Parabolizing a 20" Mirror, Part 2: Deepening the Curve
Вставка
- Опубліковано 28 сер 2024
- The curve on the 20" quartz telescope mirror needs to be deepened from a roughly spherical surface toward parabolic. Gordon Waite shows how to use 7", 8" and 12.5" pitch laps to rough in the paraboloid. Gordon performs pitch lap polishing by hand on a powered turntable.
Hi Gordon! I'm watching this from Salzburg Austria while finegrinding my 13"-project f/5.15. Your videos are awesome. Really getting down to the point of
what's really important in a very compact way. Also appreciate your willingness to answer related questions and problems. Thx and regards.
I wanted to say thank you,mi always wanted to build a telescope,mi am more like an engineer so a big stainless tube, flanges gig welding, tongue bars etc are not much problem for me I always thought 20" would be ideal and just was like a mirror like that would be like a million dollars. I learned a lot I'm watching all your videos. Thanks
Mirror making is an art, its something that I have always wanted to do. Maybe ill start of with a small 6" mirror. I have a local company who can supply the blank, grinding compounds and do the aluminium coating. Love your videos.
+paul30003 Thanks! I hope you give mirror making a try. It's not all that hard, particularly if you have somebody to lend a little help once in a while. It's really satisfying to take your first look through a telescope you made yourself!
Hi! Gordon!
May i ask of you~ 07:50 Is it overhang strokes?
Thank you very much😃
I have a question, how do you parabolized the mirror? I mean, with those movements you did on the video is what makes a circle surface become in a parabola surface?
Null testing,because of the setup is more useful for mirrors on a production basis, many of the the same size and F R, I like the test you do a Foucault with a screen. I watched a YT video of his testing: an artificial star, in and out of focus! I do not know if he even did a Ronchi grating! He stayed with me for a week many years ago.we had a hard time getting him to steep on our couch with his preference the floor. I would like to to do a 30" or 32", it is on my bucket list. Know where there is glass. In the basement at Lick! Trouble is that it wasn't the piece that came out of the secondary hole, but it is 13 inches thick, but honeycombed! Good luck and thanks for the YT!
What size polish are you using and what, starting with a sphere, are we supposed to try to do to get a paraboloid ? Thanks for posting.
Hi Gordon
Wow.... what excellent information you have provided... as usual !
Thanks for taking the time to do these videos and I look forward fpr more.
Are work the F3.5 on Cloudynights Forum? Man that thread was awesome!!!
Al l my best
Bill
Gordon,
Thanks for the excellent videos, you have been reading my mind as to what I wanted to see. A couple of questions if you don't mind:
1. What kind of abrasive are you using here, and are you doing anything special to keep it wet? I never saw you spray any water, but maybe that was just off camera.
2. Have you ever done anything to quantify how much glass you take off at what rate with the different laps, strokes, pressures, etc?
Thanks again, and please keep them coming.
Frank S
Mirrors are polished with cerium oxide aka jewelers’ rouge. A solution of that in water is the milky liquid he sprays on at the start. You ave to refresh it occasionally but that wasn’t the subject of the video so it was cut. While polishing you take off no glass, you’re mainly just moving it around. Figuring, maybe a little, but it’s still nanoscale. While rough grinding is when you get the cubic inches.
Oh Boy. I didnt know making a mirro needs so much talking !! :) Beautiful job well done. Very special work indeed....:)
Thanks for your kind comments!
You are very welcome. Love watching your work...:)
We are supposed to be flattening the curve, not deepening it!!!
Jokes aside, super cool video!
I think it must be pretty cool to look though a scope you made? I think that would make me feel so much better about what I'm looking at.
+Markinpuff It is a really neat feeling to observe with a self-made telescope. There is something just magical about working on that simple piece of glass, and then finding out that it will reflect those light rays into such special images! First light is always a joy for me, even after the hundreds and hundreds of mirrors I've made over nearly 50 years!
Super video informative everyone must know *
What came first, the egg or the hen .... OR the curvature of the mirror or the curvature of the lap ?? That's what I don't get ? Also, how do you know how the curve is going ? A spherometer ?
