Telescopes‎ > ‎

28 inch f4 alt-az photo and design page

Two end-to-end crossflow fans (also called squirrel cage fans) blow a flat stream of air (illustrated by the blue arrows) across the front surface of the mirror, which exits out the exhaust door to minimize turbulence. This has proven to be much more effective at quickly cooling the mirror, and doing so evenly, than the two traditional muffin fans attached to the back of the mirror cell. By the way, the primary mirror was figured by Steve Kennedy of Kennedy Optics and is the best optic I've ever had. It's from the first batch of 28 inch f/4 mirrors he produced in 2004 and he was a delight to deal with.
This shows the normal storage and transportation configuration of the scope. Note that the cage is secured by three red bungee cords so it can't be bumped off.


The closeup photo below shows the current configuration of the azimuth drive. It's essentially a big timing belt - the belt is wrapped around the circular edge of the azimuth ground board with the teeth facing inward while also engaged by the sprocket on the end of the azimuth drive shaft. This arrangement essentially eliminates backlash and is silky smooth.  
The large disc at the top of the drive shaft is part of the slip clutch - look closely and you'll see a smaller disc underneath it. Ebony star and Teflon are between the two discs, and tension is supplies by the spring at the top of the drive shaft. 
The edge of the azimuth ground board is edged with non-slip tape (essentially sandpaper) to make sure the belt doesn't slip. There's about 100 pounds of tension in the belt so this might seem like overkill, but depending on how tight the clutch is set, it can actually slip. On windy nights it's helpful to tighten the clutches so the wind can't move the scope off target, so this turns out to be important.
I've found the azimuth drive belt and configuration to be the most difficult part of the scope to perfect. The key turns out to be a non-stretch belt. I use a Gates Polychain belt (which is reinforced with carbon fiber) and matching sprocket, which turns out to be one of the more expensive options. At this point the azimuth drive is not only smooth and reliable, but it's also quite robust. The only detectable backlash in the system now is in the servo motors themselves and is quite small.
By the way, it might appear from the photo that the drive sprocket is also in contact with the perimeter of the azimuth ground board. There's actually about .125 inch clearance.
The next photo shows the previous arrangement, with the azimuth belt (it's the same one - the teeth side of the belt is blue) wrapped around the azimuth ground board with the teeth facing outward like a huge gear.
This worked ok but had the drawback of not being precise enough so the teeth would engage the drive shaft pulley with EXACTLY the same tension all the way around. It only took a thousandth or two make a difference! I tried floating the drive shaft on springs so it would engage the teeth more uniformly and although an improvement it wasn't as good as I hoped. The black thing on top of the azimuth drive shaft is the azimuth scope encoder.
A close up view of the mirror cell collimation bolts design. The important point about the mirror cell - a design I copied from Dan Gray - is that the collimation bolts move the entire mirror cell and mirror together, not just the mirror. This eliminates the build up of edge stress that moving only the mirror can cause. The photo shows that the collimation bolt is fixed between two brackets and the mirror cell frame moves between them. A rounded "nut" made of Delrin plastic has a threaded hole through it that the collimation bolts are threaded through and provide the small bit of compliance needed so the cell can be easily adjusted. You can see the edge of the Delrin nut as the small bit of white sticking out from the end of the mirror cell.
This view shows one of the triangle support arms and the main reason the mirror cell is low profile - the arm pivots on the bolt that goes through its center. The support triangles end up about .125 inch above the mirror cell frame.
The Sidereal Technology (SiTech) controller box, once aligned, allows the scope to be moved manually like a Dob and still know where it's pointing.  This lets the scope be pushed around like a Dob and yet track all night. It can also be used with an Argo Navis or Sky Commander for goto's, or for extremely accurate goto with a laptop. SiTech also sells compact clutch boxes allowing them to be neatly tucked away, not the big homemade discs seen on my scope. I'm a huge fan not only of the SiTech system but of Dan Gray too!
I had originally design my scope to ride on an equatorial platform but the SiTech "DragNtrack" mode allows me to use the scope as if it were on a platform that never needs to be reset. As a visual observer I'm not bothered by field rotation so SiTech was the ideal solution for this scope. Plus, it's nice not to have a huge equatorial platform to deal with.
Underneath the focuser is a filter wheel from Aurora Precision that has room for four 2 inch filters along an empty space for non-filtered viewing. This is a great way to use filters because they're always right there - just turn the wheel to the filter of choice. However, it's not good for blinking because each filter changes focus a little. I have a cover (made from ABS plastic) over the wheel so the filters aren't exposed to the night air to prevent them from dewing over. in six years of use they have yet to fog over.  
This upside down view of the cage shows the tennis racquet grips I installed on the cage tubes. They make a comfortable, secure grip on even the coldest night.


This is a shot of the 28 inch from the top of my eight foot orchard ladder. The only functional thing to note is that this is how the scope is situated during the day at a star party so any strong wind will only turn it in azimuth rather than blow it over. Other than that, this is just a cool photo!
This is jpg version of the actual design file which is available at the bottom of this page as a download in either .tcw or .dxf formats. The files are the same, only the formats are different, but it's important to note that the actual scope is about 90% accurate to the drawing. I made a few changes on the fly and have not updated the file, but a careful examination of the photos here will show the differences, which are mostly in the azimuth ground board.
This was the first time I'd used a 2D CAD program to design anything and I'm thrilled that about the results, mostly because everything fit together on the first attempt.
This is the map of mirror distortion from PLOP for the 18 point mirror cell.
A close up of the mirror cell and frame from the 2D files below.
I found this especially instructive. This is a close up of the altitude trunnions and I include it here as an illustration of the amount of dimensions needed for a local machine shop to cut them out with a water jet machine. I think it took me three tries to get them enough measurements.
This photo shows the mirror cell collimation bolt along with its matching Delrin nut. You can see that it's not rounded much, just enough so it can slightly rotate inside the ends of the mirror cell bracket to allow the mirror cell to move when the collimation bolts are turned. There's no looseness after 6 years of use.
The truss tube clamps are all from Aurora Precision. They're available in either Delrin or aluminum, and I've tried both. I prefer the aluminum version because they produce a tighter clamping force on the truss tubes. Delrin works best on smaller scopes.


A view of the back end shows the primary mirror baffle and the 25 inch cutout to allow air flow. The hole is covered by window screening to keep out bugs, which so far has really worked well. Note the three collimation knobs.
The rocker rides on three conical wheels, turned down from 95 durometer urethane skateboard wheels. Each wheel has two sets of abec 5 bearings and uses a precision stainless steel shoulder bolt as an axle - very smooth. It's important the wheels are conical - ie their vertex is at the azimuth center of rotation - so there's no binding against the bottom of the rocker. Note how the wheel bracket is shimmed so the wheel rides flat on the bottom of the rocker.


Pay no attention to my dirty diagonal mirror, its the short piece of 5 inch PVC pipe on the left of the photo that's interesting. Four slots were cut in the PVC so it could be press fit over the spider arms, effectively reducing their length to vibrate - it really helps on a windy night.

Ready for action at the 2010 Golden State Star Party near Adin, California! The scope breaks down and fits in the back of the van along with all my other observing and camping stuff. Check out the teardown video page to see how this is done.
Howard Banich,
Oct 10, 2010, 2:40 PM
Howard Banich,
Oct 10, 2010, 2:40 PM