Building a sky tracking mount for astrophotography?

[Donald Qualls]

I've seen many online plans at various times and places for so-called "barn door" mounts, simple sky tracking mounts to allow any camera with B or T shutter to be used for tracked sky photos -- tracked, so that stars, planets and nebula will appear as points (or at least small slightly fuzzy circles) instead of long trails, as would be the case with exposures over a few seconds using a fixed camera.

What all seem to have in common is some combination of somewhat complicated geometry requiring considerable precision in fabrication, and tracking time limited to an hour or so before significant rate errors crop up (even discounting any errors in aiming the rotation axis at the celestial pole). It seems to me that a simpler mount could be built using a plain motor drive (shaded pole synchronous motors are common, have good torque, and hold rate very well if not asked to work against too much resistance), suitable belt reduction (belts can be readily slipped to move the mount between exposures and don't have periodic errors as gears can), and a counterbalanced platform to minimize torque requirements.

Has anyone seen a plan for such a device, or had any experience in building one? Or would it be simpler to modify an old style telescope clock drive with a tripod ball head in place of the telescope tube?

 

[rbarker]

Even the "clock motor" driven mounts for (time-exposure) astrophotography usually include a device for making minor adjustments in the tracking by placing a magnified "guide star" within a target of some sort. Commercial versions of these are available from astronomy shops like Orion Telescopes, but are usually intended for use on equatorial-mount telescope tripods to get the dual-axis movement required.

I'd think it would be easier to adapt an equatorial mount tripod to wide-field photography use, as opposed to the other way around.

 

[Dave Mueller]

The "ultimate" barn door tracker (or Scotch Mount) would be one of Dave Trott's double arm designs (Sky & Telescope: February 1988 and April 1989), driven by a stepper motor using Ray Grover's circuit. I managed to etch a circuit board for the drive circuit and got the mount half finished before I came across a deal for an equatorial tripod for my scope. The stepper motor route is much simpler than making anything involving reducing drives, belts etc. The mechanics of the whole mount can be rather crude. I think the single arm/single hinge driven by a hand-turned screw is very accurate up to 15 minutes. I would guess that any motorized scope tripod should hold a 35mm camera and up to a 200mm lens, and would certainly be easier unless you're very handy.

Dave Trott's Homepage:
http://hometown.aol.com/davetrott/page17.htm

Quartz Controlled Scotch Mount by Ray Grover:
http://www.mikeoates.org/mas/projects/scotch/

Another page with lots of info:
http://www.astronomyboy.com/barndoor/index.html

 

[Dave Mueller]

Also, get a copy of Michael Covington's book "Astrophotography for the Amateur", ISBN: 0521627400, and check Jerry Lodriguss' site www.astropix.com.

I'd recommend a simple single arm tracker to see if you really enjoy this aspect of the hobby before you go spending a lot of time to build a double arm unit, or spend some money on a decent mount. You can probably build one in an afternoon for $20.

 

[Donald Qualls]

Well, it's the limited exposure time of the single arms, and even the double arm Scotch mounts that bothers me. I don't need guiding -- that's for high magnification; I'm talking about putting, at the heaviest, a 35 mm SLR with 135 mm lens and teleconverter or 9x12 cm plate camera on this mount, but I'd like the option, where skies are dark enough, to stop down a bit and leave the mount running for a couple hours.

And an electronically driven stepper motor has the serious disadvantage that it requires me to work with electronics. I don't have a problem with that, as long as it a) doesn't require handling a soldering iron, and b) doesn't require handling discrete compnenets or integrated circuits. I don't even mind soldering irons, as long as I'm not soldering items that will be damaged by leaving the iron in place long enough to flow on a little extra solder.

But a shaded pole motor can be taken from an old desk fan or similar, gotten from a thrift store for next to nothing.

I suppose I could use a shaded pole motor with suitable reduction to drive the screw in a Scotch mount, but I'd need to make something adjustable to allow for calibrating to the actual running speed of the motor -- otherwise, even if I use a motor/gear set from a clock (and even if such a unit has enough torque -- clock motors are notoriously weak), I'll need more precision than my woodworking normally exhibits to get accurate tracking.

