Operations
Mirror-o-Matic will take most of the manual labor (some call this drudgery, some call it joy) out of mirror making. Because the mirror is rotated on a turntable, it will have a very nice figure of revolution. The chance of producing a severely astigmatic mirror is greatly reduced.
Other than these two advantages, the Mirror-o-Matic
will do nothing that cannot be done by hand ‘walking around the barrel’.
Although the final results should be similar, the
mechanics of machine grinding and polishing can be very different from
handwork. This machine should be used as described below. Although hand
techniques such as full sized tools and mirror on top will work in a fashion,
much better results will occur with proper machine use.
Before starting machine grinding and polishing, one
should at least read several books on mirror making. Most of these books will
be about hand work, but knowledge of calculating and measuring a sagitta, pouring pitch laps, building tile tools, and most
importantly reading and interpreting the Ronchi and Foucault tests are
extremely important. Knowing how to test a mirror and apply the proper
corrections is by far the most important aspect of mirror making.
This operation manual will in no way teach you how
to make a mirror. It is only a rough guide for speeds and techniques. If
Mirror-o-Matic users will send details of their successes and failures, I will
gladly post them on the Internet for the benefit of others.
Machine grinding and polishing is almost always done
with the mirror on the bottom (MOB) and the tool on top (TOT). The tool can be
another piece of glass, a tile tool or a metal tool. The tool is always smaller
than the mirror if a concave mirror is desired. Tools of 50% to 80% of the
mirror’s diameter are normal. Tools of less than 50% diameter may cause
problems while grinding, although they are valuable when figuring a parabola
with a pitch lap.
Two topics that always come up in machine work are
rotational speed and overarm strokes per minute. Neither is really all that
critical as long as a ratio between the two is such as to prevent a repeated
pattern from showing on the mirror. Experience shows the following procedures
work well.
Mirror
Centering
It is important to center the mirror blank
accurately on the turntable. The center of the blank must be positioned over
the center of the turntable shaft. One thirty second of an inch is about the
maximum acceptable offset that should be tolerated.
The center of the blank’s figure of revolution will
be the center of the turntable shaft. Wherever the mirror blank is positioned,
the lowest spot of the mirror will be generated over the turntable shaft’s
center. If the mirror is offset, its low spot will not be the physical center
of the blank. If the mirror is removed from the table for checking or cleaning
and replaced in a slightly different location, a new center and lowest spot
will be created. A whole set of hills and valleys will be created as the two
figures of revolution attempt to merge.
This
recontouring will occur each time that the mirror is repositioned during
grinding and polishing. Hours can be added to grinding and polishing times
while new radii are being generated.
If astigmatism shows with multiple spherometer
readings while grinding or with Ronchi band readings while polishing, check
carefully that the mirror is centered on the turntable and has no run-out. This
can be checked by lightly holding a stationary pencil tip to the edge of the
rotating blank and checking for a continuous line. A low cost dial indicator
can also be positioned at the edge.
Mirror-o-Matic uses hat sections of plywood to hold
the mirror. When creating these disks, first lay out
and mark the center. Then use a compass to draw a circle the same diameter as
the mirror bottom. Keep in mind that the mirror may have beveled sides.
Cut and sand to the outside edge line then drill the
center hole right on the center mark. Use a good quality Forstner bit with a
brad point to assure accuracy.
Three or four holding blocks with
slotted holes can also be attached to an oversized turntable hat section. These
can be used in lieu of taping the mirror to the hat section. Do not tighten
these blocks against the mirror when polishing. They will cause astigmatism if
set too tight.
A stationary arm can be placed over the rotating
empty turntable and a pencil or marking pen can be touched to the turntable
face to create a series of concentric lines on the turntable. This will make it
easy to center the mirror.
Spherometers
After several years of machine
grinding, I am of the opinion that it very difficult to learn to machine fabricate mirrors without a proper spherometer.