Yes, we use a spherometer to measure the shape of the curve. The tool and the mirror always have close to the same shape. Obviously the tool lags the mirror, as the tool is doing work on the mirror's surface.
Hi. I'm a photographer and I was really surprised to learn that the best lenses money can buy are the ones that have been polished by hand. I find that extremely fascinating. Can you share your professional opinion on why that is?
In the telescope mirror business, pretty much everybody uses machines to work the glass. The biggest difference is how much time a maker is willing to spend on a mirror, and the amount of money they are able to charge. The biggest manufacturers stop improving a mirror at "good enough." They charge the least. Other makers are willing to keep testing and perfecting to make better mirrors. They have to charge more for the extra labor, but they are producing better mirrors. Nobody in the "consumer telescope" business really has a machine that can produce a mirror without human testing and correcting, although quite a few of us have computer-controlled polishing machines that automate quite a bit of the work.
This seems like it would take FOREVER. You're trying to make the mirror bowl shaped? I would have liked to see a close of of the finished product.
It takes a while. a typical 20" mirror might take a couple of weeks of full time work. A lot more if you are less experienced.
GordonWaite - There's gotta be a better way. Take your mirror to a man who does this for a living.
You are funny. I own Waite Research. Our company makes mirrors for telescopes and spectrometers for clients all around the world. These UA-cam videos are meant to help people who want the fun of making a mirror by hand for themselves.
GordonWaite - lol. I'm only here because I was trying to find out why it took such a long time to make the Hubble telescope mirror. Something tells me they didn't have one guy there doing this.
sorry i'm french and i don't understand why are you using a sort of sponge to make this mirror parabolic, is this actually acid ?
Not a sponge! It is optical polishing pitch. Not acid! It is cerium oxide in water. It's a polishing compound.
How do you determine when you need to stop correcting the curve? Ronchi test? Do you remove the mirror to do that?
During fabrication, most workers use the Foucault test to get numbers (zonal readings) on the mirror. You run these numbers into a software package (like FigureXP) and you can calculate the accuracy and quality of your work, and how close you are to perfect. To test the mirror, you do have to take it off the polishing machine, clean it, and put it on a test stand and use an optical tester to take readings on it. In general, you take a set of readings and get a report, and then plan the strategy for your next figuring session on the polishing machine. You do this over and over again, hopefully moving closer and closer to perfection. Once you get past a certain point, you can't measure the surface any more accurately, so it doesn't make any sense to continue working on the surface. Or else you are unable to figure out any procedure you are able to carry out that would help and not possibly hurt the figure. Of course, if you are making mirrors as a profession, you normally have some standard that you will work to achieve, and then quit. It's not economically viable to continue with every mirror until you have reached the physical end. When it meets your shop standard, it ships!
GordonWaite
Making a telescope mirror is the most precise thing a human being can do with their hands.
It's been said though, that a mirror maker is never really done with the mirror.. somebody just takes it away from them.
I really like and appreciate your Mad Skills And the work you put into your videos.
I am learning, and thanks to you.
I have a bit of confusion when i hear that during parabolization you have to deepen the mirror. Maths says that a paraboloid whith a focal length F is shallower than a sphere of 2F radius for the same diameter of the mirror.
So if you make an F paraboloid of a 2F sphere you should reduce the edge.
What am i getting wrong?
When you are making a telescope mirror, you aren't too concerned with keeping an exact focal length. A little longer or shorter isn't a problem. You can parabolize by taking some off the edge and some off the center. That is minimal material removal. Or you can slightly change the radius and take it all "off the edge." Or take the radius in the other direction and take it all "out of the center." The trick is, it is easiest to deepen the center, so we normally parabolize by deepening everything from the edge down to the center. The radius will change a bit, but this method is easiest to do, and keeps the quality of the surface high, especially toward the outside.
Thank you for sharing a bit of your wisdom with me :) I thik I got it. Going to make my mirror this spring and that'll help me.
Hi, are you taking orders? I have always wanted to build my own telescope. However, the mirror might be beyond my willingness. Do you have any spares? Maybe 8 inch or 10 inch... for sale? What is the price for such?
