ATM Binocular viewer from a kit - initial impressions

["Tom & Lou Krajci" <krajcit@3lefties.com>]

If you have not observed planets/moon through binocular viewers, you are
missing out.on the full observing experience.  After a view of Jupiter
through bino-viewers on a 15" scope I was hooked...I had to get a pair!

But they are so expensive.  OK, this is the ATM list, surely others have
built bino-viewers to save $$$...but so far I've not seen mention of this in
web pages, magazine articles, or the list archives.  Why?  If others can
contemplate building very large scopes, complex optical designs, powerful
astronomy related software, motor drive systems, mounts, electronic cameras,
etc....why not something simpler (yet so compelling for the observing
experience) like binocular viewers?  (I can't answer this, but the fact
remained - I had to get me a pair, but on a budget.)

Searching the list archives did turn up one or two mentions about a
bino-viewer kit put out by TL systems:
http://pw1.netcom.com/~tlsystem/index.html   but no substantive information
about it.   I asked the list about this kit a couple months ago and got back
a couple words of warning on the quality/workmanship of their kits.

Maybe I had the bino-viewer jones too strongly to heed the warnings, but I
threw caution to the wind and ordered the kit with beamsplitter/mirrors.
OK, I thought...some assembly required...no problem.

When I first received the kit and examined the contents I was initially a
bit put off.  The bino-viewer box/chassis was what appeared to be a heavy
duty aluminum project box, with holes drilled in it for the various
components to attach to.  The eyepiece holders and mirror holders were
sections of PVC pipe.  The beamspiltter holder was merely a short section of
aluminum U channel.  The beamsplitter looked OK, but the first surface
mirrors looked like they were hand cut from a larger piece...no chamfering
of the freshly cut edges.  No precision machining was evident, nor any
provision for sophisticated, repeatable, finely controlled collimation
adjustment.

Just think of the problems you face in collimation/positioning one optical
element:  movement (translation) in three axes, and rotation about those
same three axes.  For a binocular viewer that means you have six degrees of
freedom you must control...FOR EVERY OPTICAL AND MECHANICAL COMPONENT!  In
this particular case  the optics were:  one cube beamsplitter, three first
surface mirrors (and later you need to add two eyepieces and perhaps a
barlow lens).  Mechanical components included the 1 1/4 inch OD barrel, the
beamsplitter holder, three mirror holders, two eyepiece holders, and two
rails to slide the eyepiece holders for different eye spacings.  Oy veh!  My
head swam at the thought of how many ways I could screw up collimation with
so many components.

OK, I've ranted long enough to convey that I feel the kit is below standards
for "fit and finish."  (Ironic, it reminded me of the kind of work I would
do with scrap parts to prototype a device, but personally I would not feel
comfortable marketing this assembly of parts...and the documentation doesn't
give you any indication of all the intricacies of collimation you must
control to be successful.)  I decided to press on and make a workable
bino-viewer out of this kit...in spite of its shortcomings!  ;-)

Along the way I discovered a significant design shortfall:

1.  The project box/chassis has a trapezoidal cross section!!!

How do you use this trapezoidal box in a straight-through bino-viewer?  In
my case I used shims on the 1 1/4 inch barrel/nosepiece and the eyepiece
holder/slides to make the entering/exiting mechanical axes parallel.  I used
no fancy tools here, just eyeballs, rulers and carpenter square to get
things reasonably close.

There were other nuisances to overcome:

1.  The eyepiece holders would not spread open wide enough for my eye
spacing.  My eye spacing is about 73-74 millimeters..but the kit says it's
able to handle up to 70mm spacing...yet I had to remove far more than 2mm
from each eyepiece barrel.  This required lots of whittling of the PVC
eyepiece holders to allow them to spread far enough apart.

2.  The eyepiece holders had significant slop for holding eyepieces...it
would be very hard to insert/remove eyepieces in any resemblance of
repeatable mechanical alignment/position.  Lots of shim stock was used to
take up the gross slop along the inside of the eyepiece holders...and make
sure they where held in parallel mechanical alignment.

3.  There is no precise way to rotate/translate any component...no push-pull
screws against levers...no precision machined ways/guides for linear motion.
This required a considerable amount of time in assembly of the components,
evaluating collimation, disassembling and adding/removing shims...tightening
screws, only to see a part shift out of place as the screw tightened, so
loosen and try again...and in many cases...more time to think about the
entire collimation problem.

