Polishing a Small Spherical Mirror Surface on a Glass Blank
ฝัง
- เผยแพร่เมื่อ 16 ก.ย. 2019
- This video is a follow-up from the video on spherical radius grinding. In this video I explain briefly the grinding process (using sIlicon carbide, SiC) and polishing process (using Cerium oxide, CeO and optical pitch. If you want to know more about the properties of optical pitch, this video is entertaining:
• We’ve Watched This Dro...
The surface is inspected using a simple Foucault tester made of a smart phone and a razor blade. More information on the Foucault test can be found at:
en.wikipedia.org/wiki/Foucaul...
or
stellafane.org/tm/atm/test/un...
A very good book on optical manufacturing techniques is "Fabrication Methods for Precision Optics" by Hank H. Karow. A good book on optical testing is " Optical Shop Testing" by Daniel Malacara.
Gordon Waite has a great series of video's on polishing and grinding of large telescope mirrors: / gordonwaite
It is now possible to contribute financially to the Huygens Optics TH-cam channel. Please refer to the "About"-page of the channel for details. - วิทยาศาสตร์และเทคโนโลยี
This is black arts you share with the world. I am humbled someone who knows this esoteric knowledge has kindly decided to share such things in detail on youtube. What you do is pretty amazing, and I may use some of this knowledge at one point in my watchmaking for crystal polishing. Thank you!
Same here, this is extremely valuable, well into the black arts territory.
Unless you’re sticking with mineral crystal, I don’t think this will work as well for sapphire. And unless you’re working on a watch with a very unique and hard to find crystal , I can’t imagine spending hours on just a crystal
@@CaskStrength777 This level of accurasy would go to waste on watch crystal. It is only required for precision equipment that utilize optics for light focusing of some regard or other. The human eye can barely make out defects in the millimeter level, let alone the subnanometer level. Besides bragging rights that your crystals will be polished the the sub micron or subnanometer level, the practical advantage of this will be absolutely null, and as indicated in this video, the process is far too laborious and costly to justify excessive and unneccasry work to this extent.
@@imvisier9925 you misunderstand watches and watchmaking, where "black polishing" a screw head to a specular finish is mundane.
The process of polishing brittle optical materials presented here is directly applicable to many watchmakers working on older watches or making new ones. Restoring an old crystal, or custom fitting a new one, is often sensible when balanced against the time and money costs of obtaining a replacement, if one is even available.
A watchmaker wouldn't be concerned so much with dimensional precision in polishing, other than it makes work quicker and easier to produce a consistent result. The idea of using pitch or a similar conformal surface helps to this end. Also helpful is the explanation of surface damage and how inspect for it, which doesn't seem to be common knowledge.
Now I understand why high quality optics is so expensive.
@@laharl2k How much would such a machine with competitive results cost?
I can’t see myself ever needing to do this - but nevertheless, it’s always great to see a master of their craft at work. You have amazing skills and thanks for sharing them.
Cerium oxide is also used in semiconductor wafer polishing :). when doing failure analysis at the micron level we actually look for Cerium during atomic analysis to confirm if a defect was introduced in a polishing step
In the making if you pass sound waves from furnace to Cut gob the structure is more finally packed with less micro scratches
One of the most interesting channel I'v find in years. Thank you so much!
I'm currently doing this by hand on a 12.5" chunk of borosilicate glass to make a telescope mirror. Just started the polishing phase. It's amazing how this process can make a surface with nanometer precision.
This is absolutely fascinating. By pure dumb luck and no craft, I managed to polish out fungus lines from an internal convex element of a Cooke Speed Panchro lens without destroying it. The lens was full of fungus and had been written off as uneconomic to repair. Had I known its history, I would have given it more respect and handed it off to someone else who knew what they were doing.
Slowly working through your videos, just wanted to say thanks! They are fascinating, keep up the great work!
Lol just watched your oled video and then went on a tangent researching mirror grinding and here we are again haha
Ahh this takes me back to my mirror making days. I never ground anything that small though. My first mirror was a 4.5 inch Newtonian, then a 6 inch and then an 8 inch. It's a very time consuming tedious process especially figuring a Parabola on a Newtonian. We sometimes had access to Pyrex Glass but my 8 inch was regular glass and I had to wait 2 hours for the mirror to cool down after polishing before I could even test it with a Foucault Tester
At the end I wanted to scream hysterically non-stop! I SINCERELY think this is really incredible and amazing (honestly!!!) but if I had to do it myself I would rather bite my own head off! You deserve congratulations orders of magnitude greater than I could conceive in words! Just... amazing.
Your video reminds me for making my first mirror years ago. You give me confidence to try again. A very rewarding process. And something to enjoy. Thankyou.
It’s always so interesting to me how such simple things we take for granted in life have had so much time and effort put into them. Great video
I don't know why I look at this. But it's new and refreshing stuff to learn for me, so I'm going to stick around for a while.
