A 20" Dob you can lift over your head - for $1500
Introduction
This 20" f/3.3 was originally to basically take everything about the 22" that I'm also doing and shrink it down a bit. Thinner mirror, slightly smaller aperture, slightly less crazy focal ratio, about the same physical height. Why am I doing this and a 22"? Well, I need more practice if I'm to break the 30" if not 40" barrier someday. Hence two projects. However, my ambition has grown somewhat and I figured that this should not just be another neat scope but a proof of concept for a much bigger idea. You see...
The 20" f/3 is meant to be copied. With a plate glass, slumped meniscus mirror, a readily-available 4" secondary, and a low-cost construction, it is meant to cost only $1200 to construct including silvering, etc. - that is the going rate for a commercial 12" these days! This is a daunting project for anyone, let alone a first-timer - but in theory, anyone could make one of these telescopes.
To make copying if not surpassing me even easier, as well as prove the effectiveness of this design, I will be making detailed posts explaining the work involved. For general information on telescope making and mirror making, I recommend checking out the links on my homepage, in particular Mel Bartels' website.
Why the heck should I make this scope?
Commercial scopes are, quite frankly, oppressive. Today's market is a mix of Chinese-made cheap scopes which either straight up suck or just don't last over time - thin-walled steel tubes dent, particle board rots, cheap servo motors and computer boards in plastic housings won't stand the test of time. Now, there's nothing wrong with buying a cheap Chinese Dob for yourself, and they're easy enough to modify and upgrade. I tell people to buy them all the time in my reviews and ramblings. However, with the recent rise in prices, a 10" with some accessories will now set you back a thousand bucks, which is straight up out of the budget for many people, and even a 6" is nearing $500. Amateur telescope making is economically viable again, and I will talk about that concept in general somewhere else at another time. But I came to the realization that we can do better than just making a 10" for less than $1000.
The Big Three Chinese companies (Synta, GSO, and JOC) don't really make anything above 16" and their offerings above 12" are for the most part horrific, cumbersome beasts with all sorts of mechanical and likely optical trouble that I suspect they sell just so they can say they offer big Dobs. And even those cost several thousand dollars, in addition to likely leading to several thousand dollars in medical bills when you inevitably break your back moving one around. All for a 14" or 16".
Anything that isn't one of these Chinese scopes is going to be a premium luxo-Dob that costs thousands upon thousands of dollars for a base model, and even these are getting less common. Teeter's Telescopes is closing up. Obsession only offers their rather-janky UltraCompacts now, and these are set to be discontinued, too. And looking at the prices of these premium scopes is likely to send you into the early stages of cardiac arrest.
So what if I told you that not only is a "cheap" 20" scope possible, but that you can make it yourself, it'll cost you less than $2,000 to set up everything, and you need not possess any sophisticated tools, experience, or a large vehicle? This telescope can fit in a compact car and you could probably lift it over your head if you feel so inclined. Why would you ever make a traditional Obsession-style scope or buy an 8" Schmidt-Cassegrain to meet your needs? Heck, why would you ever make a standard 6" f/8, or 8" f/6, or maybe 10" f/6 mirror and telescope that will probably cost more than 500 bucks and leave you feeling confused and underwhelmed at the end? Let's do a 20"!
Yes, you'll need to grind the mirror. Yes, you'll need some power tools for the structure. Yes, it is harder than IKEA furniture. It's a challenge - but a rewarding one at that.
Your average Joe Telescope Maker is very likely to disagree with me and you. They will either tell you to simply suck it up and buy a telescope, or at least buy the mirror (which won't save money on anything small and quickly gets expensive on anything over 12-14 inches), and that grinding anything thinner than a 1:6 or maybe 1:12 thickness ratio is hard and thinner than 1:16 or 1:18 is straight-up impossible.
