When you start grinding mirrors, you wind up
with more mirrors than you know what to do with. After I made the mirror
for my 12.5" trackball, I picked up a
second mirror of nearly the same focal length at a swap meet. I joked
about making a big binocular scope, then put away the mirror and worked
on other projects for most of a year. From what I'd heard, a binocular
scope was way too difficult for someone of my skill level. Conventional
wisdom held that they were fussy, temperamental things that would make
your hair fall out and increase the national debt.
But Frank Szczepanski, a telescope maker in my home town of
Eugene, Oregon, had built an 8-inch binocular that knocked my socks off
every time I looked through it, and he said it was relatively easy to
build and to use. Just for grins, he made another one out of a pair of
4" mirrors he got from Surplus Shed. That one went together in about a
week, and it worked beautifully, too.
Admittedly Frank is a guy who can put together a scope in an
afternoon (I am not exaggerating), but his success encouraged me to give
it a try. My circle of telescope-making friends, the Scopewerks, egged
me on, thinking, I expect, that I would either fail in a spectacular but
educational way or that I would succeed in an unconventional but
mundane, seat-of-the-pants sort of way. That does seem to be my
trademark, so I went into this project with that in mind.
First off, I needed to make both mirrors the same focal length, so I shortened the second one's focal length until it was within 3/4" of the first. (I could have gotten it closer, but I naively thought it would shorten by that much in polishing and parabolizing. Not so! As any mirror maker more experienced than me will tell you.) They wound up 57.9" for the first one (f/4.63) and 58.6" for the second one (f/4.69). That's a 1.2% difference, well within the 3% difference that common wisdom says your eyes can compensate for, so I was okay.
That done, I started thinking about the design of the scope.
All the horror stories I'd heard about binocular scopes centered around
keeping the images merged. It was apparently really tough to keep the
scope rigid enough that both optical trains would point at the same spot
in the sky from zenith to horizon. The designs I'd seen tended to be of
two types: two telescopes strapped together side-by-side or two optical
trains in a framework that looked a little like a truncated Eiffel
Tower. I thought about that, and about my friends' (and my own)
expectation that I would come up with something completely new, and
decided to do something completely conventional. I decided to think inside
the box for a change. A big, honkin' box. Two of them, in fact, one on
top and one on the bottom. I'd connect them with eight trusses, two to a
side, making in effect eight triangles to provide stiffness
between the two boxes. If that bugger flexed more than a few angstroms,
I would eat it.
That still left me with a lot of
decisions to make. In order to align both optical paths, I would need to
be able to adjust the aim of at least one, if not both optical trains.
Again, the 'net was full of cautionary tales and people proclaiming that
you must do thus and so or you will be doomed, doomed to
fail. The most reasonable advice that I could find seemed to be that I
should make the primary mirror mounts adjustable horizontally as well as
tiltable. A horizontal motion combined with a little tilt to both the
primary and the secondary would move the optical train's aim while
keeping the mirrors collimated.
In order to do that I had to make floating mirror cells. I
came up with all sorts of Rube-Goldberg ideas, but finally settled on
simplicity itself: make the collimation screws run all the way through
the mirror platform to rest on the bottom of the mirror box, and use
springs to pull the mirror platforms down so they wouldn't fall out when
the scope was tilted toward the horizon. The lateral motion then became
as easy as putting in four screws, two for each mirror, that could push
the mirrors sideways. They double as the mirrors' side supports, which
should be 90 degrees apart from one another anyway. I can move each
mirror an inch or so in any direction; way more than necessary.
The mirrors are full-thickness pyrex, so I only need three
points of support. That means the "cells" (if you want to be generous
and call them that) are just plywood triangles with cork pads for the
mirrors to rest on. I've got clamps that reach up over the top to keep
them from falling out.
I put metal plates underneath the collimation screws so they
wouldn't dig into the bottom of the box. The metal was shiny, so I
covered it with construction paper disks. They don't look all that black
in the photo, but they're plenty dark at night.
The collimation adjustment rods are just all-thread. They
stick up a ways so I can reach them easily when I'm merging the images.
(More on that below.) I made the right ones taller than the left because
I only use one mirror to merge the images. I leave the left one alone so
that optical train stays collimated and becomes the reference train for
the other one. (Otherwise I could wind up moving both optical paths way
out of collimation if I wound up chasing a phantom sweet spot on a bad
merging day.)
That
left me with one more major decision: How was I going to change the
interpupillary distance? People's eyes aren't all the same distance
apart, so you need to be able to move the eyepieces closer together or
farther apart to account for that. With hand-held binoculars, you just
twist them together or apart, but I figured that might be a little
difficult to do with two 12.5" scopes. That left me with two choices:
move the eyepieces sideways with focusers or rotate the secondary cages.Using focusers had its appeal. For one, I could bolt down the secondary cages and keep my rigid framework all the way to the top. But the more I played with it, the more complicated and fussy it seemed. It would require four focusers, two to move the eyepieces sideways and two more aimed vertically to actually focus the images. Worse; every time you changed the interpupillary distance, you would have to refocus. And you would have to refocus a long ways. Plus all four focusers would have to be rigid as rocks. Any flex there would be just as bad as flex in a moveable secondary cage.
