8 inch mirror project, nearly there!

[Garyh]

Hi Everyone,
Well my 8" f/4.65 mirror is up to the last stage and the most difficult part for us newbies, figuring!:doh:
This is my second mirror so I consider myself as a beginner and still learning as I go. My 6" f/5.6 mirror came out nicely although it took 4 goes to get the mirror right. Hopefully I shall not have to do that again and try to sneak up on it this time...
Ok here is some images of my tile tool and rosin lap and mirror (excuse the dust on the mirror, just dryed it with a old towel).
It was either going to be a ceramic or glass tile tool for grinding and decided to try the glass. Had very little problems with the glass and only got a few minor scratches at the 500 grit. I decided then to bevel the edges with a diamond dremel bit.
Had some problems with that lap as well but I got there in the end!:thumbsup: Just don`t use enamel or epoxy paint near the rosin as the paint will not cure!
Grinding time was around 20 hours (had a flaw near the edge that took forever to grind out)..and around 8 hours off polishing.
Also here are some images of the mirror before my first parabolizing strokes through my tester.
Feel free to give me suggestions and advice as I start getting into this! I know I will need some advice near the end!
Gary

[Starkler]

I cant offer you any advice so I'll just offer encouragement ;)

[Satchmo]

Hi Gary

My only comment is that rosin laps are much less flexible than pitch ( great for spherical polishing ) , and you want a soft lap to parabolise an F4.5, soadding some more oil like caster or linseed may be necessary. I personally like a figuring lap that at 20 C will produce a around a 1cm wide indent of the finger nail under the weight of an arm in about 10 seconds.

The sure sign of a full size lap that isn't flexible enough would be a parabolising scrope producing a spherical hole leaving the edge zones spherical. If you are judging the bend of the ronchi bands at COC, concentrate on the shape of the bands outside focus, which gives maximum sensitivity for the edge zones. Remember that the last 30% of the mirror out near the edge contributes something like 60% of your image ( taking into account the central obstruction).

Hope this helps.

[Garyh]

Thanks Geoff for the encouragement !...:thumbsup: I need all I can get sometimes!

Mark, thanks for that info, I did add 30ml of caster oil as Mathew suggested but I thought it still seemed too hard and added another 8ml. But I still think its a bit hard...Tried your test on it and I ended up with a 4mm long depression from my nail.
Done two 20 minutes sessions with it now. The first was with the normal stroke. On testing it looked like I was just making a central depression and not affecting the outer zones...
Second session I changed to stoke that worked the edges more..This seemed to have worked much better but still the very outer edge is a little slower...
Here is a rather poor image taken outside COC.
Shall do another 20min session and work the outer zones more..
cheers Gary

[Satchmo]

Have you got Ronchi for Windows?

http://members.aol.com/RonWin20/

You can make a cardboard template if you print it out full size . I'd recommend a template of the outside edge of the centre two bands. Set a focus position that has plenty of bend on the outer zones. Just do an outside focus mask , and run a white board marker down the outsides and wipe off with metho after you've done your evaluation.

Cheers

[rochler]

Nice project, your mirror has a real kick@ss bevel hehe....

How do you get your ronchi pix? Have you got a webcam setup for this? If so, how did go about it. I'm keen on doing the same so I can accurately see where I'm going with figuring.

Looks like you even ground the edge of the mirror (how did you do that?)

Cheers & good luck on finishing it off.

:D

[Alchemy]

your 6 incher has produced some superb images , if its as good as that one, were in for some really good pictures.:thumbsup:

[Garyh]

Thanks, yes I have downloaded that for my last mirror as well as Figurexp and Coudermask. All fantastic freeware.
What a great idea of marking the lines straight on the mirror while testing!
Thanks for that tip! :thumbsup:



Thanks I beveled and done the sides with a lapidary grinding wheel. Finished it off with wet and dry paper (wet) and a block.
Did go a bit overboard with the bevel hey!
Images are taken with a little point and shot digital camera. My DSLR lense is way too big to fit in behind the ronchi grating and slit post.
I bit fiddly but can be done!
cheers Gary



Hey thanks Clive, I will be very happy if I can get a similar figure on this mirror as well!..:)
I will give it a good try!
cheers Gary

[Garyh]

Ok everyone, here is the results from my latest 2 sessions. approx 30min worth.
Looks like its getting close now! Only problem is the very outside cm is still lagging behind..the centre zones look rather good.
I am not quiet sure of how to go about correcting this.
I might try a small tool made on one glass tile 1"x1" and work around the edge a few revolutions? Would this be a good way to brig it down a little or would I be better to use a different tactic with the full lap....
I think another 10 min or so after getting the edge right shall get me really close and time to make a mask up!
I knew I get stuck somewhere!..:doh:
cheers Gary

[Satchmo]

Yes the edge is still undercorrected. What kind of stroke are you using and MOT or TOT ? It is always best to stick with as large a tool as possible: it takes a lot of experience to get smooth results with a 1" lap.