This is fantastic, really super that you mention approximate times for each step, and really awesome hack with Cell phone there for knife edge test + great illustrations and ACTUAL values for things (which is super helpful) ~ Looking forward to learning more [Thanks for making these videos ! ].
Thanks for posting this! I'd love to see how you add the reflective coating. Hope you make some more videos. Thanks!
Very interesting! I found your channel from your DLP video, and it's been fascinating to see how just hand polishing can achieve such smoothness.
Really interesting. Im moving to the mountains of spain soon and i’d love to make a small newtonian scope. Looking forward to the mirroring video. Cheers julian
I love that phone Foucault test
This is a wonderful tutorial, thank you
This was extremely informative. Thank you very much!
Excellent, again! Congratulations!
I dont 'like it when you say "maybe in a future video". I like it when the video is 10 minutes longer and you explain it. :)
I learnt more from this video than 1 semester of school
Very cool!
I got that Pitch some cerium oxide, pitches love cerium oxide 😂 jokes aside, I love this channel now that I’ve found it. I love honing surfaces of things in my free time. I even have a pc a honed the surface area of the cpu and heatsink, then honed the heatsink to the cpu (to mostly accommodate for any convex/concave surfacing that may have been introduced by my lack of skills to either the die or the heatsink) and tested it with and without thermal paste and all cores run cooler without the thermal paste and I did long benchmarks and rarely saw over 45 Celsius on any core with a very modest, stock aluminum air cooler. Now this isn’t a ryzen but it is a amd “A” series, I can’t remember exactly which, but I still use it, as is, for CAD design and software testing and never had a hiccup.
This is really excellent. I have a question. I have a telescope eyepiece in which the coating on the outermost (field) lens is deteriorating. I want to completely remove the coating so that the appearance will be uniform across the field instead of darker where the coating is missing. Can I accomplish this by pressing the lens into styrofoam and then using polishing compound with light pressure on the lens against the styrofoam? Thanks!
Interesting, thanks you!
Does it become a more standart process, if you use ultrasonic waves to make thrill up abrasive particle within solution?
Interesting work! Is it a personal preference to grind these by hand, or is (partially) automating it very hard to do?
Hi Jeroen, Fantastic channel. I wonder, could you give advice on where to buy small glass blanks like these. I'm interested in getting three BK7 blanks, size is not critical (maybe 35-40mm dia 14-16mm thick) . Can you give any pointers? I'm in the UK. All the best Matthew
Bedankt U motiveert mij weer.
Will using your hands as point of contact for shaping the curvature trough any chance of give a person to getting a good measurement. thanx a bunch. highly education.
Extremely well explained.
What I was wondering is: where do you get your raw grinding surface from? I mean the one that already has the right radius on which you're grinding the your glass mirror.
Initially both surfaces are flat, but through a special grinding method and using abrasives with large grains, one of the surface becomes concave and the other convex.
How do get the surface curvature that you require. You impressed the blank on the pitch but the blank doesn't have the final surface. I don't understand.
Amazing channel. Is this how microscope lenses are made as well?
Hello, very interesting video, thanks! Can I ask if you have done any irregularity/power/flatness measurements (with an interferometer) for such hand made items? What levels can you reach?
Hi Tomas, I did do interferometry on small mirrors (not this one by the way, this was just for the demo of the process). So, if you put some time in it and avoid the pitfalls related to temperature (friction/warm hands) you can get down to 1/8- 1/10 Lambda. As you can see at around 11.05 , there is some irregularity in the surface. This is due to pressure and can also be largely avoided by milder polishing in the final stage.
Thats funny. Long ago a had polished a 135mm spherical newtonian mirror. But i was not succesfull in the job and sadly i had gave it up. Now i want to try to polish a flat one. Can i use this method for polish a flat by hand? I worry about turned edge. What kind of tip you give me to succesfully complete the work? Thanks
I hate to ask a question that has been repeated before but if I'm interested in learning to make my own lens, do you have a video or know of a video that shows how to source these materials?
What about grinding a optical lens? Is 1200 still sufficient as the finest grain size?
Nice
what temperature for the pitch ? where do you get the supplies ? Thanks.
Generally you use the pitch at room temperature, so let's say around 21 deg. C. There are different types of pitch which have a different melting temperatures. This influences the "hardness" (viscosity) at room temperature. The type of pitch you want to use is dependent on your surface. Generally, you use softer pitch for a-spherics. Since I am in Europe I orderd at a German Company named Pieplow & Brandt. However if you search for "optical pitch suppliers" you will probably find a supplier in company.
Hello. How hard is the pitch you use?
I have some lenses to polish. The soft has several grades of hardness.