I'm asking this because I'm trying to decide if it would be simpler to build a platform for the camera to replace the telescope in the mount for my little Meade reflector, and recalibrate the (wrong model) motor drive to match sidereal rate, then live with the little jumps that happen periodically with the (cheap, better than hand tracking for visual observation) tracking drive.

The other consideration, of course, is whether I could align the mount well enough to be good beyond an hour in any case. A simple sight tube sized to put Polaris at the edge, along with a chart to know which direction to offset, should allow aligning within a few minutes of arc -- is that enough to avoid problems visible with a normal lens?

Yes, I could probably build a manually operated single-arm barn door for under $20 (one piece of hardwood 1x6 a couple feet long, one hinge, threaded rod, assorted nuts and fittings, and, initially, wedges, hooks and rubber bands to mount and point the camera, plus a stopwatch or similar to know when to turn the screw), and from my yard, where the sky has a significant glow (even on the edge of a town of about 80,000) a five minute exposure is pushing things a bit. But I'd like to avoid building the whole thing again when I'm ready to drive out in the country a bit and try exposures of an hour...

 

[rbarker]

It would seem to me, Donald, that without precise and almost constant tracking, you'd be likely to get some "wander" of individual bright stars. Doing the tracking for multi-hour exposures would get pretty tedious, I'd think. So, I think the requirements really revolve around what you want to accomplish with the images, and where the dividing line is between a circle of confusion and what might be more like a nebula of confusion.

But, I've only poked at the problem enough to realize how difficult it is to precisely align a regular telescope's equatorial mount, thus enabling the clock drive to function properly. You really need to get input from someone with far more experience than I have. An astronomy forum might be a better place to seek such experience.

 

[Donald Qualls]

Alignment is one of the big issue, alright. With a telescope, and photography though the telescope, it's a *really* big issue, because of field rotation -- even with accurate guiding, the field will appear to rotate around the guide object and you'll get trailing on objects in the edges of the field.

However, the equation has changed, a little.

I developed some images yesterday, and scanned them today, taken with a fixed camera -- "star trails" -- and it seems to me there are other limitations on exposure time, at least when shooting from my yard in town. I found the sky background at EI 800 was quite bright after only a five minute exposure at f/8 (even with the relatively poor reciprocity failure characteristics of Tri-X); five minutes at f/2.9 was enough to render the sky nearly as bright as in a normally exposed daylight image. So, I'll either need to stop down a good bit (like to f/16) or shoot at a lower EI, or both.

My next experiment with this will likely be with my Zeiss-Ikon 250/7 Ideal plate camera -- not only will I get a much larger negative, but I'll be able to use ISO 100 film (I don't have any 100 speed film on hand for 120 or 35 mm), and stop down, if I choose, to f/45. And, of course, I can shoot and develop individual frames instead of dealing with a roll format. That will be much better for trails -- the trails will be longer if I can expose for, say, 20 minutes without the sky background starting to compete with the star trails -- but also suggests that tracked images could be effective with as little as five minute exposures, which is well within the capability of a single-arm barn door mount.

So, perhaps I can get some wood and make a mount sometime soon.

Can anyone recommend a (very) reasonably priced ball head that would let me aim the camera in an arbitrary direction while the mount axis remains pointed to celestial North?

 

[Lee L]

Donald,

Are you using one of the Meade ETX models? It might be simpler and easier to use that for a drive system. I've heard they're pretty accurate drives, but haven't used one myself. If you can build a camera cradle for the ETX mount that balances around the same axis as the OTA (optical tube assembly), then you should be in pretty good shape. Keep your camera as light as possible and the setup as balanced as possible.

Another idea is to use a curved bolt drive in the same "barn-door" configuration as the Scotch drives. You just need to get threaded brass rod and curve it to the correct radius, usually around a routed wooden form. I use an old AstroKits drive that does this with good results on up to 180mm lenses and 15-20 minutes. The tangent error is eliminated in this form of drive, but you'd need better than 30 degrees of arc to get to two hours. Putting a tripod head on the end of an axle that is polar aligned and driven by a belt & disk has also been done, but can get bulky, and requires accurate machining and motor speed.