Rotating the secondary cages began to look like the better option. I had already come up with a design I liked for my 12.5" trackball, and one of its features was a nice round bottom ring. I had made that one out of lightweight foam to save weight, but I wasn't worried about weight for this scope so I decided to make its cages out of plywood. I used 1/2" Baltic Birch for the rings and 1/8" doorskin for the panels. I cut the rings with a router on an arm pinned to the center, which made them nearly perfect circles.
The secondary cages are held onto the top platform with only four bolts each. The bolts have nylon spacers between the wide washers on top and the platform, so the bolts don't pinch the bottom rings of the cages. The cages rest on Delrin pads so they'll slide smoothly. There's a thousandth of an inch or so of play to the cages, but that barely affects collimation or image registration. Your eye compensates way more than they move.
The wire spiders give me another neat advantage: I can move the secondary mirrors around in the cages just like I can move the primaries around in their box. I can move them up and down, sideways, and change their tilt just by tightening and loosening various wires. I also have tilt adjustments at the center hub so I don't have to mess up the other adjustments for simple collimation.
Notice the flattened faces on the top rings just above the focusers. That's forehead space. I wanted to keep the light path between secondary mirrors and eyepieces as short as possible (to keep the secondaries a reasonable size), so that meant getting the observer's head in as close as possible.
I ran into one snag early on: the tertiary
mirrors — actually just diagonals like you'd use in a refractor —
wouldn't come close enough together for most people's eyes. The corners
banged into one another! So I did the only logical thing: I ground
off the corners. Now they come together just fine.
I made the tertiary mirrors adjustable, too, by putting foam
tape between the mirror plate and the diagonal housing. I can tighten or
loosen the mounting screws and get a little tilt that way; enough to
help collimate the entire system. (I collimate secondaries and primaries
first, then adjust the tertiaries to match.)
At first I tried just rotating the secondary
cages by hand, but it was way too easy to over- or undershoot, so I
started thinking about how I might be able to make a screw drive for it.
I mentioned my problem to the Scopewerks guys, and a couple of days
later Frank Szczepanski said, "I've found your adjustor." It's a cargo
strap tightener. The far end of it used to have a big coil of cargo
strap on it, and you would tighten the strap by turning the handle and
pulling on the strap anchors. I cut the strap holder off and left the
tightening arms, and connected the arms to the secondary cages with a
couple of aluminum bars and some vertical rods made out of a knitting
needle. The linkages allow the upright rods to get closer together or
farther apart as the cages rotate. It works beautifully! It's intuitive
enough that I can tell people at star parties "This is where you adjust
it for the width of your eyes," and that's all the instruction they need.
The focusers are odd-looking little things. I didn't really
want to buy two commercial Crayfords. Besides the expense, they're
heavy. I wasn't worried about weight at first, but as the design
progressed I began to realize that I was creating a monster, so I
decided to see if I could keep the focusers relatively light.
I had made a really light one for my 12.5" trackball, but that
design was a little too flexible to hold up under the weight of a
diagonal and an eyepiece. So I decided to think inside the box again.
How about a square set on its side? The bearings could go in two of the
faces, and I could use a stiff wire for an axle. Turns out that worked
pretty well. So well, in fact, that Sky & Telescope magazine ran a
separate article about the design, and I've got a separate web page for
that here.
The flex rocker is pretty conventional. I used three roller bearings
and one teflon pad on the bottom, and four teflon pads on the top. The
latter turned out to be a little too stiff, so I replaced the front two
pads with roller bearings. (This photo still shows the teflon pads, but
you can see the bearings in the image below.)
The altitude bearings are half-inch Baltic birch with aluminum
strips for runners. They're removable so the mirror box can sit in the
back seat of a car. I made them a little smaller in diameter than they
should have been for perfect balance, figuring that would reduce the
amount of "fin" sticking out in front. The operator has to stand
inbetween them to view but go around to the back of the scope to use the
finder, and I didn't want to be tripping over the altitude bearings all
the time in the dark. I compensate for the lack of balance with springs
on either side, and that works really well.
The ground ring is from the 20" scope that Kathy and I bought
from Mel Bartels. I was glad to see that we could re-use that, saving me
the trouble of building (and storing!) a second one.
I stained the wood "Early American" and varnished it with
Varathane. I think the finish is one of the things that surprised me
most about this scope. It turned out pretty! Who knew? I sure didn't
expect that.