An 8" f/4.5 costs about $1,190. A 10" f/4.5 is $1,725. A 20" list the one in this video starts at around $7,500.
Is the center hole not completely perforated through during grinding / polishing / figuring ?
This mirror is intended for a Newtonian telescope, so there is no center hole in the mirror. When I do need a center hole, I cut it through most of the way from the back before I start, then finish it from the front when the mirror is fully figured. Makes me really nervous to do those!
its making me dizzy watching this, but it is very educational lol
Do you use the null test? Do some videos on testing please! Have you seen how Dobson use to test?
Garth Elliott Hi, Garth... I don't use a null test. During rough polishing and rough parabolization, all I need is a quick look at a Ronchi to make sure the mirror is smooth and moving in the right direction. When push comes to shove and I need to know exactly the shape of the mirror, the Ross null test takes too long to set up, has too many variables, and doesn't give as accurate a measurement of the surface shape. Plain old Foucault is much faster and gives the numbers that people want to see. I'll try to do some video on testing in the near future. I'm designing a new tester and test stand, so it would be a good time to make some video. Dobson was pretty much a star test guy. Once he gave me figuring advice... "Good enough is good enough. Don't be afraid to stop."
rats arsed except that’s not what happened with the Hubble mirror, they tested but their test apparatus was improperly set up.
Do you sell mirrors?
I'm tempted to try make one my self but i feel there's to much that can go wrong.
And not to speak of all equipment i need to build.
Even after watching / reading a lot of tutorials there's to many question marks throwing me off.
I'm looking for something in the 20" range.
Hi, Luftbublan! Yes, we sell finished mirrors. If you would please email me at gordon@waiteresearch.com and let me know what you are interested in, I would be happy to help.
Hi, quite curious how this turned out. Have you taken any photos through the telescope?
beomagi Hi! I actually made this mirror for a client, and it was shipped out. So I didn't have a chance to do any photos. I just got a new camera adapter, so I'm hoping to do some photography through one of the new Renegade telescopes we are building now. Thanks! -Gordon
A new 20" f/4.5 today (2018) would run about $5,500 or so, coated. More common are the f/3.5 to f/3.5 mirrors, which would go for $7,200 to $7,500. People like the shorter telescopes now.
Gordon, how would you tackle figuring a much larger mirror? Wouldn't the proportionately larger figuring tools be harder to manage manually?
With any larger mirror, I would not approach figuring using hand-worked polishers. On commercial mirrors, we are using machines for all figuring work. You can still get it done by hand, but it is a lot of work!
Thanks Gordon. I am watching your videos again. Each time I do I learn even more. The Couder Mask doesn't seem feasible for a 40"f/3 I will rely on the Bath or Ceravolo (Fizeau type) Interferometer. Thank you for what you are doing for us mirror and even flat makers.
When you have the time please make a video on how you figure large mirrors.
Hi, Dan. You may want to rethink your selection of testing a 40" f/3 with an interferometer. Your Bath uses a spherical wavefront for testing. The f/3 paraboloid is very far from spherical, and this will lead to all sorts of issues for you. I've made a lot of successful f/3 surfaces using Foucault and a Couder mask. And when I think of the few people that have succeeded with large f/3 mirrors, all of them are using Foucault (and star testing). Please study how astigmatism works in a Bath instrument before you go down this road. You are going to have a LOT of fringes, and you will need to know how to choose among them with your software. If you really want to use Bath, you might go with autocollimation, although you would need a large reference mirror for that! Check out this thread on CloudyNights: www.cloudynights.com/topic/476304-bath-interferometer/
If I could make a large Couder Mask that would be a big accomplishment. I do have David Harbor's little booklet. I can't rely on a printer to make this template initially and then have something more rigid to be the final material. I will need very large "compass" or a means of making very large circles accurately...then make the inner and outer and middle of the various masks. No I don't have the means of making or purchasing a reference flat (mirror).
Hey Gordon! You say in this video that you would work on this mirror for about two hours. What percentage of this time would be spent on working at the centre/ 50%zone / edge?