If you have a library nearby with back issues of Sky & Telescope, you can look through them for ideas. They used to have a column called Gleanings for ATMs (amateur telescope makers) that you could browse for ideas.

You might find that you can't get to 2 hours very effectively anywhere near where you live, and with the newer films with lower reciprocity failure, exposures of that length will almost certainly be unnecessary, or will be overwhelmed by skyglow. I'd expect that you'd need to be under very dry clear desert skies in the SW of the US or some other very remote location to get away with a two hour exposure.

Even a well constructed commercial drive with a polar alignment scope costing $800 only promises 35 minutes with a 135mm lens on 35mm film. So you might be asking a bit much unless you can DIY with very high accuracy. http://www.sciencecenter.net/hutech/kenko/skymemo/index.htm

I've had excellent luck with an unguided Synta drive, available from Orion as the Atlas mount with lenses up to 100mm and 20 minutes (and probably longer if skies permitted). Nothing but pinpoint stars across the field.

Oh, yeah, I assume you already know this, but use a camera with mechanical time exposure shutter operation, and pull batteries if you have LEDs in the finder that can fog the film.

As I wrote this, you posted again, and I see you found the skyglow problem.

The field rotation is not an issue with a properly polar aligned and driven equatorial mount, but is with an alt-az mount.

Read Robert Reeves' site for recommendations on astro films.
http://www.robertreeves.com/filmtest.htm
It's a little out of date, but worth a look.

Hope this helps,
Lee

 

[NikoSperi]

Wow. I salute your courage if you undertake this endeavour. Precise polar alignment (or lack thereof) will bring you crashing back to earth, even if, and it's a really big one) you manage to DIY a clock drive sufficiently precise to track at 135mm.

Even a $10,000 Astrophysics GTO requires guiding - i.e. some feedback mechanism to iron out the tracking irregularities inherent in any drive mechanism.

The Jerry Lodriguss reference is a good one to get feel for the degrees of precision required for photographic quality guiding.

For your sky glow issue, as you're using B&W film, one thing that will help you keep it down (depending on the type of illumination causing the glow) is to use a red 25 filter. If you want to invest a little to get something specific astro, then look into a night sky Hydrogen alpha filter (not to be confused with the solar hydrogen alpha daylight filter used to view eclipses etc.) . With that on your lens, you no longer care about moon/skyglow... but your exposures will get long.

Good luck!

 

[SteveGangi]

Any drive you choose will still need to be monitored and corrections will need to be made. Have you thought about piggybacking the camera on an existing equatorial mount and scope? That way you could use the scope for guiding and checking. With a Scotch drive, you would still need a finder scope or a small telescope for alignment and tracking, and it all might cost more than a decent second hand setup.

 

[Donald Qualls]

What I've found so far is that five minutes is near the practical limit with Tri-X pushed to EI 800. I need something that will put the sky glow farther down the reciprocity curve than the stars to keep them separated; I intend to try on the next clear night with Fomapan 100 through an f/4.5, which ought to let me use exposures up to 20 minutes without sky glow being too intrusive (though ten minutes would be more comfortable in terms of trying to print down the sky without losing the fainter stars, at least for trails). Of course, I can also stop down, but that will tend to lose the fainter stars due to reduced aperture. Also, the night I took the test shots (including the one I attached a couple replies back) might not have been the clearest in terms of transparency, nor was it much more than an hour past dusk at the time I made those shots; I can probably do a bit better on the clearest of nights by staying up quite a bit later.

I was rather surprised at how faint the stars can be and still be visible; with a 50 mm f/2.9, I was recording stars in a trail image that I have trouble seeing with naked eye. A tracked image should pick up faint stars better (I expect) since they will burn into the same point on the film rather than spreading along a trail.

However -- with the practical upper limit still under 20 minutes, I can pretty readily get away with a simple barn door for now.