4.  As you alter position/collimation, you also alter optical path length,
which makes simultaneous focusing of the two eyepieces problematic...another
factor to juggle.

There were several evenings I walked away from the work table pissed off,
realizing this wasn't going to be a simple, quick assembly...and feeling
like I had been duped.

After enough attempts I made progress at juggling all the collimation
issues...locking some things down that would not have to move again, and
leaving only a couple items to move about to get a final collimation.
Eventually I was able to collimate the bino-viewer well enough by daylight
at the work table so that it only required one or two types of adjustment at
the telescope to make it easy to fuse the two images.

Surprisingly I was able to get near identical optical path length, while
controlling up/down image placement and image rotation so that image fusing
was easy and comfortable.  I could verify this by observing through a
telescope during daylight, with my eyes a bit back from the bino-viewer...I
could look at the images in the bino-viewer and see that when the two images
were fused, then I could glance off to the side and look at distant objects
without the bino-viewer...no feeling of discomfort, crossed eyes, etc.

Unfortunately this particular kit doesn't do the smoothest job of changing
eye spacing, but with practice I can quickly adjust the spacing for my wide
eyes, or the more normal spacing of my teenage son's eyes...and re-tweak the
collimation by rotating the eyepiece/mirror holding assembly to correct the
up/down image placement...which is the only collimation error that crops up
during eye spacing changes.

In my scope I can't rack the focuser in the additional 5-6 inches, so I use
a barlow lens to shift the focal position back far enough.  (This is a great
reason why you should avoid a low profile focuser in your newtonian...or the
barlow/extension tube will intrude into the main mirror's light path.)  The
inclusion of the barlow might be a good idea for this kit no matter what
type scope you have (unless it's a slow f/ratio system, say f/10 or slower.)
With the barlow I use 25m eyepieces to get about 300-400x from my 16 inch
f/6 scope.  The barlow kicks my effective focal ratio to about f/18-20.
Vignetting is not a problem with such a slowly converging light cone...but
more importantly...you use low power eyepieces, which are probably much more
forgiving to collimation errors in this bino-viewer kit, making the odds of
success higher.

So...after all this...how does it work on the planets?  I have not been able
to test it in good seeing conditions, but this bino-viewer makes observing
much, much more comfortable and relaxing...and at least in bad seeing it
makes it easier to see cloud details on Jupiter compared to observing the
planet with one eye.  Also, color differences in clouds and festoons is more
obvious with the bino-viewer.  My 16 year old son, more of a novice
observer, notices the same improvements in viewing comfort, detail, and
color perception.

There are two other optical issues that are inherent in this design:

1.  The beamsplitter cube uses a dichroric reflecting layer, which imparts a
slight color shift.  One eye gets a slightly bluer image, the other eye gets
a slightly redder image.  This is a subtle effect that is apparent when you
use the same eye to observe the left, then right image, but when I use both
eyes at the same time...after a few seconds the eyes/brain adjusts and I
don't even notice the difference.  (A different type of beamsplitter would
avoid this problem.

2.  The two images are also different with respect to polarization.  One
optical path has vertically polarized light, the other has horizontally
polarized.  I suppose the dichroric beamsplitter is also to blame for this,
but I'm not sure.  This made for some interesting views of objects on my
work table, with polarized glare coming from the table top, etc...but I do
not notice this as a problem when observing Jupiter or Saturn.  I do not
think the bright planets have a strong amount of polarization to their
light, so this may not be a big factor at the telescope.

But this is only an initial report.  What's next?

1.  I need to evaluate this system in better seeing conditions to see how
much better this kit is compared to a single eyepiece.

2.  I want to encourage more ATM's to consider the building of binocular
viewers.  They are an amazing tool to the visual observer.  We need
websites, drawings, diagrams, etc. to help build a knowledge base.  Maybe a
future ATM book will consider a chapter on building binocular viewers?
Maybe an article in Sky & Telescope?

Let's not let this thread die...ATM'ing of bino-viewers  has been neglected
too long.

Thanks,
Tom Krajci

PS.  This is not the only way to skin the bino-view cat.  For example I
think Apogee sells binocular microscope heads for $99.  That sounds like a
good, relatively inexpensive start.  However, they have a 45 degree angle to
them.  (In my particular case I preferred a straight through design.)