You will be told that you will need to spend hundreds of dollars on making a useless 6" f/8 first, followed by perhaps a 12" f/5 that's probably obnoxiously thick and expensive Pyrex too, that you need to understand Foucault readings and Couder masks, and essentially to stay in your lane and do the exact same things people have been doing since the 1920s. That's right, you should follow the same telescope making techniques as in the 1920s. Some people may even tell you that an amateur mirror over 12" or faster than f/4 is impossible, and you should just buy the mirror and abandon all hope. A 20" as a first mirror is likely to get you laughed out of the room. Ignore the naysayers. Telescope making isn't supposed to be competitive, nor should it be about taking the easy route; it's about innovation and trial and error. To follow the instructions may guarantee success - but what's the fun in that?
With newer testing methods, a positive attitude, and hard work, it is possible to make an acceptable 20" mirror on the first try. People have done it. And with silvering, you can always strip the coating, try again, and recoat for mere tens of dollars and hours of turnaround time instead of hundreds or thousands of dollars, perilous shipping and weeks of waiting. As for thickness, Mel Bartels, Tom Otvos, Rob Brown, and the Oregon Scope Werks are leading the way in making mirrors that are getting slimmer and lighter faster than the most dedicated diet fanatic. A 20" x 0.555" thick meniscus is thin, but it's not pushing the absolute limit and you can always go thicker if you wish.
I've chosen f/3.3 for the focal ratio to avoid ladders, avoid a secondary mirror larger than 4", still work with cheaper coma correctors, and keep the whole scope short - but you can go as fast as f/2.6 or as slow as f/4 if you choose. A meniscus slower than f/4 won't be as stiff, so don't go any longer than that. Figuring gets dramatically more difficult as you go faster and you'll need nicer accessories and a big secondary mirror, too. So for best results keep it between f/3.3 and f/4 - and keep in mind BVCTek also already has a mold for f/3.3 which can save you quite a few bucks and some waiting!
There are still unanswered questions. Among them:
Can I cram this thing into my girlfriend's Mazda Miata? If not, what is the smallest vehicle I can cram it into? I think I can fit it in the Miata.
How much exactly will this weigh? The mirror is 15 pounds. I've calculated a ballpark estimate of 40 pounds for the whole scope.
What kind of edge support is best for these mirrors? I'm planning on no adjustments to the mirror cell and a hexapod truss.
Will figuring require any fancy pitch laps beyond maybe a spider lap? No clue, hopefully not but we'll see.
Can I add GoTo to this thing economically?
Who else wants to do this? Is this an idea that I can bring to the masses?
Cost Breakdown
20" x 0.555" plate glass blank from BVCTek, slumped to f/3.3 - $200 + $75 shipping to the USA - yes, the 1.1" thick blank is listed at $320, yes you could and probably should go with 1.1" thick, but then the scope is heavier and costs more. Ditto for BVC vs. plate glass, it's objectively a superior material (apart from scratching more easily) but same story.
104mm secondary mirror from GSO - $170 (okay, I got one for less, but you get the point). Can also get one from Antares with better quality control for $280-$500 if you feel so inclined or GSO is out of stock again. Plus there's always used.
Grits, pads, and pitch from GotGrit - $180 - pitch is reusable though, so you can sell it to your friends when you've converted them to the cult of telescope making, or just use it to make more mirrors. You need 2 kilos of pitch, less than a hundred polishing pads, and only need grits starting with #120. The full kit GotGrit sells is more than you actually need; 1 pound of cerium oxide is also more than enough for polishing.
Dental stone off eBay for tool - $36 - I'm not listing a link because it'll expire, but just search "dental stone". I think I usually use Type IV, you can get it in any color of the rainbow but it will just look like bubblegum-flavored vomit, and you want at least 15 pounds of it. You will need to pour a second tool to grind the back smooth to #220. Also, having extra is good if you screw up something. I ended up mixing my dental plaster with Perlite to make my main tool a little lighter, this is about $20.
Hex tiles - $13 - no, you do not glue them on, just bake them into the tool. Epoxy will not hold and costs a ton for a mirror this size, and it's disgusting and smells bad. If some of the tiles are partially submerged in the tool and useless because you baked them in, oh well. Better than half of them falling off during fine grinding and having to pour a new tool and start over.