A six-inch spherometer
with a ten thousandth gauge is accurate to about 0.3 inches in radius of
curvature (or 0.15 in focal length) for the typical mirrors that we create.
This usually is plenty adequate. A spherometer with a thousandth gauge has
one-tenth the accuracy, which is not enough.
It
is my opinion that a spherometer should not be longer than half the diameter of
the mirror. Many people feel that longer is better and that the best
spherometer is one that is about 1/4 inch shorter than the diameter of the
mirror to get the most accuracy. I feel that using a spherometer of this size
reduces the usefulness of the instrument greatly.
A
sub diameter spherometer can be used to measure the center area of the mirror
and then the outer areas. If there is any difference, the mirror has zones. (Anytime that the ROC of the mirror is changing when grinding, you
will have zones.) Polishing out a zone of even a few ten thousandths is time consuming.
Take readings all over our mirror and compare them before polishing.
If
readings are taken in several concentric areas of the mirror and there is any
difference, you have astigmatism. Again, it’s difficult to polish out. Always
proceed with fine grinding until all spherometer readings are the same.
A
long spherometer will only show the relative depth of
the central sag verses the edge, everything between can be rippley and it will
never show until you start polishing.
I
have checked a lot of mirrors over the last few years. Many machine ground
mirrors have zones. It’s the nature of machine grinding. Many hand ground
mirrors have astigmatism. Plus zones. Again, without a
small spherometer, you have no way of finding these problems until you polish
and test and then you will probably tell people that you just polished a huge
hole in the middle of your mirror when in fact, its been there all along.
If
you are using a micrometer, you will want to make a three-legged instrument. If
you are using a dial indicator, you can use a 2 or 3-legged unit. Two legged spherometers need to be tipped inwards until reading the
highest and correct value.
Another
advantage of small spherometers is that you can get by with a dial indicator
with only 0.08 inches of travel.
MSC
sells
Rough
Grinding
Rough grinding (hogging out) is the first stage of
grinding performed on a new flat blank and tool. The purpose of rough grinding
is to generate a ‘dish’ in the center of the mirror. This dish is deepened
until the correct radius of curvature (R.O.C.) is produced.
The procedure to set up the machine for rough
grinding is to first set the eccentric to zero stroke.
Then set the overarm to tool center over mirror center and set the overarm
platform bolt. Now set the eccentric for desired stroke. You may want to tip
the overarm up and remove the tool before plugging in the machine. Then after
verifying that the overarm has a safe amount of stroke, stop the machine and
replace the tool.
My current recommendation for rough
grinding is as follows:
Tool on top. The tool should be 50 to 80% of the mirror size. It must be
smaller than the mirror. The turntable should turn 40 to 80 rpm depending on
mirror size. Smaller mirrors can spin faster. Too fast will sling grit and
water off the mirror and make a mess.
Set the eccentric so as to stroke the tool center over center
edge to edge with about 1/4 inch tool overhang on each side of the mirror. Stroke
Use 60 to 80 grit. Add grit and water
every 2-3 minutes for 6 to 10 hours.
I check the sag with a homebuilt 6-inch spherometer with a
1/10,000th micrometer every 30 minutes of so.
Without a spherometer it is hard to tell what is happening.
An eight-inch plain glass mirror should take 3 or 4 hours. A
twelve-inch mirror will take about 5 hours. About one or two cubic inches of
glass will be removed every hour. Pyrex may add 50% to these times. Some very
hard glasses will double the times.
If I am starting with a flat mirror blank, I use a flat tile
tool. They will grind spherical quickly. Tiles are soft enough to recontour and
make good contact. Avoid pre-curved tile and metal tools while hogging. The
curved tool surface will not allow good contact until it is worn flat. It will
then be compromised for fine grinding. If you are starting with a concave
blank, use the blank as a mold to pour a tile tool with the same curve. Do not
use a curved tool with a flat blank or a flat tool with a curved blank.
