, A) using
Eq. (5). Final collimation of the instrument is carried out according to this final design.

The Mirror-Warping Problem

The purpose of the warping harness is to produce, by mechanical flexure, a
displacement of the mirror surface which is everywhere equal and opposite (to within
acceptable optical tolerances) to the wave-front aberration, astigmatism. The aberration,
astigmatism, is given by the following equation:
-9-

This equation tells us that the amount of flexure should be proportional to the square of
the distance from the center of the mirror and in going around the edge of the mirror (say
clockwise) the displacement should take the shape of a sine-curve and it should go
through two complete cycles (two maxima and two minima) in going once around the
mirror.

If we were to balance a mirror blank, face-up, on a small rigid support located under
the center of the mirror and then hang two equal weights from the edge of the blank at
opposite sides, the glass would bend under the load as a simple beam. The following
formula gives the flexure, h, for a beam so supported and loaded:

where Kh is a quantity which is determined by the force applied and other factors, Z is
the radial distance from the center to the point in question and r is one half the diameter
of the mirror blank. This equation tells us that in the central part of the mirror the flexure
will be proportional to Z2, but out toward the edge it will flatten out a little compared
with this curve. This suggests that as far as the Z2 feature of the astigmatism formula is
concerned, our flexed mirror will follow the desired curve fairly well in its central part,
but out toward the edge it may deviate a little from the desired curve.

If we should support the mirror blank at two points at opposite sides at its outer edge
instead of at its center, and then hang our two equal weights at points midway between
the two points of support (90° each way from the
-10-

support points) the edge of the mirror will be flexed upward over the two support points,
and downward under the other two points, midway between. In going once around the
edge of a mirror so loaded we will encounter two high areas (maxima) and two low areas
(minima). Although the curve may not be a perfect sine-curve, it will be a fair
approximation to such a curve.

The first warping harness to be constructed was designed to apply flexure forces to the
edge of the mirror blank in just this way - two "pushes" and two "pulls". Tests of this
device in the optical shop showed it to do a fairly good job on a 6-inch-diameter mirror
everywhere except in the immediate vicinity of the the pressure-points, where the flexure
was found to be too great. This is just the opposite of what might be expected from a
consideration of just the Z2 and Z3 features of the flexure curve. It indicates that other
factors are acting to increase the curvature of the mirror blank in the areas where the
simple beam formula says the curvature might be found to be a little too small.

Another way of looking at the problem is to consider the actual flexure curve as we go
around the edge of the mirror (in contrast to going across the face of the mirror). From
this point of view the actual flexure curve peaks too sharply at the maxima and minima
(as compared with the desired sine-curve shape). A rather obvious solution to this
difficulty is to distribute the loads at the edge of the mirror as compared with the single-
point design. To carry out this idea, a short bar was placed under each of the four pressure
members and each of these divided the load equally and distributed it to two points
separated 30° apart around the edge of the mirror.

The original warping harness was modified to employ this stratagem and then tested
on a 6-inch-diameter mirror. The final design of this warping
-11-

harness is shown in Figure 2. As far as could be determined by sensitive tests in the

optical shop, this warping harness was completely satisfactory in every respect. Although
the warping harness for the SVAS telescope shown in Figures 3a and 3b is more
sophisticated in its entirety, it employs the same basic pattern of load distribution.

A second stratagem, which can be used to reduce the excess deflection of the mirror
surface in the immediate vicinity of the pressure points, is to reduce the thickness of the
mirror blank. The total deflection of the mirror surface at any given point is largely due to
bending of the mirror blank; but in addition, and to a lesser extent, is due to shear-action.
The deflection due to shear-action varies directly as the applied load, inversely as the
thickness of the mirror, and builds up rather rapidly as we approach the pressure point.
The deflection due to bending, on the other hand, varies inversely as the square of the
thickness. A moderate reduction in thickness will, therefore, result in a rather large
reduction in the force required to bend the mirror; and this will result in reduction of the
shear-action component of the total deflection at each point on the mirror surface.