Silvering chemicals, Angel Guard, silver cleanup supplies from Angel Gilding - $155
Sprayers for silver - $20 or less from any store
Distilled water, cleaning supplies, talc, buckets, etc. for silvering - ~$30
Anti tarnish cloth - $15
Poles - $81 from DX Engineering for the hexapod design I chose
Focuser - Basically $0 - make one yourself out of wood and some other stuff. Or buy an HC2, or wait for Moonlite to get more aluminum and make a manual focuser again, or buy a GSO, or buy something used, or steal one off another scope. Do not use a Feathertouch as its height necessitates a larger secondary mirror (4.5" or 5") which will drive up the price and make the upper end of the scope heavier too.
Wood, nuts, bolts, screws, wires, glue, related hardware - ~$120
QuInsight - $60
Lycra/Spandex fabric for a shroud (optional): $20
Pink insulation foam to use as a test stand: $10
Ronchi tester: $10 to print a Ronchi screen onto transparency film and get some kind of a light for it. Stick it on a tripod or something and you've got a precision tester. Some guy attached one to his glasses. I don't care how you do it, it works. Use Mel Bartels' matching Ronchi test and then star test in the scope. If you really don't trust a piece of plastic with lines on it, you could make a Bath interferometer and use that, but that costs money and requires more brain cells than I have.
Total cost: theoretically ~$1195 USD. In practice it's a bit more than that. The Heim joints for the hexapod I'm using are pricey, as is the Alucobond for the structure. Strictly speaking, though, you don't need those. A regular wood structure would be fine. If you were to really scrounge - get your sprayers for free, get some scrap tubing for poles, find a secondary mirror used, use some scrap fabric, steal pitch from your local astronomy club, and share the silver with others to coat their mirrors - you could probably get the cost down to $600-700 or so. But that is a best-case scenario, and if you budget $1500 and end up spending less you've got more money to buy additions to the scope. So I'd plan on spending around $1500.
For comparison, a 22" Obsession UC has similar capabilities (the silver is more reflective), costs $11,195 USD (yes, a whole extra $10,000), takes months to deliver, and has numerous quality control problems. An Apertura AD12, which has a base that weighs by itself nearly entirely what this telescope will in total, costs $1299 and is heavier, bulkier, about as tall, and has less than 1/3 of the light gathering . And a bare-bones 8" Schmidt-Cassegrain is now $1600, and this thing can be stopped down to an unobstructed 8" and walk all over it. Yes, we need accessories for this scope to work well, but so do almost any scopes you can buy.
This telescope is designed around the Baader MPCC to preserve the native f/3.3 focal ratio and save on cost. To achieve sharp images I will make the mirror slightly hyperbolic/overcorrected. You may not wish to do this, in which case an Explore Scientific HR Coma Corrector or Tele-Vue Paracorr is for you, albeit more pricey and annoying to get used to tuning etc. The MPCC just screws onto your eyepiece or 1.25" adapter. It's probably also possible to adapt the GSO coma corrector to this setup, but I'm not a masochist.
If you've never had a telescope before, or at least not a big one, here's the minimum of what you'd expect to spend on accessories, assuming you buy new, which you should avoid if you can help it:
MPCC - $217
Astro-Tech 16mm UWA - $120
Astro-Tech 10mm UWA - $120
GSO 2.5x Barlow - $50 (not needed, you could get a nicer Barlow or a good high-power eyepiece, but this is the budget solution for high magnification)
You'll probably want 100-degree eyepieces, more of them, and a nebula filter, but these will suffice if budget is a concern.
That's $507 in accessories. Okay, that's $1775 for a complete 20" scope with some nice eyepieces, a good finder, a weight equivalent to a typical 8", a height equivalent to a typical 12", and it can fit in a pretty compact car. Not too bad, isn't it?
You can obviously spend more, and there's nothing wrong with spending more. Get a BVC blank, an oversized 4.5" secondary spec'd to 1/30 wave flatness, some 100-degree eyepieces, a UHC nebula filter, a used dual-speed Moonlite, some electronics for fans and dew control, heck, add GoTo or digital setting circles. You can easily sink double or triple the initial budget I've outlined into the scope. But the bare-bones system I've outlined is all you really need, and will show you 95% of what all those upgrades will get you (apart from the nebula filter - do go get one of those).