If your strokes are too long, the mirror will stay flat. If they
are way too long, the mirror may go convex. Adjust the tool overhang to produce
the desired radius of curvature. Short strokes make deep curves. Long strokes
flatten curves. The fastest hogging will occur when there is no overhang.
A tile tool may completely grind an 8-inch f/8 mirror without
reapplying tiles. Bigger mirrors will
probably need a second and third layer of tiles. It is possible to stand the
tiles on edge within a form set on a Saran Wrap covered mirror and encase them
in Pourstone Plaster. This will give you a full inch
of tool thickness which will definitely last through-out grinding.
A tablespoon of grit and a squirt of
water will be required every 2 to 5 minutes. Listen to the sound of the grit
tumbling and it will be apparent that a recharge is necessary.
Weight should be added to speed the
grinding. I find about 20 to 30 pounds to be ideal. Five-pound barbell weights
are only a couple of dollars and can be stacked directly on the arm over the
swivel. Too much weight can deform the tool or turntable and create zones. It
may also crack tile.
A second method of rough grinding is to use a 55% - 80% diameter tool with a stationary overarm. Set the eccentric for zero throw. Set the arm offset so that one edge of the tool has about 1/8th inch overhang. Make sure that the other tool edge covers the mirror center if you are using a small tool.
Add grit, water and weight as above
and let the tool ride on the rotating mirror. The tool will spin on the mirror
with their outside edges rotating at the same rate. Because there is little
relative speed between the outside edges and a lot of relative speed at the
mirror center, a sphere will be ground into the mirror and the tool will become
convex. The longer that it spins, the deeper the curve gets. Move the arm out
and the edge developes a chordial motion, which grinds the edge down and makes
the curve shallower. Find just the right spot and the tool doesn't change the
ROC. This is the place that you do all your fine grinding and the start of
polishing.
This is my favorite way to grind.
Without the twisting moment caused by pushing a heavy tool across the mirror,
additional weight can be stacked on the arm which speeds grinding. It is easy
to add grit and water since half the mirror is always exposed. I strongly
recommend that you use this method.
If you are starting with a new flat mirror blank, I
would recommend that you begin with a side to side arm movement until the curve
becomes established. Occasionally, the stationary arm method does not work
until a shallow curve is begun.
Fine
grinding
After the mirror has been rough ground
to the proper sagitta, fine grinding can begin. Fine
grinding on Mirror-o-Matic can be done with the stationary overarm
method. The overarm does not need to swing side to side if the tool is large
enough to cover half the mirror diameter when in the neutral position described
below. The procedure is to first set the
eccentric offset to zero throw (position of the eccentric rod end bearing is at
the center of the eccentric drive) and tighten the bolt. Next set the overarm
offset so that the tool hangs over the edge of the mirror about 15% of the
tool’s diameter.
The
inner tool edge must extend over the center of the mirror. Tighten the bolt in the overarm offset.
Turntable
speed can be fast or moderate. If desired, the eccentric can be hooked up at a
slow speed and a very small amount (1/2” to 1") of back and forth motion
can be introduced. This will blend away any tile tool marks that may appear.
Tools should be in the 75 to 80
percent of mirror diameter range. Smaller tools can be used for fine grinding
if they are swept across the mirror surface enough to grind the mirror center.
Sweeping the tool will require larger tool overhangs to maintain the neutral
position mentioned below.
Tile works fine except small pieces may break off and leave very small scratches. If you used tile for fine grinding, seal the tool edges and areas between tiles with casting resin. It is best to use unbroken tiles as the edges break off easily.
I fine grind about 30 minutes with each type of
smaller grit. Check for a uniform grayness all across the mirror and check for
pits and scratches with a magnifying glass before moving on to the next grit.
Check the curve after the first
fifteen minutes. If it is getting deeper, move the arm out a bit. If it is
getting shallow, move it in. You will
soon find the "sweet spot" where the curve maintains its R.O.C. This
is the neutral position.
It is important to maintain the R.O.C.
when fine grinding. If the R.O.C. changes during fine grinding, zones will be
created which will take a long time to polish out. The
neutral position uniformly removes glass across the entire mirror surface
without changing the mirror’s radius of curvature.