Also, how do you measure the deltas? ( I read an blog at CN about you making twin 22" mirrors, and you kept saying
Perfect deltas 34 34...
Current deltas 30 37....)
Greetings from Hellas
Hi! Thanks for watching the videos, and thanks for your questions. When you parabolize a mirror, you have to remove more glass from the center than from the edge. So you will spend a little more time doing that. But it always seems like you spend more time at the 70% zone and outside that mark. The "deltas" are easy. When you bring up the Foucault software, you will be given "perfect" readings for each zone, starting at 0 in the center, and getting larger as you move toward the full correction at the outside zone. The "deltas" are simply the difference between the perfect readings of any two adjacent zones. So if your perfect Foucault readings should be 0, 27, 52, 77, 102, then your deltas would be 27, 25, 25, 25. The first delta is always a little higher, and the rest are all the same. Thus they are easy to remember while you are fabricating the mirror!
Thanks for your reply!
Let's say I want to parabolise a 12 inch f10 mirror, what are the ideal deltas?
If you setup a 12" f/10 mirror with a moving-light-source tester and 6 zones, the perfect readings would be: .000, .014, .027, .039, .052, and .064... So the delta's would be: .014, .013, .012, .013, .012... If you had another digit of accuracy, the last four numbers would be closer together. These show some rounding from the software. These numbers are in inches.
Ok thank you!! Last question
I have read that if you make a mirror that has a long focal length, it doent matter if it is spherical or paraboloid. Is there a significant difference??
On a 12" f/10 mirror, if you leave it spherical, it would have a wave rating of 1/3.62 wave, which is just a little worse than the classic 1/4-wave standard. That wouldn't be great, but it wouldn't be awful. Of course, if you even parabolize it a little bit, it would likely make it to 1/4-wave!
How does it not sand grooves into mirror?? Looks like it leaves lines
The lines you see are just the places where the facets on the polisher channel the liquid on the mirror. The polisher is taking off mere millionths of an inch. They would only leave "lines" with the worst technique by an incompetent operator.
GordonWaite gotcha 👍
Hi there,
Where do I get the glass? And how do I know the type and thickness of the glass I need?
Thanks
One source of mirror material are old recovered navy ship porthole glass. I think the largest are 12".
You can get glass from many glass fabricators, especially companies that do architectural glass. They can usually cut glass on a waterjet. For most mirrors you want a good borosilicate glass like Schott Supremax-33, which runs around $40 per pound. You can make mirrors from regular soda lime (plate) glass, which is cheaper, but has thermal issues. Most mirrors are made from glass between 1" and 2" thick, with 1.5" or 1.625" common for mirrors in the 14" to 16" diameter range.
Would it be worth building a cnc arm that you could program to move forward and back and then left to right , rotation too , program patterns and speed rates . The attachment would come from the centre . It would take most of the brut force grinding out until the very latter stages .
Hi, Kaveman! We started building CNC grinding machines about 10 years ago. The one I am using today controls a fixed-post grinding and polishing machine. I also has computer-controlled abrasive and polishing agent slurry management. Now days I don't have to do much manual labor at all, except the occasional tricky or short figuring step.
How time total to complete a parabolic grind by hand?
Grinding a 20" by hand is quite a bit of work. If you used just a simple fixed-post machine, the grinding and rough polishing might take somebody anywhere from a few days to a couple of weeks, if they worked at it steadily. Hand figuring a 20" f/4.5 takes quite a while as well. Typically, an experienced amateur might be able to finish rough parabolization and figuring to 1/4-wave or so, in anything from a week to a whole lot more. Getting to "premium" status by hand would absorb more hours of work. Hand work isn't all that fast, but it is very satisfying.
How many hours does it take to make such a mirror and what does a blank cost ?
Hi, Kaveman. With the computer-controlled machines, the physical labor time in minimized. You do have to set up for each session, and clean up afterwards. And all of the Foucault and Ronchi testing takes manual time. Surprisingly, machine grinding and polishing is quite a bit slower than doing it the old fashioned way, in terms of hours of work. But I don't care how long the machine takes, because I don't have to pay it an hourly wage!! With all the computer-controlled machines, we can knock out smaller 12" mirrors in a couple of days, elapsed time. The 20" f/3.0 to f/3.5 mirrors take a lot longer, because there is considerably more testing to meet the much tougher measurement and quality standards.