The Primary Mirror Blank
The primary mirror is, objectively, the most important part of the telescope. It is also the costliest and most difficult part to make. You start with a raw disk of glass. In this case, it is a meniscus - that is, convex to the same degree as it is concave on the opposite surface. Meniscus mirrors cool down more evenly, and have less mass, than the equivalent flat-backed mirror, which in this case would be about an inch thick. Thin, but not borderline nonexistent. The meniscus mirror is also somewhat easier to support in the telescope than a flat-backed one and we can get away with an 18-point cell.
The primary mirror blank for this telescope is 0.555" thick plate glass from BVCTek, slumped in a kiln. It weighs 15 pounds.
Setting up for Grinding
For grinding a meniscus mirror, the first thing we will need to do is, counterintuitively, to smooth out the back. Even a well-slumped blank will have irregularities and we want to prevent the Twyman Effect as much as possible. Thus, we will pour two tools: one for the back, and one for the front. These are dental plaster with the tiles baked in - no glue here.To make a mold, the blank is wrapped in Saran wrap as tightly as possible and aluminum flashing is wrapped around it, secured with tape. The concave tool is poured first, with the mirror resting on it to stay level while the convex tool is poured after. Any irregularities in the tools' shapes can be smoothed out with a file, sandpaper, or a blade. Pouring the dental plaster is pretty easy - I mix it with a drill in a paint bucket and add water to the plaster slowly to prevent spillage.
When the plaster starts getting hard, it is important to immediately remove it from the mirror blank. Dental plaster heats up rapidly during the curing process and this could actually crack the blank. To make the tool for the front of the mirror lighter in weight for tool-on-top work, the convex tool for the front is 50% dental plaster and 50% Vermiculite.
The concave tool also serves as a support for the mirror when working tool-on-top.
Rough Grinding
For rough grinding, we start with #120 grit, then moving on to #220. Fine grinding constitutes finer particles. In rough grinding, we are mainly trying to smooth out irregularities in both sides of the meniscus. The back can be ground to #220 and left there, any smoother is unnecessary. The grit is sprinkled on with a spoon and water is sprayed on. Even #120 is rather loud, and it's somewhat messy. These grits also have to be rinsed off outdoors and cannot be poured down a drain lest they clog a pipe.
Rough grinding the back took around 4 hours (2 with #120 and 2 with #220), and I left a few spots untouched. The blank has a general slight saddle shape when it comes out of the kiln.
In retrospect I should've started with #80. Grinding with #120 took about 8 hours; #220 was another 6 hours. This was mainly to get down to the annoying low spots on the blank. #80 would've been a lot faster.
Fine Grinding
For fine grinding, we now focus solely on the front optical surface of the meniscus, and can pre-mix the grits in a slurry. Keeping clean is easier - you have to avoid contamination between grits, yes, but final rinsing in the sink is now acceptable.
Fine grinding with 25 micron took around 1 hour, as did 9 micron. I am surprised I was able to do 9 micron with no scratches.
Polishing with Pads
For initial polishing, we will use optical polishing pads meant for eyeglasses. These leave a rough surface but they work faster. You simply need to stick them on the tool. I heat mine with a heat gun and press them on afterwards to ensure they do not fall off during polishing.
For pad polishing, we simply squirt cerium oxide slurry onto the pads very occasionally and otherwise work just like fine grinding.
Polishing with Pitch
Initial Figuring
The Secondary Mirror
The secondary mirror in a 20" f/3.3 can be a 4" minor axis mirror, which for whatever reason is drastically cheaper than anything even a few millimeters larger. You can buy one of these new. I scavenged a 4.25" quartz secondary for a low price from what was left of Clyde Bonne's 20" Mersenne telescope.
To avoid the annoying diffraction spikes present in many reflectors, as well as save on cost and complexity, the secondary will be mounted to a wire spider.
Building the Upper Cage
Building the Lower Tube Assembly
The LTA is pretty simple in a hexapod truss. The actual "mirror box" is a circle with sides cut off for mounting the bearings, made out of one layer of foam with Alucobond sandwiching it. The mirror cell is suspended from the back.