Even if we were to make the mirror blank wafer-thin, the deflection of the mirror
surface would not fit the desired curve perfectly, so we should not try to carry this idea to
an extreme. If we should make the mirror blank too thin, the optician would have
difficulty in controlling the surface curve of the mirror to within the desired tolerances. In
order to give some idea as to what is recommended in this respect - the 12.5-inch-
diameter pyrex blank for the secondary of the SVAS 12-inch Yolo was thinned down
from the standard 2.125-inch thickness to 1.75 inches.
-12-

The third stratagem which can be employed to eliminate localized excess deflections
produced by the warping harness is to use what might otherwise considered an oversize
glass blank and thus move the pressure-points on the glass well outside of the area of the
mirror which is used to reflect the cone of rays of light. This stratagem has the further
advantage of relieving the optician of some of his working tolerances and making an
optical surface with a fairly bad turned edge perfectly acceptable - as long as the turned
part of the edge does not extend too far in from the edge of the blank.

This device has the disadvantages that it increases the cost of the glass blank and it
makes necessary larger angles of tilt in the design. Although larger tilt-angles result in
larger tilt-aberrations (which are compensated for in the Yolo design) we should charge
only a small fraction of this disadvantage against this idea since the larger tilt-angles
reduce very appreciably the problem of baffling out the unwanted stray light which might
get into the eyepiece. Although the warping harness might be applied successfully to
either the primary or the secondary mirror, the increase in the cost of the blank resulting
from the employment of this stratagem will be much less if secondary is warped rather
than the primary. Furthermore, if the amount of this increase in blank diameter is chosen
properly, the diameter of the blank for the secondary will be equal to that of the primary.
When this is done, one grinding tool and one polishing lap will suffice for both mirrors;
and the economics from this may more than offset the added cost of the increased
diameter of the secondary blank. It is for this reason that we recommend that the warping
harness be applied to the secondary rather than to the primary mirror.
-13-

Design of the Warping Harness

And Mirror-Mount

The principal structural part (or parts) of the warping harness consists of either a pair
of crossed beams, as shown in Figure 2, or an x-member, as shown in Figure 3a. It should
be capable of providing four equal forces - two pushes and two pulls - equally distributed
(90° apart) around the edge of the mirror. The four forces will be equal if, in the crossed
beam design, the two beams are pivoted at their midpoints or, in the x-member design,
opposite arms are of equal lengths. Preferably, all four arms should be of equal length.
Tolerances for lengths of the arms have not been calculated, but it is felt that if reasonable
care is exercised in laying out the lengths of the four arms, this specification will have
been met satisfactorily.

The force required to warp a glass mirror has not yet been measured or calculated. As
a result we can only give a rough value for the strength requirements of the warping
harness. The force (at each of the four points on the edge of the mirror) required to bend
the mirror is given by the following formula:

where P is the required force, K is a constant which depends primarily on the modulus
of elasticity of the glass, t is the thickness of the glass blank, D2 is the calculated
illuminated diameter (not the diameter of the glass blank), and wa is the amount of
astigmatism which must be produced
Fig 3A

-14-

by the warping harness. This is given by the following equation:

where U is the calculated separation between the artificial star and eyepiece to be used in
the test set-up for checking the performance of the warping harness in the optical shop,
and is given by Eq. (19).

The weight of the warping harness shown in Figures 3a and 3b is carried by a thin
sheet-metal spider and a set of counterweights. With this design the actual weight of the
warping harness is unimportant so far as flexure is concerned. With the design shown in
Figure 2, however, the total weight of the harness is carried by the mirror. This will
produce unwanted deflections, or errors, in the surface curve of the mirror. In order to
keep these errors to a minimum, the warping harness should be designed as light as
possible. Lightness could be achieved by constructing the harness out of high-strength
aluminum alloy; but it is felt that the high coefficient of thermal expansion of aluminum
alloys might make this material unsatisfactory for this application. For this reason steel is
preferred as the material for the warping harness. In order to avoid excessive weight with
steel, some guide lines are needed.