Use a spherometer of one half the
mirror’s diameter to check the outer portion against the inner portion of the
mirror’s surface. The ROC of the two must be the same. The ROC of the mirror
should not be changing while fine grinding.
If the readings are not the same or if the ROC is changing, you have
created zones. Adjust the arm in or out as necessary to find the neutral
position.
If you do not have a spherometer, mark
the mirror face with a Sharpie felt tip pen and assure that the surface is
grinding uniformly. If areas are still black after a few minutes, fine grinding
is not complete and a spherical surface has not been established. It is
important to maintain the neutral position for at least 30 minutes to assure
that there are no zones ground into the surface.
Polishing
Most polishing problems are due to
improperly formed pitch laps. The pitch lap must be fully pressed to conform to
the mirror's sphere. Make sure that the
lap's channels are uniformly cut to allow for pitch flow. Cold press overnight and check the channels.
Wavy channels are an indicator that there were high and low spots in the lap
and the pitch has recontoured itself. Straight channels are an indicator that
the lap is in proper contact with the mirror surface. Recut and repress if
necessary.
It is important that the pitch lap be
symmetrical about its center. A lap that is not symetrical
will not spin correctly when polishing and will drag across the glass causing
zones and grooves.
A pitch lap of 70% to 80% of the
mirror diameter works well for polishing. Use the same stationary overarm
techniques as in grinding. It is easier on the machine than using a stroking
motion. I use 70 rpm for a turntable speed. It is desirable to introduce a
small amount of eccentric motion (1/2” to 1” total tool movement). Use the
slowest eccentric speed (8-rpm). Good results can be had by setting the
eccentric to produce 1/4" of
overhang when the arm is all the way in and 15% of the tool diameter
when all the way out.
When polishing, we are looking for the
neutral position as in fine grinding. If the arm is in too far, a second inner
sphere with a shorter ROC will be polished into the mirror. If the arm is too far out, a shallower sphere
will be overlaid on the original sphere and it will grind the edge down. Each
will show as a turned down edge when testing.
If you have just finished fine grinding, you can use
a pitch lap of the same diameter as the fine grinding tool and use exactly the
same overarm position and stroke as in fine grinding. If the pitchlap is
properly pressed, the surface will polish to a perfect sphere quickly. This is
the surest method of getting a perfect sphere.
After the first 30 minutes of
polishing, check to see if the surface is taking a uniform polish. If the
center is polishing faster, move the arm out. If the edge is polishing first,
move the arm in.
Use the Ronchi test to check that the figure of
revolution is spherical with no edge problems. The center is not as important
as the edge. If it is slightly high, it will be taken down in figuring. If it
is slightly low, it’s part way parabolized.
Under the Ronchi test, if the mirror center is
black, the pitch lap is probably high in the center. It must be pressed better.
Check to make sure that the channels are really clean and straight (so the
pitch can flow properly) and press overnight with 5 to 10 pounds of weight.
This method almost always produces perfectly straight Ronchi lines if the lap
is properly pressed and the neutral grinding and polishing positions are used.
If the Ronchi test shows sever zones, you probably
ground them in. You may want to go back to grinding with 9 micron grit until
the spherometer shows equal readings.
I have had good results with Gogolz 64 pitch. Add 3 heaping teaspoons of quality cerium in 32 ounces of water and let set overnight to dissolve clumps which can cause scratching. You can also pass the cerium mix through a coffee filter. With a fresh pitch lap there will be quite a bit of resonance. The arm will vibrate and the machine will shake. You will only be able to use five pounds of weight at first and maybe 10 pounds after the lap breaks in because of the vibration. Faster rotational speeds will reduce the resonance.
Polish the mirror fully before moving on to
figuring. Again, the edge should be perfect before moving on. It will take 3 to 5 hours to fully polish a
spherical mirror.