I generally use the medium grade Gugolz pitch. Since temperature is an important parameter, you can still vary the hardness a lot by polishing at a different temperature. For making spherical surfaces, the pitch should preferably not soft.
Love this channel! Now I'm hooked on this post apocalyptic garage made ultra precision cyber {possibly crypto} optic thing.
where can I find all these materials?
Can you aluminize a ground blank and just polish the aluminum? Maybe you'd need to re-aluminize but it seems like it might be easier
Aluminium is actually way more difficult to polish to optical quality then glass.
You should make a video about Holograms, that would be cool
do you polish again after adding a reflective surface?
No, all polishing and figuring is done before coatings are applied. These processes remove material, so you'll only be removing the thin coating you just applied if you attempted to polish after coating. Sorry this is 3 months late, but I hope it was helpful either way.
I can’t imagine the difficulty making the mirrors for the JW Telescope must have been.
10:40
Where can one person find the lens holder mechanism shown holding the small lens?
Thanks
This thing is sort of an antique, I got from an astronomy club that were throwing stuff away. So, the best advice I can give you is see if you can find something similar on ebay. Look for "self centering lens mount".
if I use 1\8 glas as a tool and flip the tool at each 30 laps do the result can be be convec lens I need a beconvec lens of 50mm dia and 50 footh Focal for projection of the sun but do not know howto make lems of this kind
Hello, can you tell me what can I use to polish watch mineral glass?
Guess the best way is to use a dremel, some polyurethane polishing pads and cerium oxide as the polishing agent
@@HuygensOptics thanks for answer! I tried cerium oxide but with felt wool pads. It is working but very slow. The water solution dries up to quickly its problematic. Maybe can you recomendet some polishing past?
I have got a freeform lens to be polished, anyone knows how to do it?
The Free-Form Surface Lens' buffing is very close to the process as in this video, just use wool grinding head instead of the pitch. Thank you Huygens Optics!
It's almost 1 year you haven't update any video, hope every things is OK with you!
why dont you use a lapping wheel to polish the glass blank?
Actually, I just wanted to show how easy and fast you can do this by hand with a small mirror like this one.
4:50 what if the particles are bigger than the uneveness
I would have put the glass in a drill, and rotate it at higher speeds, with less friction. Causes smaller cracks, while still allowing for fast work completion.
That is exactly what I want to make. What emitter are you using? Osram or Cree? Are you making your own driver too? I love the Anduril UI. The FET drivers are good for high amperage but need a lot of heat sinking. I hope you post more on this, would love to see details.
1:02 regarding subsurface cracks. Extreme hazard considerations aside, I wonder if either grinding with hydrofluoric acid or pre-soaking the lens with HF, for some soak-grind-(wash) time cycle to be experimentally determined, might reduce crack propagation by loosing up/freeing up the glass surface "bits" (not sure exactly of the microstructure of glass) so they would break off more easily and not transfer force into subsurface layers and induce cracks. Has anybody ever tried that (and lived :))?
I'm sure that you could etch the subsurface damage away if you etch for a long time (it's a relatively slow process). However, the fastest (and safest ;-) way to remove it is just to use sequential grinding and polishing steps.
@@HuygensOptics thanks for the reply. I meant less about etching away damage after it happens and more about preventing it altogether--improving the grind process itself by (maybe) allowing more aggressive grinding with less subsurface damage, if the HF could somehow "loosen up" the top layer of glass. So the glass bits would break off more easily and (maybe) not cause forces to propagate down into the glass as much.
@@perspectivex Aha. Well, I'm not sure how that would work. The loosening of particles from the surface is actually the result of crack propagation in the glass. Anyway, I would never try to attempt grinding with HF as an additive to the SiC slurry...
@@HuygensOpticsI'm not going to try it either... I'd like to keep all my finger bones.
...do people ever grind their own mirror flats for star diagonals? I can't find anyone doing that but I don't know why not.
You need a lot of Zen for that job.
Good Lord, you made 500 mm mirrors!!!
now imagine how the biggest telescope which is about to be put in space was made and why it is so expansive and why it took so much time... James Webb I mean
Why don't you make some mechnical tool and automate polishing process? Why doing it by hand?
why cant the parabolic shape be cast into the blank so that all you would have to do is polish it to its final parabolic shape?. if you can cast a flat blank you can cast a parabolic curve with the different focal points!.
for smaller mirrors (
@@HuygensOptics what about taking a smooth mirror blank (talking about the
@@perspectivex , yes in principle it is possible if you start out with a very parallel glass plate and an extremely accurately machined mold. However, 10-20 microns is still a long way from the final shape, which needs to be accurate to 20 nm for a good mirror. That is a factor 1000x more accurate. Polishing away 20 microns of, let's say, astigmatism is so difficult that it is better to start of with a nicely grinded spherical surface, polish it (spherical) and then deepen the center area to make it parabolic.