The warping harness shown in Figure 3a was made of annealed mild steel. Assuming a
yield strength of 40,000 lb per square inch and dimensions as measured, it should have
been capable of sustaining a load of about 250 lb at the ends of each of its four arms.
Substituting this value for P and putting the known values for t, D2, and wa for this
unit into Eq. (14), a value of K = 14.5x106 lb/in3 is obtained. This warping harness was
tested

Figure 3B
Figure 2

The two angles indicated in the Plan View are 15° and 30°.
The text between the two views reads "Pressure pads".
The text below the isometric drawing, between the two detail drawings reads "Relieved
with oversized drill".
The text below the rightmost detail drawing reads "All pivot joints of similar
construction".
-15-

on a 12.5-inch mirror and found to be adequately strong. A value of K equal to 14.5x106

will, therefore, give a maximum force that the warping harness need be able to provide.
The actual working force needed to produce the required amount of flexure in a pyrex
mirror is considerably less than this.

In order to keep the weight of the warping harness to a minimum, the cross-beams or
x-member should be designed to have the material of that member stressed to its yield
strength when a load of this magnitude or a little less is applied to the ends of each of its
four arms. If the crossed-beams or x-member is made of annealed, or hot-rolled mild
steel, this condition will be met if the maximum width (at midsection) b, and the depth,
or thickness, a, conform to the following equation:

where D is the actual diameter of the mirror blank.

For small telescopes where only a small amount of warpage will be sufficient, a
simple design of harness which contacts the glass blank with four small circular shoes
should be satisfactory. These shoes should be articulated, or pivoted so that they can seat
themselves well on the surface of the glass. In larger telescopes, where larger amounts of
astigmatism must be neutralized, the shoe design shown in Figures 2, 3a, and 3b should
be employed. These shoes have two contact areas each, separated by a distance which
subtends an angle of 30° at the edge of the mirror blank. The force from the crossed-
beams or x-member is applied to the midpoint of each shoe which, in turn, divides it
equally between its two contact areas. Each shoe is pivoted or articulated so that it
-16-

can seat itself properly on the surface of the glass. The forces developed in the crossed-
beams or x-member are transmitted to the shoes by tension or compression-struts.

In the crossed-beam design, the amount of astigmatism provided by the harness is

varied by merely turning the pivot screw at the midpoint of the two beams (see Figure 2).
In the x-member design, adjustment is accomplished by merely tightening a single screw
against one of the compression struts. (see Figure 3a). In both of these designs the
warping harness contacts the glass at eight points and the linkage is such that the applied
force is the same at all points, no matter what its magnitude might be.

One of the requirements of the warping harness is that it should hold its adjustment
more or less indefinitely. When we say "hold" we mean that flexures should remain
constant to within less than one-millionth of an inch. Since the forces involved are
considerably more than the weight of the mirror blank, and since one-millionth of an inch
is a much closer tolerance for mechanical parts than we are accustomed to dealing with,
some new precautions must be observed. Since the pressures in the system will normally
take up all play at the joints, the main precautions to be observed are to be sure that all
joints are fitted well and that no material is used in the harness which might creep or
otherwise yield slowly to the forces. One joint in particular which is important is the joint
between the shoe and the glass blank. In order that this joint should fit well, the shoe
should be lapped-in against the glass. This can best be done after the mirror is rough-
ground or during fine-grinding stages. This should be done without any abrasive between
the metal shoe and the glass. If abrasive is used, the contact areas will usually turn out to
have turned-down edges.
-17-

The most that should be put between the metal and the glass is a single sheet of onion-
skin paper. A material such as sheet cork should not be used between the metal and glass
because, under the existing pressures, it will yield slowly and thus make the harness lose
its adjustment. All the articulated joints, or pivot points in the warping harnesses shown
in Figures 2 and 3a consist of one-eighth-inch steel balls pressed into drilled recesses in
mild steel members. In order to facilitate assembly, the balls were soldered to one of the
members at each joint, care being taken not to overheat the balls to the point of
discoloring them in the soldering operation. It is felt that this construction is preferable to
sharpened points, because with small points, pressures may easily reach the yield strength
of one of the materials of the joint thus permitting plastic flow to take place.