In a fit of energy not long after, I built an observing chair
and stool to match. They turned out pretty, too. (See top photo. Yeah,
that's glow-in-the-dark tape around the edge of the stool.)
How it works
That's pretty much the design. So how does it all work?
It takes me about 15 - 20 minutes to set up. The collimation stays pretty tight even after tear-down and re-assembly, so I normally just have to tweak that a little at the beginning of the night, stick in a couple of eyepieces, and merge the images by tilting the primary mirrors. That throws off the collimation a bit, but hardly enough to matter. If it ever gets too bad, I shift the primary sideways a smidgen and recollimate. It took me a while to figure out which direction to move things in order to improve things rather than make them worse, but once I did I realized it's surprisingly intuitive. The big bugaboo about binocular scopes — merging the images while keeping the scopes collimated — turns out to be no big deal.
For quite a while I didn't even have the adjustment
rods extended up to where I could reach them while looking in the
eyepieces. I knew from experience how much motion and in which direction
a turn of each screw would give me, so I looked in the eyepieces,
figured out where the images needed to go, and reached down and moved
them that much. I usually had to fine-tune it with another bend-down.
The horror! It took at least five seconds' work! Sometimes ten. I
eventually extended the rods just to eliminate the bending down, but it
was really not a big deal.The extended rods are made from flexible water supply hoses, the kind you use under bathroom sinks. Those are connected to lengths of rigid PVC pipe, with sections of rubber hose stuck on the ends to make grippy knobs. You can see them in the photo to the right rising up from the middle of the mirror box and extending up to the tops of the trusses just under the eyepieces. Now I can merge the images while looking through the eyepieces. It is a little easier, I have to admit. (Click on the photo for a larger view.) The adjusting rods cross over from left to right so they don't get in the light path, which means the left knob controls the right mirror and the right knob controls the left mirror, but that doesn't really matter. You get used to the way they work relatively quickly. (I can often get people at star parties to merge the images for their own eyes even if they've never used a binocular scope before.)
Once the two images are merged, the sky takes on a depth and richness like nothing you've seen before in even the largest single-mirror scope. It's simply stunning. Star clusters look three-dimensional, rather than like salt spilled on velvet. Nebulae look like they're floating right out there in front of you. And the Moon! OMG, the Moon! If that was the only thing I ever looked at with this scope, it would be worth it. It looks like a big bumpy rock floating about five feet away. The horizon curves away from you, and the shadows along the terminator have depth. I used to feel let down when the Moon would get into first quarter, but now I'm eager to get out and observe on moonlit nights just so I can look at it in stereo.
The distance even to the Moon is way too great for real parallax, of course, but the image processing parts of your brain don't know that. You've got two separate images going into two separate eyes, so the wetware does its magic and you see everything in stereo.
The rigid design does its job: the images stay merged from zenith to horizon. They stay merged when I swap out eyepieces, too, except for my highest-power ones (4.7mm Explore Scientifics, which give me about 315x). For some reason, those eyepieces don't like the same settings as all my other ones, so I have to tweak the merging when I use them. Like I said above, it's the work of a few seconds to do that, so it's no big deal, but I keep trying to figure out why they're different. I think it may be their tapered barrels interacting with the set screws differently than the straight barrels of the other eyepieces.
I occasionally have to tweak the merging even when I don't use the high-power eyepieces. Stuff gets bumped, or temperature changes shift the tolerances, or cosmic rays addle my visual cortex. I've learned that different people's eyes have different abilities to merge. Some people are slightly cross-eyed, while others are slightly wall-eyed. Most of us can meet somewhere in the middle, but it doesn't take much deviation before you start dropping the outliers off the bell curve. I've learned to ask them "Which way does the right-eye image need to go?" and I tell them how to use the merging knobs to quickly dial it in for them. (After which it's usually wrong for everyone else.) Fortunately this only happens for maybe 1 out of 20 people.
The light-gathering ability seems to be about the equivalent of an 16 or 18-inch mono scope. I've been able to see 16th-magnitude galaxies through it, which I can't do with just a single 12-inch mirror. Planetary nebulae like the Ring and the Dumbbell are amazingly rich in detail through this scope.
All in all, it's an easy scope to use and the view through it is way, way more exciting than through a mono scope. It's rapidly becoming my favorite.
People kept telling me I should name the scope "Gemini" and call one mirror "Castor" and the other one "Pollux." I liked the idea, but my favorite star names are over in Libra instead. I put the names on the mirror covers.
I'd
love to hear from people who are interested in this scope design. Please
feel free to email me at the address on the right. (Sorry you can't
click on it or copy and paste it; it's a graphic file to thwart spambots
that search the internet for addresses to send junk mail to.) I
have no idea how much mail this idea will generate, so I can't guarantee
a response, but I'll do my best to answer everyone who writes with a
genuine question or comment about the design.