GordonWaite -- i was very interested in the blanks cost also.
Depends on material you want, and size, of course. You can get plate-glass blanks like 10" and 12" for $80 to $120 or so. A 20" blank on Supremax will run maybe $1400 to $1600.
GordonWaite hello can you create one and sale it to me i will buy it for sure
GordonWaite Thanks, that gives me a base.
the abrasive looks like cerium oxide. is that the case?
bgggroove I use cerium oxide for polishing, and aluminum oxide for fine grinding. The coarse abrasives are SiC.
mr. Gorodn
If i want to do a 16" mirror with an f/5, how thick does the glass need to be to be able to parabolized with this diameter?
is it possible with 19 mm?
Your 19 mm glass is pretty thin for a 16" mirror. That is a challenge that will require good technique, and very careful test-stand mounting and testing. Even a 1" thick blank is tough at 16" diameter. And, of course, you will need a very good mirror mount in the telescope.
GordonWaite mmm ok. can I do this mirror with two glass of 19 mm united with Polyvynil Butyra?
I don't think that will help, and might make it worse.
Yes, and there is also the major problem of face-up polishing without introducing astigmatism. He needs to grind the back flat, and also rest it on a cushioning base, rotating the mirror regularly.
Is this even taught anymore.
There are a few books, but nothing new has come out in a long time. I used to teach a mirror-making class when we were in NJ. Would like to take it up again here in AZ if there was enough interest, and a place to do it.
Making a mirror by hand today, you would only use slightly different techniques than those taught in the 30's and 40's. People used rouge for polishing back then, and today you use cerium oxide. Glass tooling was way more common then, versus ceramic tile or steel tooling today. From the viewpoint of the text, though, nobody has written any mirror making book that is more fun reading than those written prior to 1960.
Great stories. Takes me back to my own high-school mirror-making adventures! Thanks...
All I hear is bitchslap :P Great work!
Thanks!
This is making telescope mirror polishing an art. Unfortunately, it's a science. Doing it by hand is oldschool, inaccurate, and ineffective. Using chaotic strokes by hand will ensure it will never be perfect. Get a machine, or build one. It's incredibly easy. And map your cut on a PC.
being an amateur machinist cant this be replaced with a cast aluminum blank that is then machined. I can turn a 13 inch arch on my metal lathe. I can also swing a 20 to 30 inch arch over my vertical mill and power a turn style. Seem that using this method to achieve a concentric curve is sort of primitive. I would appreciate your input. Thanks
Michael Hursh Hi, Michael... the surface tolerance on a finished telescope mirror is in the range of 5 to 10 nanometers, or .0000002 to .0000004 inches. No machinist can hold tolerances like that. Also, the surface of a mirror isn't a sphere, it's a paraboloid. That's a very difficult shape to cut on a metal-working machine, especially with those tolerances.
BaSH PROMPT Nothing is ever perfect, my friend. I own many mirror-making machines, and use them every day in my mirror-making profession. But most folks that want to learn mirror making are going to do their mirrors by hand, not by machine. So I try to make videos to help those folks succeed in making their mirrors. In any case, making a mirror by hand is neither inaccurate nor ineffective. I typically finish mirrors with a surface error of 2 nm to 10 nm, even when working by hand. And your statement about mirror making being a science isn't exactly right, either. There is considerable argument in the field about exactly how polishing glass actually works. I would define mirror making as a craft, probably, instead of an art, as their isn't really any "creativity" to it. By definition, you are just creating a mathematical shape on the surface of a piece of glass. Also, you might be surprised to learn that it's generally faster to work a mirror by hand than by machine, especially on smaller mirrors. The machine just takes the work out of the job. But machines don't inherently make the mirrors any better.
Thanks for the comment. Makes me want to try this.
Good stuff. I am also a professional mirror-maker, and your instructions are good. keep it up!