Even though the areas of plastic flow may be quite small, they are still subject to a
small amount of creep in the metal, and in time, may allow the warping harness to get out
of adjustment.

One more adjustment - rotation - must be provided. In order to obtain satisfactory

cancellation of the astigmatism, not only must the tension in the warping harness be
adjusted to within about one-half of one percent, or less, of the required magnitude, but
the phase, or rotation of the warping harness must be corrected to within about one-tenth
of a degree. Provision for this adjustment can be made by attaching the secondary mirror-
mount to a circular plate (parallel to the mirror blank) and then attaching this plate to the
telescope tube so that it can be rotated a few degrees in either direction within its plane.

So far, we have spelled out requirements to be met by the warping harness. Since the
mirror blank is a partner in this warping operation, it too must be
-18-

manufactured to close tolerances. Its modulus of elasticity must be quite uniform

throughout, it must be closely circular, and uniform in thickness. The glass as it is
manufactured is probably uniform enough in modulus of elasticity, and the molded pyrex
blanks are probably close enough to circular. Since flexure varies inversely as the cube of
the thickness of the blank, the thickness must meet fairly close tolerances. It is
recommended that the back of the mirror to be warped be ground flat or slightly concave
or convex and the thickness of the blank at the edge should be made uniform to within
about 0.001 inch.

The warping harness of Figure 2 has a fairly firm grip on the mirror. This suggests that
the mirror-mounting lugs might simply be attached to the warping harness. If this is done,
any adjustment of the warping harness will change the collimation of the mirror and
make recollimation necessary. This is not in itself a serious objection to its use. If the
mirror is to be used in a fixed position, such would be the case in a wind-tunnel
application, this method of mounting the mirror is probably worth considering.

With this method of mounting, gravity acts on the mirror which, in turn exerts forces
against the warping harness. Equal and opposite forces are automatically set-up in the
warping harness. These will, in general, merely change all the individual forces exerted
by the warping harness by an equal amount. This might throw the warping harness out of
adjustment, but if the mirror is mounted this way and then adjusted, the difference in
pressure will be taken care of automatically - and no one will be aware of it. If the mirror
is mounted this way in a movable telescope, however, each time the telescope is pointed
in a new direction, the magnitude and distribution of these forces
-19-

in the warping harness will change and this in turn will change the amount of warp in the
mirror. This means that with this method of mounting the mirror each time the telescope
is pointed in a new direction, a small adjustment of the warping harness may be required.
In small telescopes where the weight of the mirror is small, this method of mounting may
not produce changes in the tension of the warping harness large enough to necessitate
readjustment each time the telescope is moved, and so, may be quite acceptable. In larger
instruments (eight inches aperture and up) this method of mounting the mirror may be
found to be objectionable because of the necessity of making frequent adjustments of the
warping harness.

The warping harness and mirror-mount shown in Figures 3a and 3b grips the mirror
firmly at three points around its edge. This carries the weight of the mirror and provides
positive collimation no matter what direction the telescope is pointed. The x-member of
the warping harness is held in position by a thin, flexible sheet metal spider which
maintains it in the proper position relative to the mirror, but does not otherwise restrain
its movement. Counterweights balance the weight of the warping harness in all positions
of the telescope so that only the flexure forces produced by the x-member are exerted
against the mirror. All gravitational forces on the warping harness are taken care of by the
sheet-metal spider and the counterweights. Thus the net warping forces exerted on the
mirror remain constant no matter in what direction the telescope points. It is felt that with
this design of mirror-mount and warping harness, Yolo reflectors in sizes up to 30- or 36-
inch aperture can be made to operate successfully.
-20-

Shop-Test of Warping Harness

Part of the success of the Yolo reflector can be attributed to a simple but rigid test
which is available for certifying the design of the warping harness. For this purpose, the
mirror with its warping harness is set up as shown in Figure 4. The separation-distance,
U, is given by the following expression:

where the angle, B, is expressed in radians.