For longer tracking, the mention of a curved bolt made me wonder -- if the bolt is curved, doesn't it tend to move (something) around an awful lot as it rotates? Oh, no, I guess one would turn the nut in this case. But one could also use a threaded rod as a worm to drive a gear made with JB Weld on the edge of a disk -- I've heard of folks doing this to drive telescopes, and as long as the radius is right it works well; one can, in fact, bend the rod around the edge of the gear to get improved engagement, as long as the radius of bend is long and the rod isn't permanently deformed. This one, in fact, might be workable from the (slightly adjustable) clock drive motor I have, which runs on AA batteries.

Bottom line at present: I"ll start with a manual barn door and short exposures and see what I can get. I'd like to be able to do things like make printable images of objects like the Andromeda galaxy (M31), Orion Nebula, and the Pleiades as well as the Moon (much easier, though, since it's a short exposure sunlit subject). Longer term, it sounds if it would be less work to recalibrate my Meade drive (it's a Meade 4400, a cheap consumer "Christmas special" telescope from shortly before the computer drives came out in that class) and make a platform to replace the telescope tube.

Eventually, I'd like to make an adapter of some kind and use my telescopes for this, but the Meade's mount doesn't do well handling the extra mass of a camera on the focuser, and my 8" is on a Dobsonian mount, so I'd need a complete mount rebuild to use it for tracking. Rich field photography works better with much shorter focal lengths, anyway...

 

[SteveGangi]

The 8 inch Dob may work. Build an "undercarriage" for it and tilt it according to your latitude. In short, make it work more like a fork equatorial.

 

[Donald Qualls]

The 8 inch Dob may work. Build an "undercarriage" for it and tilt it according to your latitude. In short, make it work more like a fork equatorial.

I've thought about that, just to make it easier to follow objects when observing visually -- but even here in the South, 36 degrees of tilt is enough to put serious stress on the base bearing, which will most likely make it very hard to get smooth movement in right ascension. On top of that, most of the weight of the tube will then be on one of the two bearing pads on each side of the mount, making the declination axis stiff as well (though that's less of an issue). Finally, there's still no drive, though that could be managed by inserting a disk with another teflon/formica bearing in between the base plate and existing box, driven by a variable speed motor mounted in the box and a belt around the intermediate disk.

My dob isn't a low-slung ultralight, I'm afraid; the balance point is less than a foot from the middle of the five foot tube, which is made of steel furnace duct. It was easy, and it was cheap, and five feet doesn't weigh that much (it's no big deal to put in and out of my van) -- but it doesn't balance six inches from the mirror like a truss tube ultralight might. That gives a rather long lever arm -- my base would need an extended and very stiff plate on the ground just to keep the whole thing from tipping, and the forces on the R.A. bearing might well be too much for teflon...

However, it isn't out of the question to design a fork mount that would use the same tube bearings I use with my rocker box. That's a long term project, though, that will take considerable thought, time, and probably money (and money is very tight right now).

 

[nick mulder]

I'm quite keen to build a scotch mount myself - I happened along this website:

http://www.astunit.com/tonkinsastro/atm/projects/scotch.htm#Photo

... when googling 'CMOS 4060' for another photography project I'm upto (time-lapse drive for a 16mm cine cam)

It got me thinking and I have a question for the astro types:

could you have two or more of these mounted on top of each other so when the top one (or bottom one, whatever) completes its full action it triggers the next one and so on ...

the first would be aligned with the pole for the time it started and it follows that my question is:

>> if I knew my mounts wound down after 60 mins could I align the next in the stack for polar alignment an hour ahead ? ... and the next two hours ahead etc ??

Actually, i can see it working fine, just need a bit of foresight and i have starry night pro (http://www.starrynight.com/) here on the powerbook to help me with that, but has it been tried before ?? easy ?? obvious issues i'm thinking of ? weight ?

We have good dark nights just out of town here but i'd try just two for starters, very keen to try it with cine film, seeing the earth rotate around the stars ...


(;
cheers,
nick

 

[Donald Miller]

I have a 5 inch celestron Schmidt Cassegrain with equatorial mount. I have never used it for astro photography. I agree with those who indicate the problems with precise alignment.

