29 months after I first put grit to glass, I've put my mirror on the test stand for the first time. It's not as good as I'd hoped.
To recap: I last touched this project 10 months ago. Since then, the mirror and the polishing lap have been sealed into plastic containers, and left to rot in the loft while life overwhelmed me. But today, I got in my car and hauled everything down to Joburg and the ATM class. When I unpacked everything, I found the mirror caked with dry cerium, and the pitch lap had slumped down into a very flat mushroom shape, and gone rock hard (Also, some mildew had taken root in the pieces of damp fabric that lined the containers to provide padding and keep everything from drying out).
The first order of the day was to check my progress on the mirror. Last time I checked, I was pretty sure I had successfully finished polishing the mirror, so I wanted to test it to see what the next step would be. I washed it, rinsed it, dried it, left it to cool to ambient temperature, and popped it on the test stand. I then spent about a half hour struggling to get the foucault tester to work... turns out there's an art to it, and that one does not simply apply book knowledge to do a quick Foucault Test. It takes a bit of skill and finesse.
There was another tester set up, this one with a CCD camera up against the knife edge. I brought my mirror there, and let one of the resident experts set it up for me, and there on the monitor was my mirror. It was obviously not a sphere (In the Foucault Test, a perfectly spherical mirror will show up as a smooth featureless white disk. A fully corrected mirror, with a paraboloidal figure, appears to have a shape closer to a donut. Or a very softly rounded impact crater), which was slightly disappointing. More experts came to have a look, until Chris Stewart gave his diagnosis: "Your mirror is filthy. You must have wiped it after washing it. Never do this. Also, you have a turned down edge. Also, it's an oblate spheroid".
What does that mean? Well a turned down edge is exactly what it sounds like: the outer edge of the mirror is flattened, and does not follow the rest of the curve. To fix this, the entire surface of the mirror needs to be polished away slightly, sinking the curve a little deeper until it extends all the way to the edge. This is my first task, and the stroke to correct it is short (1/4 diameter) center-over-center strokes with the mirror on top. This is to be done briefly, and the mirror should then be tested again. Repeat until the edge is fixed.
After this, we have to correct the overall shape. But what is an oblate spheroid? Take a sphere, and spin it hard enough for centrifugal forces to distort it. You now have an oblate spheroid. Planet Earth is an oblate spheroid, because it spins around its axis. Slice off a thin sliver, from around one of the poles, and you have the shape of my mirror. How do we deal with this? Put simply, certain changes have to be made to a spherical mirror to turn it into a paraboloid. If the exact opposite of those changes were to be applied, you get an oblate spheroid. So all I have to do is apply the standard parabolising stroke, and my mirror will gradually acquire a spherical shape, and then reach the desired paraboloid. Easy! (It's a third of a diameter wide 'W' stroke, in case you're wondering)
But before I could do any of this, I had to create a new lap. Step one is to remove the petrified pitch from the glass base (Which, you'll remember, began life as the grinding tool). This was done by violently chopping at it with the edge of a putty knife. This causes the pitch to chip away in shards, like very soft glass. Pro tip: Do this at arms length, while holding your work deep inside a rubbish bin. You should probably wear safety goggles too, although squinting slightly seemed to work just as well... As I type this, my hands are still speckled with tiny crumbs of pitch that will probably stay with me for a few days. It's messy stuff.
But once most of the pitch had been knocked off, leaving just a sticky film and a few black specks, I took it back inside and (with the kind assistance of Dave Hughes - thank you Dave!) made a new lap. This new lap, incidentally, sings. Every few strokes, it emits a soft whistling squeak, like a very boring birdsong! I'm loath to add more cerium, in case this lovely sound goes away!
So I'm going to have to finish building my own tester if I'm to make any progress, or get any work done at home. And once it's built, learn how to actually work the thing. But I'll be back to my usual schedule of attending the class every second weekend until this telescope is done. And for the third year running, I hope to have it ready for ScopeX
I figured out why I'm taking longer and longer gaps between work: My kids are getting old enough to climb into cupboards, and curious enough to dig through and dismantle everything they find. Therefore, if my mirror and supplies are to survive, they have to be packed and hidden away very thoroughly in hard-to-reach places. I'll do a few hour's polishing, stash it all together in the highest darkest recess of a cupboard, and then forget about it. It's the old "Out of sight, out of mind" principle at play. Perhaps our new home will have a room we can use as a studio or workshop, where our stuff can be left out and the door locked. Ah well.