Your 8" F4.5 only has about 4 waves departure from the paraxial sphere or 1 wave at center and edge from the 70% zone , so certainly no need for any drastic measures :) You need to push the curve out to the edge and the centre may raise up in the process. Conditions are fairly cool at the moment so you might consider the option of making yourself a plastic tent around the polishing table and warm things up .

Don't be impatient to get a good figure so quickly. Enjoy the journey.

[Garyh]

Hi Mark,
I have been using MOT and using the www stroke with emphasis on more wear on the outer zones so more over the centre strokes with a wider stroke towards the sides..
Yes has been real cold up here and I have been taking over the dining table at night..:)
Being much warmer last night it really made a difference with the amount of glass removed!
Should I try a different kind off stroke.
Thanks again Gary

[Satchmo]

Gary,

You need to deepen the curve a little and hence push the curve out to the edge. Probably just keep going the way you are but slow steps. It is easier to work the edge with a sub diameter lap TOT. A simple way to achieve this is to press a paper annulus between lap and mirror to reduce your lap diameter to 6" . It is easy to press the lap full again over night when you want to. You would really benefit from having the scope constructed so you could star test the mirror at the end of the process.

[Garyh]

Thanks Mark,
I shall do a 10min session tonight and see if it gets to the edge. I might do more over the centre stokes as well as the centre of the mirror seems deep enough. Shall warm up the house as well to get the lap working better!
Well the scope is just a pile of parts at the moment and still have to get a good focuser. So will have to put the mirror aside for a while till the scope gets finished then give it a test.!
Shall post another ronchi image tomorrow with the results...
I find it quiet exciting figuring and watching the shape change. Certainly is a art this!
cheers Gary

[Garyh]

Hi again,
Done two sessions last night. One for 10min then a shorter one for 5 min with pressing of around 45min with about 3kgs.
Getting closer to the edge but still not quiet there.
Here is a poor image through ronchi grating with a printout beside it from RonWin. Looks rather close by eye just have to get that last 5mm on the edge better before I start testing.
Would a quick few minutes of TOT WWW help or would that be too drastic? or better just keep doing what I am doing?
cheers Gary

[Satchmo]

Gary

The physics of your Ronchi setup means the white lines are much thinner than the simulation, so its a little hard doing a visual comparison as it is. . For the purposes of a comparison I would make up a cardboard `hourglass' mask whose outside edges were the centre lines of the white simulation ronchi bands on either side of the centre one.

As your actual white bands are quite narrow this will give you a pretty good handle on things. At this stage I would work more on clarifying the testing than any more figuring. You'lll find the lines drawn on the mirror with a whiteboard marker easy to see due to them having bright edges due to diffraction effects.

[Garyh]

Thanks Mark,
I think I have seen on the net a program that actually simulates the diffraction effects for a more realistic image.
Shall give that mask a go tonight and post a image tomorrow..a better than my last hopefully!
Thanks again for all your help!...much appreciated :thumbsup:
Gary

[Satchmo]

Its called Diffract and is by Jim Burrows. I was one of the Beta testors. Search Google and you should find it. I once used that method ( simulating the shape of the actual central diffraction lines between the Ronchi bands with Diffract) to figure a couple of tiny 1.3" F3 Parabaloids as part of Gregorian beam diameter reducing telescope for the Sydney University Stellar Interferometer. A pinpoint laser source was needed to give strong clear diffraction lines. As the ROC was so short this was the very best means to do it. Even my interferometer was useless as it needed too much back focus.

[Garyh]

Now that sounds like a real challenge! They say the small mirrors are just as hard as large ones!
Shall have to download Diffract and have a look.

Last night did some testing with marking out the outer ronchi lines of the two inner bands outside ROC as Mark suggested. Tried to match up as close as I could to see how well they match...not too far off...:)
Piccy attached below.
After this I made a quick and dirty couder mask with 4 zones, not including the very centre . Done some quick measurements and put them into FigureXP to see what I have there...
1/4.3 Wavefront!....:thumbsup:

Here are my readings with the ideal reading beside..

Reading: Ideal:

Zone 1: 0.000 0.000
Zone 2: 1.390 1.293
Zone 3: 2.530 2.544
Zone 4: 3.420 3.791

By this it still looks like the 2 outer zones need more work.
The edge doesn`t look too bad in the Ronchi but with the mask the difference is more obvious.
Shall I go and do another 10 min MOT www with most of the stroke over the centre?
I think this should get it real close?..:)
cheers Gary

[Garyh]

Well I have done another 6 min last night of figuring and now have a 1/5.8 mirror!
Still that outside edge is undercorrected while around the 80% zone its slightly over corrected...Inner zones are almost right on!
Shall give it another go tonight and see if it changes.
If not I might have to change my tactics and fixing this...
Stay tuned! :thumbsup:
Gary

[Satchmo]

Hi Gary

Looking good. Unfortunately the mask type and position is not quite ideal. Because the Ronchi program calculates -even- Ronchi spacings ( even thickness black and white lines) yet your setup shows thick black lines and thin white ones.