There is a filter (not red 25 and fairly expensive) for the sky glow problems that you mention.

A 135 mm lens is probably going to include a fairly substantial chunk of sky and it will be pretty slow for night sky photography. I would want to be up in the 1000 mm+ area and have the capability to capture more light.

 

[NikoSperi]

There is a filter (not red 25 and fairly expensive) for the sky glow problems that you mention.

A 135 mm lens is probably going to include a fairly substantial chunk of sky and it will be pretty slow for night sky photography. I would want to be up in the 1000 mm+ area and have the capability to capture more light.

Actually, a plain Jane red 25 filter will do a world of good for B&W astro in cutting out skyglow. Worth using. Specific astro filters for camera lenses are expensive as they need to be cut to front filter size, rather than the standard 1.25" for eyepiece or 2"...

Speed is not dependant on focal length. In fact in terms of astro, it is absolute aperture that is king. Wide field astro is alot of fun and alot easier to track. 1000mm and up means your guiding and alignment woes are going to be multiplied. Guiding at 1600mm on my (reduced) LX200 is practically beyond the mechanical capabilities of the fork mount and motors.

Start with short focal lengths - trust me!

 

[Lee L]

Michael Covington suggests the use of a didymium filter called a Hoya Color Intensifier to kill the sodium emission lines from high and low pressure sodium lamps. This is often called a Redhancement filter, and is available in the Tiffen and B+W lines as well. It passes much more light than a Wratten 25 or 23A. Search on Covington didymium in google to find sample shots. It's much cheaper than a typical LPR (light pollution reduction) filter made specifically for astrophotography. Unfortunately, lighting designers and mfgrs are now promoting full spectrum metal halide lamps that these filters don't work on as well. The metal halides are also less efficient and their blue spectral content scatters much more, contributing up to 6 times as much skyglow as the same wattage high pressure sodium lamp. (Why is the sky blue?)

Niko is correct about aperture area (square millimeters) alone being what counts in astrophotography, and a 135mm f:2 or f:2.8 can be a nice astrophoto lens. Go for primes, not zooms. Going anwhere near 1 meter or longer focal lengths will require a very solid mount and long guiding sessions staring through a high magnification optic. That length is totally unsuitable for a barn door drive. Most folks using longer focal lengths (1 meter or longer) are now going to computer guiding using a CCD chip and software that keeps a star at sub-pixel registration by making adjustments to tracking motors. This is definitely not what Donald is asking for.

Nice fast lenses from 35mm to 135mm or maybe 200mm with a solid and reasonably accurate barn door drive can get you a lot of beautiful shots. I have a curved bolt barn door drive (yes, you turn the "nut", actually a gear with a central 1/4 20 thread, not the bolt) that I've gotten consistently good results from with up to 180mm at up to 12 minutes.

Film is also important. You need sensitivity at 650nm for red nebulae, and low reciprocity failure as well. Most of the really great astrophoto films are out of production, but Provia 400F and Elite Chrome / Ektachrome 200 do reasonably well. You can find the spectral sensitivity of most films in the tech data sheets, but you have to watch for reciprocity failure swamping red sensitivity. Robert Reeves suggests HP5+ for B&W astrophotography. In color film, you need to watch for differential rates of reciprocity failure in the different emulsion layers, which can cause radical color shifts. Ever shoot Kodachrome 200 at 1 sec or longer? YUCK!

Film recommendations (all somewhat out of date, but the best info to my knowledge) at:
http://www.covingtoninnovations.com/astro/films.html
http://www.robertreeves.com/filmtest.htm

The P factor in Reeves' tables is the value for the Schwarzschild exponent used in determining reciprocity adjustments, and is an indicator of the rate of reciprocity failure.

For construction ideas go to "Gleanings for ATMs", a regular column in older Sky & Telescope magazines. Probably at a library near you. Lots of tangent drive designs, melted plastic teeth formed with a heated threaded rod driven with a worm gear made from threaded rod, Poncet and D'Automne(sp?) platforms for "equatorializing" your dob mount, etc.

Lee