To recap the past few work sessions: I tried to re-open the channels on my lap by warming the pitch and pressing a steel ruler into the existing channels. This caused the edges of the facets to mushroom up, so I tried a warm press to fix things. This was overdone (too hot, too much pressure, left for too long) so that I ended up with a perfectly smooth featureless lap. I then tried fixing it by cutting channels in with a craft knife, but my knife-fu was weak and I ended up chipping great divots out of the surface. So I tried the warm press - steel ruler combination again, this time paying more careful attention to the variables, and ended up with a normal-looking lap again. I put in an hour's polishing, during which the lap behaved like it was brand new: Sticking, skipping, and generally misbehaving.
Last night I returned to regular procedure: Cold press, with a thick layer of cerium oxide slurry, followed by 1/3 MOT W-strokes. That lap finally seems to be fixed again. I only put in a half hour's polishing, and did not examine the surface to see my progress - I assume that it hasn't changed much since last time - a few cm's near the edge still remaining to polish out.
On a different topic, the family and I have been looking to own a home for a number of years but have struggled to find something that's big enough, to our taste, in a suitable area, and within our budget. So now we're looking for vacant land and planning to build. Yesterday we visited a property that's a lot further out from the city that we'd previously considered. Virgin land, proper African savannah, and if you pay attention to the ground you see many different species of plant woven through the grass. And it's quiet. And it will be properly dark. So if we buy it, expect a new series of stories in this blog about the construction of my observatory!
This has nothing to do with my telescope, but it is absolutely ATM relevant: ScopeX
this year will be held on Saturday, 21 July 2012
. I'll be there to see what's new in the scene, get ideas for my 5th telescope (After I finish my 1st one, I've already decided what to build for numbers 2, 3 and 4!), shake the hands of other enthusiasts from around the country, and hopefully catch a few interesting lectures. Entrance is open to all at the Military History Museum
in Johannesburg, which charges R25 for entry.
Features of the expo include telescopes (self made and commercial), an astrophotograpy competition, talks, science shows, telescope auction, camera obscura, sundials. There's something for everybody, young and old!
When last I wrote on this blog, I had decided to store the mirror on top of the lap for 24 hours, the idea being to give an extended cold press to try and fix the 2 facets which weren't making contact with the mirror surface. Several days passed before I managed to get back to polishing, and I found that, while the contact issue had not been resolved, the gaps between the facets had narrowed noticeable. I put in an additional hour's worth of polishing and decided to repeat the experiment.
The following evening, there was still no improvement, apart from the gaps between the facets having shrunk even more. They were now only about 1mm deep, and maybe 0.5mm wide. On the advice of the ATM class, I prepared for a bit of lap maintenance, excavating a newer and deeper set of channels. I planned to do another hour's polishing before making any change, but to my dismay I found that I hadn't packed the lap/mirror combination level. As a result, the mirror had slid a few centimeters to the side before coming to rest against the side of the container, leaving a pronounced lip pressed into the pitch. Bugger.
So, to remedy this, I decided to try a warm press. Filling a basin with hot water ("hot" meaning "Slightly uncomfortable to keep my hands in there"), I soaked the lap and the mirror together. The heat softened the pitch to the point where I could easily dent it with my fingers. I pressed in some deep channels with a steel ruler, then laid out the mirror on a table and spread a nice thick layer of cerium oxide slurry. Then I put the soft warm lap face-down on the mirror, stacked a few books on top of that lot to build the pressure, and poured myself a nice hot cup of tea.
Extended tea-time over, I removed the books and had a look at the lap surface. Wow... completely smooth - I not only pressed out the ridge, but all signs of faceting. Also, a fair bit of pitch had mushroomed out the side and had to be trimmed away. I could try the process again, but I'd noticed that simply pressing channels in caused the edges of the facet to rise up. Pressing them down again would only tend to close the channels. What to do...