If you find deep scratches during polishing, you
will have to determine whether to live with them or go back to fine grinding.
It takes forever to remove even 220 grit scratches by polishing. It is better
to go back to 15 then 9 micron fine grinding for 20 minutes to remove the
scratches and to polish for another 2 hours again than to try to polish out the
scratches by polishing for 10 more hours. Or you can live with the scratches.
They are of no optical significance.
I do not recommend using the fine grinding tool for
a pitch lap. If you find scratches, it will be difficult to go back to fine
grinding without the original tool.
If the pitch lap does not seem to be biting the
glass surface, run hot water over it for a few seconds to soften it a bit and
then scrub it with a brass of stainless steel brush to produce a micro-faceted
surface. This will be necessary after each hot or cold pressing.
Figuring
Figuring is the process of changing
the mirror surface from a sphere to a parabola. We do this through addition
polishing of the center of the mirror while leaving the edge essentially
untouched. The amount of material removed is usually measured in ten
thousandths of an inch or less.
Sub diameter pitch laps are usually used (40-60
%). The polishing lap can be used for
figuring by pressing a ring of aluminum foil or waxed paper into the outer
portion of the lap. When the foil is removed, the new lap will be effectively
smaller. If you wish to change the mirror surface back to a sphere, you can
repress without foil and regain the area.
Slow the turntable to 7 rpm and speed up the
eccentric to 16-21 rpm. This will give a
proper stroke ratio (5 or 6 to 1 is about right). Slowing down the turntable will prevent the
pitch lap from spinning and grinding a new sphere into the surface. When
figuring, we want to 'scrub' a selected portion of the surface away, not create
a new sphere. The fast arm motion will cause scrubbing.
Since turning a sphere into a parabola requires
deepening the center of the mirror, the pitch lap should stroke from edge to
edge. Use only a few pounds of weight and concentrate the polishing towards the
center at first. Later work your way out by increasing the stroke. This will
blend the zones smoothly together.
You can use a Ronchi
computer program to show what the Ronchi lines should
look like for your mirror. When figuring, the more the lap rubs on a zone, the
farther apart the Ronchi lines will move. Figuring is
the process of deciding where to place the lap and for how long it should rub
to correct the Ronchi lines.
It may be useful to slide the overarn assembly
rearward to figure out zones that can appear in the outer radius of the mirror.
You can also offset the arm over high zones and use small side to sidestrokes
to blend them into the correct zones. Several builders have cut a 5/8-inch slot
down the overarm and hinge to allow positioning the lap in the correct zone.
The Mirror-o-Matic 20 has this groove cut into the overarm for this purpose.
Small laps and rose or star shaped blending laps are
often used with machines. I have had good results with 60% shaped laps.
You may want to disconnect the eccentric and move
the arm to the up position or remove the overarm assemble totally and just use
the turntable alone for hand figuring.
It is up to you to experiment. Please pass your
experience back to me, so I can share it with others.
Turntable and eccentric rotational speeds.
With
the recommended 1725-rpm electric motor, the following RPMs will be available.
Intermediate
shaft = 258 rpm
Slow
speed shaft = 38 rpm
Shaft
Pulley |
RPM |
Belt
Required |
1-1/2” |
32 |
|
2” |
43 |
|
3 |
65 |
|
4 |
86 |
|
Shaft
Pulley |
RPM |
Belt
Required |
1-1/2” |
4.7 |
|
2” |
6.3 |
|
3 |
9.5 |
|
4 |
12.6 |
|
5 |
15.8 |
|
6 |
19 |
|
Shaft
Pulley |
RPM |
Belt
Required |
1-1/2” |
7.1 |
|
2” |
9.5 |
|
3 |
14.2 |
|
4 |
19 |
|
5 |
23.7 |
|
6 |
28.5 |
|
Note:
Belt sizes are not given due to the different combinations that can occur from
varying speeds with low and intermediate shafts. Space is given for you to
enter your own preferred belts. The belts called out
in the parts list will be adequate for most use.