I recommend plotting the centers of two white bands on either side of a centered white one . In practise then you'll have nice narrow white bands to compare and no error due to picking the edges of the fatter dark ones, which have overbled due to diffraction. . I suspect that the way you have done the mask may produce an undercorrection which is backed up by your KE readings.

[Garyh]

Thanks Mark, that sounds like a better method with the lines marked and the mask. Shall give it a go in the next few days.
I did try some more figuring (trying to fix the edge) but only a little progress.
First 5 minute session with the www just made the center and middle zones overcorrected while the outer zones stayed the same..

Rechanneled the lap, pressed and done some TOT for a few minutes with a 1/3 w stroke which brought it back to slightly under corrected and about 1/3 PV. Done a little TOT figuring for a few minutes which actually improved the edge slightly. Figure XP now gives me 1/7.5 PV.
Tried to find information on figuring with a full size lap TOT but I can`t seem t find any useful info....
Last time I borrowed Jean Texereau`s book from the library but someones nicked it!

Shall give a update and image in the following days..
Cheers Gary

[Satchmo]

Thats because a full size tool on top should push the edge spherical: not ideal. If you have 1/8 wave *wavefront* then you are done.

[Garyh]

Ah that explains why with the TOT it fixed the over corrected center! :P

Tonight I pulled out my mirror for another go.
First I tried something I found on the net and that was TOT working the edge with small W strokes to lower the outer zones. Done this a few laps around but no real improvement, actually made the wavefront drop to 1/6 PV.
Pressed for another hour then done a few minutes the normal figuring strokes but with a few big strokes right over the center and brought the center of the mirror right to the edge!
This has worked great and the edge now looks much better....no TDE and the mirror looks very smooth right to the middle.
Done two tests with the mirror with the coudermask and put the info into figureXP..First lot of data gave me 1/8.8 PV...(20min after figuring) and the second test done 30min later gave me a even better readings and FigureXP now gives me close to 1/20 PV. I shall have to give it a few more tests when it cools down some more to get a more accurate reading.
But it looks like it shall be better than 1/8 PV..:):thumbsup:
Shall get some images posted tomorrow after some tests with the ronchi mask and the coudermask.
cheers Gary

[Garyh]

Hi,
Well finally got around to giving the mirror some tests with a mask like Mark suggested and done some knife edge readings with a 4 zone mask.
Looks pretty good to me...:thumbsup: Edge is still under corrected but with the eye its not really noticable.
Have attached a image which shows 1: lines marked on the mirror at center of white bands..2: knife edge nulled at around 80% 3: ronchi outside COC and a ronhi shot inside COC....Not the best pics but its not easy getting nice pics with my setup..
Mirror looks fairly smooth through the zones but I notice a bright ring/line on the same side as the light source when the outer zones are nulled..
Would this be a diffraction ring? or be caused by the very edge being under corrected? Also the edge of my mirror in my image outside COC be from diffraction? as it doesn`t look quiet like this by eye.
Have attached a image of the results from a average of three readings from FigureXP.

cheers Gary

[Satchmo]

Yes you can see the edge is a little flat ( turned up ) on the foucaultgram. The bright edeg on the same side as the light source would confirm a slightly steeper turn up at the very edge. While I'd always treat wave ratings on a limited sample of points cautiously I 'd say you are done. It is always a good idea to star test before aluminising just to check for astigmatism.

[Garyh]

Thanks Mark for your input!
Always handy to get advice from someone who knows what he is doing!
Just one more question,
Is a TUE as bad as a turned down edge of the same magnitude?
Everyone talks about how bad a TDE is on image quality but not much is mentioned about the effects of a TUE?..
Shall put the mirror away till I get the OTA finished to do that star test! :thumbsup:
Thanks again!
cheers Gary

[Satchmo]

The very narrow flatter edge looks to me very ninimal ( <1/10 surface and the very edge doesn't look that bright. Its hard to calculate such impacts but keep in mund the relative surface area is very small. Zambuto mentioned on his forum that many of their machine finished mirrors have some small turned up edge. I would just give it a star test before you aluminise and etablish you are happy with the infocus image: at the end of the day that is what counts.

[Starkler]

Perhaps much of that could be removed by beveling the edge a bit more :confuse3:

[Garyh]

Thanks Mark, yes it is only a very narrow edge maybe 1mm or so wideand would be only a very small part of the surface area....
I suppose it would be easy to bevel out with care but knowing me I will take a big clamshell out of it!
I shall consider that Geoff if it shows on the startest!
cheers and thanks Gary