I decided to use one of the other classic methods: Carve the facets in with a sharp utility knife. I used the steel ruler again, just to score the surface and give myself some guidelines, then got to work digging. This turned out to be very slow work, but seemed to be going well until the surface of the pitch suddenly cracked and a huge chip the size of a 5c coin broke away. I was now right back at the beginning, with minimal channeling between facts, and a region of lap that was severely recessed away from the mirror surface. That was when I decided to put everything away, go to bed, and leave it for the experts to fix at the ATM class.
But by the next evening, I'd regained some confidence and had some time to think about the problem.
I decided to go for a warm press again. Same procedure as before, softening the pitch, smearing slurry against the mirror, steel ruler to press in more channels, stacking books... but this time I was careful to only leave it for a minute. The results were a lot better - the newly pressed in channels survived, while the nasty divot was still there but much shallower. I let everything cool down, gave it a cold press and got on with polishing. The lap was very grabby at first, just like when it was first made, but it took only a few minutes to settle. More cold pressing, more polishing, and things seem to be working as they should. The hole is still there, but steadily getting shallower and hopefully will vanish completely as the pitch continues conforming to the work.
Current status: Mirror seems to be polished out to within a few millimeters of the edge. Total time spent polishing (not including lap maintenance, pressing, etc): about 6 hours. Prognosis: Positive!
Let me say up front: Polishing is boring. There's no real sound, no tactile feedback, and the work is slow going. Unlike grinding, you can put in many hours of work and not have any obvious result. As a result, motivation slides and work slows. Since my last update I've put in another 2 hours. Even so, in that time I've learned a bit more about the process. Firstly, to my disbelief, it seems I cut corners at the beginning of the process - I have a few facets on my lap that are not making good contact. I know this because, when viewed through the back of the mirror, they're a distinctly milkier colour than the rest of the lap, suggesting that they're not quite touching the glass and trapping a thin film of cerium slurry. I guess this means more pressing is in order. So rather than my usual procedure of packing mirror and tool separately, I've packed them together, mirror on top of the tool, and will work some more tomorrow night. That gives 24 hours of cold pressing with just the weight of the mirror. We'll see what that does.
Meanwhile, the current state of the mirror is as follows: To a casual glance, the mirror appears polished with a nice transparent surface. Closer inspection reveals what looks like a faint grimy film on the surface, though, which is thinnest at the centre and gets worse towards the edges. This reveals that the polish is incomplete - a fully polished surface should be absolutely spotless. In fact, if I shine a thin beam of light onto it, the surface shouldn't be illuminated at all.
So ya. Much more work.
There are reasons for my recent silence: First, I'm in the middle of exams (one left to go, wish me luck!), and I've taken some forward steps in my career which have made me a whole lot busier. But this doesn't mean nothing has happened. I've rebuilt the polishing board, to sort out a few problems that only became apparent while I was using it, put microfaceting on the lap, and done about an hour's worth of polishing. Also, I've taken the opportunity to measure the focal length of my mirror more accurately than before (the semi-polished surface doesn't need wetting anymore when testing), to close on 950mm, which translates to a focal ration of f/6.2
The polishing board had two problems: It slipped and slid around on the workbench while polishing, and the rubber pegs (made by stacking tap washers and screwing down onto the board) were too soft and short, so that the rounded lower edge of the tool tended to deform them and then climb up over them. I solved both problems with more washers, and longer screws. The new pegs follow the exact same design as before, only they're taller and the mounting screws have much wider heads. Result is a secure grip, and much less deformation. An additional four washers were then glued to the underside of the board (one per corner), making nice grippy feet. A brief polishing session confirmed that I now have a good, workable board.
Once exams are done, I'll get back to work. But for now I'll (and you'll) just have to be patient!
Wow... Fifteen month's after I first laid hands on my glass, I have finally begun polishing. Let me say up front that so far no part of the process has been anything like I imagined. I've read the literature, I've read personal accounts like this one, and none of them conveyed the sheer diabolical nature of a brand new pitch lap at the beginning of polishing! The coefficient of friction was a meaningless term, as it switched from "superglue" to "Greased ice", sometimes several times in the course of a single stroke. Happily, this didn't last long, and things settled quite rapidly, although I think it will continue settling down for some time yet.