Where R1 and R2 are equal and T is 1.2 times S, the following formula will hold:

The artificial star, in order to be suitable for this test, should be reasonably round, and
its diameter, d, should be no greater than that given by Eq. (21):

Since this is smaller than the pin-hole light sources used by many telescope makers,
optical reduction must be resorted to in order to adapt available equipment for this
service. A microscope objective is recommended for this purpose.(4)
-Figure 4-

-21-

With the light source, mirror, and eyepiece set up as shown in Figure 4, and the
warping harness unstressed, the image seen in the eyepiece should show only one
aberration - astigmatism - but lots of that. When the eyepiece is moved toward the mirror,
the image should take on the appearance of a sharp vertical line. When the eyepiece is
moved away from the mirror, a point will be reached where the image has the appearance
of a sharp horizontal line. Next, put a little stress on the warping harness. Now, in the
eyepiece, the vertical and horizontal line-images will be shorter and will become sharp
with less travel of the eyepiece. As the stress on the warping harness is increased, the
images will continue to shorten and may start to rotate a little. If image rotation is
observed, the mirror and warping harness should be rotated a little in the same direction.
This should make the images return to the vertical and horizontal attitudes. Further
increases in the stress and rotation of the warping harness should eventually result in a
perfect Airy diffraction pattern in the eyepiece. Until a very good Airy diffraction pattern
can be achieved in this test, there is little point in putting the mirror and warping harness
into a telescope.
Light-Baffle System
In order to provide good performance on deep sky objects, the Yolo reflector needs a
well designed system of light baffles. Since its eyepiece looks right up at the sky through
the open end of the tube, it has some of the problems of the Cassegrainian reflector. Its
main problem differs from that in the Cassegrainian, however, in that the open end of the
Yolo is confined to only one side of the secondary and does not extend all the way
around, as it does

-22-

in the Cassegrainian. The eyepiece and eyepiece tube are, or course, pointed right at the
secondary. By designing the Yolo so that the tilt angles are a little greater than the
minimum required to make the light path just clear mirrors of the minimum required
diameters, the central illuminated part of the secondary is moved a short distance away
from the open end of the tube. This helps very appreciably in the solution of this problem
and is a device which cannot be used in the Cassegrainian.

Having the eyepiece located at the bottom of the tube and looking up instead of at the
top of the tube and looking across, as is the situation in the Newtonian reflector, is both a
disadvantage and an advantage. It is a disadvantage in that it makes the light baffle
system necessary in the Yolo; but it is an advantage in that once the system of baffles is
installed, the field of view in the Yolo is much darker than that which can be obtained in
Newtonian. In the Newtonian, the eyepiece is pointed squarely at the diagonal mirror, but
it also "sees" the far side of the telescope tube surrounding the diagonal. This part of the
tube is quite close to the open end and thus is fairly well illuminated by direct starlight,
general sky light, and perhaps light from other sources. Thus the eyepiece in the
Newtonian is looking right at a surface which receives quite a lot of illumination. Since
the eyepiece in the Newtonian is quite close to the diagonal, it gets a fairly good look at
the mirror cell of the primary and any surface roughness or baffles located in the lower
part of the tube, all of which are looking right up at the sky and are, therefore, fairly well
illuminated. This feeds more unwanted light into the eye of the observer and spoils his
view of faint objects.