@@HuygensOptics hm, ok. I unfortunately don't really understand how long grinding really takes per grit change and how much is a lot of glass (I would have thought grinding off 20 microns would fall on the "little" side but I guess not). I also thought good telescope mirrors were 1/10 wave, which would be 500nm? But now that I start to read up on it I see someone saying 1/10 wave is 50nm, which doesn't make sense (yet). oops, no, my mistake--I'm always thinking of light in Angstroms and sometimes get the factor of 10 backwards. I get it...it's 50nm. p.s. can you point me to any references on how to make a machine to grind telescope mirror blanks (in the 200-300mm range)? I will google but you seem to have a lot of experience and might know a good reference. A machine for roughing it down to shape or one for polishing (not sure they can be the same machine).
@perspectivex. You are right about the grinding of 10-20 microns, that can be done very quickly. However, I was under the impression that the goal of the whole casting exercise was not to grind any more and start polishing immediately. Since polishing is much slower, 20 um is quite a lot and it will take many hours, whereas with grinding it may take as little as a few minutes.
Generally we we refer to labda's as an accuracy when we mean wavefront errors. 1/10 Lambda surface error will result in 1/5 lambda wavefront error. So a 1/10 of a lambda wavefront error will require 1/20 lambda surface accuracy, which is 32 nm at lambda = 633nm. If you want a simple machine for 200-300mm, look for "mirror-o-matic" either on youtube or google. cheers.
9:59 - all these worlds are yours...
Small concave mirror surface
I wish they would do something to bring down the cost of telescope parabolic mirror, so that people don't have to sell a kidney just to buy one!.
Making a good parabola is quite labour intensive and requires many hours of measurement and correction. That is why people make their own mirrors: if it's a hobby it's not that important how many hours you put into it. And you can actually make very high quality optics by hand - if you know what you are doing. You can buy cheap telescope mirrors (mostly from China), however their quality and optical performance is generally not very impressive.
@@HuygensOptics why can't they cast the parabola shape at different focal points into the mirror so that all would be needed is a final polish and then get coated. it would save a lot of time and labor, or even maybe cut the shape with a diamond tipped lathe like on a betts mill. only the betts had a turntable thats 12 ft across. I mean if the cost could be brought down then more people could get into the hobby.
@@davidbrandenburg8029 Glass doesn't cast like that. It just doesn't hold form like would be needed to do that.
@@davidbrandenburg8029 So, simple answer on the diamond tipped lathe process - that is an optical manufacturing process called single point diamond turning (SPDT). SPDT only works on ductile substrates, there has to be some plastic deformation in the substrate at the tool tip for the chip to form in the right way to make an optically clear surface. On materials such as Germanium this is very straightforward, however glass is too brittle for chip formation. There are advances recently in this technology with Ultrasonic or Laser assisted SPDT where ductile mode machining can be induced in brittle materials, however it is far too recent to be the "conventional" manner of producing parabolic surfaces in glass, or hobbyist level technologies.
0:14 400mm? That can't be right? That looks like 40mm. And maybe not even radius, but diameter.
400mm is the radius of curvature in the front surface, not the dimension of the mirror.
@@HuygensOptics I see. That makes sense. Thanks.
Your wobbly table gives me unrelenting anxiety ngl
Having ground a 700mm mirror, it gives me a giggle to watch the fingertip grinding!
Why am i even watching this
sir thats a fuckin cylinder
Ah yes. I always make sure to keep the strokes short.
Why do you keep calling it 'spherical' if you've been polishing it for hours, making the surface finely planar? Surely spherical and planar are 2 opposite things.
The surface is not planar. It spherical in the sense that it's surface is a small part of a sphere with a radius 400mm
wait what is 400mm radius?
Cause that would be 80cm, almost a meter in diameter - 0:19
Edit: ooooh, the curvature has that radius, okay
Interesting. I use to make glass lenses for eyeglasses. The way your going about grinding the lenses seems primitive to me.
Doing all that seems unnecessary to me. There are machines for that. Also instead of pitch we used cast iron tools cut to the correct curve they had slots in them to. But later they used wire meshes pressed onto the lap to avoid wear on the tool. Even later they used thin zinc pads. To polish they used pellon pads which is a though paper like what is used for shoulder pads in dress making. Same polish you use. With a machine spinning the lens in tight circles while rotating the tool or lap you could make them much faster. The polish is continually pumped onto the lens. Same with the abrasives.
Lol you poor thing. Can't read? This was a demonstration video to show the process of hand polishing...
@@mikeznel6048 Snooty response with an insult thrown in. That's what I get for trying to be helpful. Disappointing though, I'd hoped one such as you might have more class.
That's not 400 mm. That's ~40 mm 😁
The radius of curvature is 400mm.
Wouldn't you want to wear gloves?