I arrived at the class, packing the mirror and tool in separate containers, and with a nice new polishing board. The board was made from an old square of supawoodTM with retaining pegs made from stacks of rubber washers fixed into place with screws. To keep it waterproof, I gave it several coats of eye-scorchingly bright green enamel paint which we had lying around from an earlier project. This turned out to be the first mistake, as I hadn't allowed enough time for the paint to cure - although it was dry to the touch, it remained slightly tacky, and was still wet beneath the surface. Any significant pressure or friction caused it to break free from the undercoat, and spread a thin layer of wet paint. The result is that there are now pretty little spiral ridges of recently dried paint directly underneath the tool. I'll have to fix this.
The thing is, after 15 months of grinding, I was pretty comfortable with the routine. I would set up and get to work quite quickly, I even helped some of the newbies out without worrying about the quality of my advice. But polishing was entirely new territory, so I ended up standing around hoping somebody would volunteer to help me. And they did! One of the younger members of the class, Montague, helped get me started with pouring a lap, although Johan stepped in after a while to offer the benefit of his years of experience. The process was simple: Clean and dry both the mirror and the tool. Wipe a few drops of motor oil on the surface of the tool, leaving the thinnest possible film, as a wetting agent to help the pitch to stick more firmly. Place the mirror face-up, with an appropriately sized mould resting on top, so that it conforms to the shape of the mirror. Melt some pitch (previously confirmed to be of a suitable hardness), and fill the mould. Place the tool face-down on top of the pitch and gently hold it in place while the pitch sets. As the pitch becomes brittle, break away the sections of pitch overflow (when working with plastic moulds, this excess material is called 'flash'. Does that word also apply here? Answers on a postcard to email@example.com
), until the pitch has set enough to become brittle. Peel off the mould, take a putty scraper and chip away the rest of the excess pitch (anything protruding past the diameter of the tool). And then the lap is created and ready for pressing!
This was when I realised I'd forgotten to find a bottle to hold the polishing compound and had to use the class bottle. A few splashes of the compound (actually a thin slurry of cerium oxide powder suspended in water. It's an interesting colour - I remember when I was a child, my crayons always included one labelled 'Flesh', which was about the same colour. Incidentally, I never met anybody of any race whose skin looked like that crayon...) onto the mirror later, place the new lap face-down into the mirror, and apply pressure. For the first few presses, I simply leaned on the glass with my full weight. This puts a lot of pressure on your arms, so I soon switched to just balancing a bucket of water on top of the stack and letting it stand for a break. Pressing is important because it presses the pitch into increasingly intimate contact with the mirror surface, until it conforms exactly to the shape. Once this is done, we say that it has good contact, and we know that all parts of the surface will be polished evenly. This process also embeds some of the cerium into the surface of the pitch, which is critical to the polishing process.
Once that first pressing was complete, I could begin the actual work of polishing - with the mirror still on top, I pushed the tool back and forth in a wide W stroke with one third overhang - exactly the same as the early grinding stages. And this is when I experienced the practically uncontrollable conditions I described at the beginning of this update. After about 15 minutes, I put down a fresh charge of cerium oxide slurry, pressed some more, ground some more. After the third pressing, I turned everything over and polished mirror on top - I will continue working in this way until polishing is complete and I have achieved a perfect spherical shape.
Once that is done, I begin the final stage: Figuring. I'm given to understand that this is a form of sorcery, so I won't spare any thought on it until I absolutely have to.
A final point: The reason a mould is used when pouring the pitch is simple: a lap must be channeled, so that it looks like an array of square facets (although I'm sure the actual shape is unimportant. Squares are just easier to make). Traditionally, they would be pressed into the soft pitch, or cut out with a knife. This sounds like a terribly messy and fiddly process. We used soft re-usable moulds instead. No mess, no fuss, just a perfectly regular set of well-spaced squares.
So why are the channels necessary? Two reasons: First, pitch is an extremely viscous liquid. You know the popular myth that glass is actually liquid and flows slowly over the centuries? Well pitch really is like that, which is why it's used to polish optical surfaces. It's rigid enough to polish the surface, while continuously conforming to the surface of the glass as it gets polished to new shapes. But if it is to flow, it needs somewhere to go, and this is provided by the channels. The second reason for the channels is that most of the polishing work is done at the edges of each facet. Without channels, the only edges would be at the outer perimeter of the lap, causing it to polish around the edge of the mirror while barely touching the centre.