Once the baffles, B and G, in the Yolo (see Figure 5) are properly adjusted, all direct
starlight, and any other light coming down the tube, is
-23-

screened off; and the eyepiece can see only the back sides of the baffles. Since these
surfaces are not directly illuminated through the open end of the tube, they provide a very
dark background for the eyepiece. Since the secondary mirror in the Yolo is relatively far
from the eyepiece, the eyepiece does not get to see much of the primary mirror cell
reflected in the secondary; and, since the primary in the Yolo is located at the bottom of a
much longer tube than in the Newtonian, it receives many times less illumination
(roughly inversely as the squares of the tube-lengths) than its counterpart in the
Newtonian. Very little light from this source, therefore, gets into the eyepiece of the Yolo.

The light-baffle system for the Yolo reflector is designed as shown in Figure 5. First,
the two mirrors and the field lens of the largest eyepiece (or photographic plate) are laid
out to scale. Next, the outline of the cone of rays from a point source is laid out (light
solid lines). Then the outline of the cone of rays from an extended source which will just
fill the field is drawn in (light dashed lines). The next step is to locate baffle G. If the
telescope is intended for visual observation only and the field lens of the largest eyepiece
is no more than about 2 inches in diameter, baffle G can be located as close as two
primary mirror diameters in front of the tube proper. If, on the other hand, the instrument
is to be used in photography, with a plate as large as 5" by 7", then baffle G may need to
be located as far out as four- or five-diameters.

With the location of baffle G decided upon, lay a straight edge on the drawing so that
it will pass through points O, B, and G. With point O (inside edge of eyepiece or plate
holder) as a pivot, adjust the straight edge until both points B and G lie about the same
fraction of the distance between
-24-

the solid and dashed lines. If B and G are quite close to the solid lines, there may be an
objectionable amount of vignetting for objects near the edge of the field. In this case,
baffle G should be located a little farther out from the main structure of the telescope
tube. When a satisfactory location for baffle G has been found, points B and G are
marked on the layout. The locations of all the other baffles are then decided upon and
laid out in plan approximately as shown in Figure 5.

Once the locations of all baffles has been decided upon, the baffles themselves can be
designed. The outside outline of each baffle must conform to the inside cross section of
the telescope tube at its location. These can be laid out to scale as shown in Figure 5.
From the plan view we locate the center and radius of the cross section of each cone of
rays that passes through a baffle. These centers are the located on the baffle layout and
the cross sections are drawn in as circles. Points B and G locate the straight portions of
the inside outlines of baffles B and G. The rest of the inside outlines are drawn as arcs
of circles which clear the cones of rays by about one-inch, all the way around. This is
done so that possible convection currents in the air inside the tube can get around the
edge of a baffle without being pushed too far into one of the light paths.

If the telescope is intended for use on bright objects only, such as the
moon and planets, the baffle system may be omitted entirely. However,
even on these objects, a baffle system will give more pleasing views. A
very good compromise between the complete baffle system which has just
been described and no baffles at all, is a muffle with baffle G on its front
end plus just the straight portion of baffle B. These two baffles by
themselves will screen out all direct starlight from the eyepiece, and will
give a field of view which
-Figure 5-

-25-

is noticeably darker than that which can be obtained in a good Newtonian reflector.

Tube Design
Because of the shape of the light path in the Yolo reflector, a tube of rectangular cross
section will generally be found to be advantageous. Plywood is a satisfactory material for
its construction. If the light-baffles too are made of plywood, they may be built integrally
with the sides and thus serve as internal stiffeners for the tube. The muffle (that part of
the baffle system extending forward of the secondary mirror) can be made of relatively
light-weight material, circular in cross section, and detachable from the telescope tube,
proper.

In order to minimize the deleterious effects of possible convection currents in the air
within the tube it is recommended that the inside surface of the tube should be no closer
to the light path than one-half of the radius of the primary mirror. This means that the
shorter dimension of the rectangular cross section of the inside of the tube should be no
less than one and one-half times the aperture of the telescope.
Collimation
Collimation of the Yolo reflector requires both mirrors to be tilted through the proper
angles and the tilt-lines of both mirrors to be aligned accurately perpendicular to the
principal plane of the instrument (the plane of the paper in Figures 1 and 5). When the
warping harness is properly oriented, the two tension members ("pullers") which reach
around to the front surface of

-26-

the mirror (see Figure 3b) should be accurately in the principal plane of the system. The
two "pushers" lie approximately one mirror radius above and below this plane.

Satisfactory collimation can be achieved with the aid of a pair of fixtures, one made
for each end of the telescope. Each fixture consists of a slender rod which can be attached
to - or mounted directly in front of - a mirror. Each rod is long enough to extend across
the mirror and out past the center of the light-path passing that mirror. Each rod carries
two movable markers.

The following procedure is recommended:

1. Attach each fixture to its mirror, and make adjustments to have each rod pass
directly in front of the center of its mirror and slide the mirror-marker along each rod
until it is squarely in front of the center of its mirror.

2. Sight along the center of the eyepiece tube to the secondary mirror-marker and then
rotate primary fixture until its rod falls in the line of sight.

3. Slide light-path marker along the rod until it is on the line of sight to the center of
the secondary. This, then, places the light-path marker on the line running from the focal
point to the center of the secondary. The primary mirror-marker is now on the line
running from the center of the primary to the center of the secondary.

4. Measure the distance between these two markers; and, with the measured distance,
S, calculate the angle, B.

5. With this value of B and Eq. (5), calculate angle, A. With A and S, calculate the
distance from the center of the secondary mirror to the center of the light-path to a star.
-27-

6. Slide the light-path marker for the secondary fixture until it is at the calculated
distance from the mirror-marker (center of secondary).

With the rod of the primary fixture passing through the centerline of the light-path
running from the secondary mirror to the focal point and passing directly in front of the
center of the primary, the principal plane of the optical system is now defined by this rod
(a straight line) and the center of the secondary mirror (a point).

7. While sighting along the centerline of the eyepiece tube adjust the three collimating
screws on the secondary mirror mount until the light-path marker on the eyepiece axis
and the two mirror-markers appear to fall one behind the other on a single straight line.
When this has been accomplished, the secondary mirror has been adjusted to the proper
tilt-angle and its tilt-line must be accurately perpendicular to the principal plane.

8. Rotate secondary fixture about the axis of the secondary mirror until secondary rod
appears to be parallel to - and lie on top of - the primary fixture rod. This then places the
light-path marker of the secondary fixture in the principal plane of the instrument.

9. Adjust the three collimating screws of the primary mirror until the light-path
marker of the secondary fixture appears to fall in line with the others. This, then places
the primary mirror at the proper angle of tilt and leaves its tilt-line perpendicular to the
principal plane. Remove both collimation fixtures from the telescope.

10. With the telescope pointed at a bright star high in the sky and a high-power
eyepiece, adjust tension and rotation of the warping harness until a perfect star image is
obtained. Follow the procedures as described under the heading of "Shop-Test of Warping
Harness."

-28-

References
 1. ‘First Order Tilt Aberrations in Mirrors and Lenses,’ by Arthur S. Leonard,
Proc. Of Seventeenth Annual Convention of Western Amateur Astronomers,
University of Nevada, Reno, Nevada, August 19-21, 1965

2. ‘A Null Test for Paraboloids,’ by Horace E. Dall, A.T.M. Book 3, pp. 49-155.

3. ‘Convention Highlights from Reno,’ by Allan McClure and Leif J. Robinson,

Sky and Telescope, Vol. XXX, No. 4, October 1965, pp. 206-210.

4. ‘Schmidt Camera Notes,’ by Henry E. Paul, A.T.M. Book 3, pp. 149-155.

NOTE: Fig. 3a Secondary mirror, mirror-mount, and warping harness.

Rear view. SVAS 12-inch Yolo reflector.
Fig. 3b Secondary mirror, mirror-mount, and warping harness,
Front view. SVAS 12-inch Yolo reflector

-Equation Summary-