It is official - my mirror is done grinding. Although I had planned to stop at #800 grit, I found myself with time to kill at the last class, and decided to give it a few wipes with #1000. I spent about 45 minutes working with this stuff, so that I'm confident my mirror is now a perfect spherical section, within a tolerance of about 8 microns (since that's about the average size of individual grains of #1000 carborundum). Which is pretty impressive, considering that this is done by hand, with no instrumentation. It blows my mind to consider that these techniques were developed hundreds of years ago, that there was a time when opticians would test their mirrors with monochromatic light created by burning salt in an oil lamp. The art of mirror making has been unchanged (for amateurs, at least - professionals use machinery) for centuries. Sure, for the longest time the mirrors were ground out of a hard alloy called speculum, instead of glass, but the techniques did not change with that transition.
But back to my mirror: I have not tested it in a long while, but it has a diameter of 155mm, and a focal ratio which I seem to recall being around f=6.5. Fine grinding is complete, and I managed to avoid scratching the glass in the process (no mean feat - my workstation at the last class was liberally dusted with what was either iron filings, or carelessly spilled #80 grit. I didn't investigate, for fear of it sticking to my hands. The effect of one of those falling on my glass while I work would be pretty devastating). So, now I start polishing. Before I can start, I need to collect or build a few items. First a polishing platform, which is something to grip whichever glass is underneath formly. I'm told you overcome a great deal more friction polishing than grinding, and if the work is not firmly anchored, it will slide all over the table. I was shown an example at the class, and it was suggested that I simply use that one, but it wouldn't work for me: My tool is quite thin, at maybe half an inch. That makes those supports taller than the tool, so that I wouldn't be able to move the mirror at all (Unless I was making 1mm strokes...). Instead, I will cut a square of chipboard from one of those incredibly useful offcuts and scraps that all men keep lying around, and make retaining pegs from some rubber washers I have left-over from the last time I tried to fix some plumbing. I'll stack them up two at a time (for a peg about 8mm tall), and screw them down into the board. I also have about a litre of leftover white enamel paint which should be salvagable, and will use this to seal the wood with about 3 or 4 coats before screwing down the pegs. Finally, I have some rubber mats from my car (I removed them because they slide around from day to day and like to bunch up under the clutch pedal. Irritating), which I might attach to the bottom, or simply place under the board, to ensure that the board itself grips the table.
The next item I need is a plastic bottle - the sort used to dispense vinegar or tomato sauce in restaurants. This will hold my cerium oxide slurry, and make it easy to apply whenever necessary.
And finally, it's time to start working on a mirror cell. The literature is full of talk about complex multi-point flotation systems, active support, slings.. but my mirror is small and it's thick. I'm pretty sure it won't flex, and I live in a country where indoor temperatures seldom differ from outdoor temperatures by more than 20 degrees. I don't need to think about airflow to stabilise temperature quickly. So my design will be a simple one: Flat round plywood base, with a radius several centimeters larger than the mirror. A circular ring will be attached, with an inner diameter as close to the radius of the glass as possible, forming a recess into which the mirror will fit snugly. Line this recess with felt (or possibly some other material to cushion any irregularities in the plywood), and fit a few retractable retaining clips to prevent the mirror from ever falling out once it's installed. Simple and easy, if my woodworking skills existed... still, I know enough people who can help, so that shouldn't be a problem.
Things have gotten quite busy at work - servers have crashed, and I've been putting in a lot of late nights putting out fires, rebuilding services and restoring backups. This has left me no time at all for mirror grinding, so yesterday afternoon at the Class was the first bit of work I've managed since last time. I made a near-fatal error, however, which almost ended the project right there...
Johan was there with an interesting toy: a vintage naval gun-sight (Correction: Not a gun sight, but a moonscope - see the note at the end of the page) converted with the usual ATM flair into a compact little 5 inch refractor. The objective lens was quite heavily masked to hide the many large chips (some the size of a thumbnail) which had presumably been caused by decades of rough treatment. The original tube remained intact (a heavy metal tapered tube, shaped to approximately follow the shape of the light cone), and had a diagonal with helical focuser stuck to the end. What made it particularly interesting to use, though, was the altazimuth mount: The altitude axis was centered on the optical axis of the eyepice, meaning that the observer does not have to move his head as he tracks across the sky. Obviously such an arrangement puts the entire weight of the telescope to one side of the axis, so it was balanced with a heavy weight projecting some distance from the axis. It sounds pretty clumsy and I'll admit that it looks that way in the flesh as well. But it was surprisingly smooth and fun to play with. Besides, telescopes are for looking through, not at.
Having seen the new scope, it was time to work. I put in about an hour's work with the #500 grit (not very nice to work with - some moisture must have entered the salt shaker I use as a dispenser, so that it formed clumps and would not pour. I eventually took to removing the top and pouring unavoidably huge piles directly from the bottle. Wasteful, but at least it worked), and check with the loupe.
This testing for pits business is getting to be more and more time consuming. The pits left at this level of grit are approaching the limit of what can be seen with a simple magnifier. Still, I managed to identify a lot of the features I'd discovered the last time. To check whether they were actually pits, or just microbubbles and other flaws beneath the surface, I shone a bright torch across the surface while inspecting the same areas, lighting them up from different angles. My conclusion at the end of this: NO pits left. Fantastic!
So, the usual clean-up procedure began, although with a little more care - previous sessions had seemed to create a fair number of new pits, which made me worry about contaminants and scratching the surface. Not wanting to put the glass anywhere where it might pick up old grit, I rested them back-to-back in the water bucket. I then rolled up the wet newspaper I'd used to line the workspace, carefully packed away the saltshaker full of grit and the sponge I'd been using to wipe the glass clean between wets in a ziplock bag, and scrubbed my hands thoroughly. I then opened a packet of #800 grit (Such a light grey, and so very fine that it actually raised a little dust cloud every time it was disturbed!) and poured it out into a brand new salt shaker. I moved the bucket and spray bottle to the floor, and laid out clean newspaper. Only two tasks remained: Wash the bucket and spray bottle as thoroughly as possible, and clean my hands again. I took the bucket outside, and threw the water into the bushes. A sickening glass-on-glass sound rang through the air. I had forgotten to remove the mirror and tool. I had just hurled 18 month's of work into a garden dotted with heavy bushes and rocks.
I died a little on the inside.
After a little thought, I reasoned that, since I hadn't heard glass shattering, it might be worth trying to salvage the situation. I further concluded that, since the pieces of glass had already come to rest, they could be left to lie where they were until I'd decided what to do next. So I went ahead with clean-up, scrubbing out the bucket and rinsing off the spray bottle, wetting the new newspaper, arranging everything where it needed to be, and only then ventured into the garden to find my mirror.
I found the tool first - it was lying on a bed of loam, between the garden's concrete edging, and a spiky bush. I rinsed off the dirt, and put it in place at my workstation. I then went back to the garden and searched for the mirror. And searched. And searched. Several times I stopped, wondering if perhaps I'd removed it from the bucket after all... but it eventually turned up, wedged between the branches of a particular dense bush, suspended about 20cm from the ground.
To cut a long story short, the mirror and tool both survived, with no chips or cracks. Which is why, for just a few hours, I was the luckiest man alive. Incidentally, I was too nervous to inspect for scratches, so just dove straight in with the #800 grit. Everything I said about the #500 grit applies doubly here. It takes a good minute of each wet working back and forth before the film of water is thin enough for the abrasive to even touch sides and actual work to begin. And it's such a subtle feel, that the only way to tell when a wet is done is to watch for when the slurry changes colour from light grey to very light grey (indicating that it's now composed more of ground glass than carborundum). But it's very very fast - less than half an hour after beginning, it looks almost complete with just a few pits near the edge. If I can only get in a session or two at home, I might even be able to start polishing next time I make it to class. And I seriously genuinely cannot wait!
If you've ever made a telescope, I'm sure you'll have a few horror stories of your own. Come share them on the Urban Astronomer forum.
Update: Chris Stewart wrote on the Urban Astronomer facebook page to say "The scope Johan had is not a gunsight, it is a moonscope - a 5" achromat refractor with a roof prism. Many of these things were distributed around the world in the early stages of the space program, when teams of volunteers would track and time satellites and send the data to the US for analysis. This allowed scientists to deternine orbital element, ascertain the degree of atmospheric drag on orbital devcay, and so forth. It is a lovely little scope with a long and illustrious history."