Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
I'll post the pics. First one is a centered pattern, and the other allows the "bullseye" to be offset. The area of interest is in the central 3" 8mm thick glass, A Schott equivalent borosilicate similar to pyrex, blocked up with pitch on a 1" thick aluminum plate.
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
This second pattern shows the fit fairly well, at just slightly concave. About 1 set of fringe colors evident. It also shows an offset pattern, with a slight hook on the edge.
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
The maker of these ~3" secondaries made a groove on the front/polished face, which seems to leave a residual issue with the edge of the groove. I assess the error at about 1/8th wave surface, overall. ROC is a convex 214". Comments welcome!
The testplate is from an old Yolo project. Central 3.5"
of it are just about a perfect null.
====================
I would have preferred a thicker glass to start with, and no groove on the front face. I think that would have helped avoid the edge defect better. I have 2 of these finished. I can post 2 more pics of the second one.
Mark
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
here's the first from the second piece
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
The second pic:
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
These 2 depict a convex fit of about 1 fringe. Still depicts an edge issue, seems to go with the way these were prepared initially. I give this one an approximate 1/8th wave as well. When touching the testplate, there is a certain amount of distortion to account for. In this case, the angle viewed at seems to counter the effect of touching the testplate, and shows a fairly round center to the bullseye.
One of the key things to note, is how just a fringe or so visible from manipulating the test plate makes the edge defect blatantly obvious on this second example. It's "almost" invisible with the many lines showing with the first picture on this second example. Beware of more than about 3-5 fringes visible in any interference test!(IFgrams included!)
Mark
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
Can anyone guess why a convex fit (2nd pr. pics) are not as desireable to judge errors?
Mark
--------------------
So many projects, so little time!
|
Biff
Carpal Tunnel
Reged: 09/04/05
Posts: 2372
Loc: Courtice, Ontario
|
|
Just a WAG (but I love guessing) if you have a concave fit you can at least measure the ROC of the piece in question accurately, and you can be reasonably sure of its quality.
--------------------
Ryan
Antares 200mm f/6 Dob & 130mm f/5 Travel Dob.
Projects on the go...
- a couple 80mm SS refractors on the back burner.
- a few small mirrors awaiting polishing
- 260mm f/7.15 mirror... still polishing
Member of DRAA
 My house.
 DRAACO
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
Most of the time, the task of a smooth regular surface with these smaller items isn't an issue; but edges still are. With a concave fit, the bands curve the opposite way that the turned edges do, therefore far easier to detect.
With a convex fit, the general direction is the same, making it much harder to tell its presence, or magnitude. A little study of these pics illustrates the aspects very well.
Mark
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
You can also find the ROC of the convex example. Having the testplate curvature well established, just add half a wavelength to the sag of the testplate which will result in an overall greater value, and a shorter ROC. Remember, one fringe is half a wave. Testplate ROC is 5443mm, and after going through the whole mess, I get 5430mm for the convex fit secondary.
M.
--------------------
So many projects, so little time!
|
Mike I. Jones
Carpal Tunnel
Reged: 07/02/06
Posts: 1572
Loc: Fort Worth TX
|
|
Hi Mark,
I built a sodium interferometer that might be of interest to you. It gives much higher contrast fringes and a normal incidence view of the part under test. I used a piece of 1/4" thick 50/50 glass from PPG that's used for office buildings to reduce light and heat into the building. The uncoated back side faces up, and the coated side faces down, giving a ghost-free image. The low-pressure 18W sodium light and ballast cost about $65, the glass was about $35, and the rest is scrap plywood.
Here's the design, then I'll post the pictures.
Mike
--------------------
56 mirrors, lenses, 16" f/6 Newt, 6" f/10 refractor, TOA-130S, Tinsley 5" f/15 Mak, 6" f/4 RFT, Coronado PST. Still to build: 24" f/10 Modified Dall-Kirkham, 10" f/26 Mak, 8" f/12 apo, spectrohelioscope, Herrig, Schupmann, and others.
|
Mike I. Jones
Carpal Tunnel
Reged: 07/02/06
Posts: 1572
Loc: Fort Worth TX
|
|
This is the finished interferometer. I can put 24" diameter parts under it. This is before I put the rubber mat on the test table.
|
Mike I. Jones
Carpal Tunnel
Reged: 07/02/06
Posts: 1572
Loc: Fort Worth TX
|
|
This is what the fringes look like. The fringes have very high contrast due to the very narrow line emission from the low pressure sodium light. 18 watts doesn't sound like much, but when it's fully warmed up (takes about 10 minutes), it is blinding to look at, and plenty bright through the white Plexiglass diffuser and 50/50 beamsplitter glass for photography.
This particular part is just a dummy shined 5" Pyrex piece I grabbed to take the picture, tested against a 1/20 wave 6" master flat.
Mike
|
DAVIDG
Carpal Tunnel
Reged: 12/02/04
Posts: 1990
Loc: Hockessin, De
|
|
One can also filter the light from a CFL light bulb to obtain a monochrome view or image by looking through a filter or holding it in front of the camera. Eyepiece filters work well and so do just inexpensive plastic gel filters used for stage lighting.
Mike's box with the beam splitter window is a very nice feature since it allows one to view the fringes straight down and on the optical axis. If you view the fringes on an angle there will be false curvature introduced.
- Dave
--------------------
Homemade 'scopes 8"f/7,6" f/5", 6"f/4, 4.25" Schief. 60mm Coronagraph,60mm H-alpha system, 4.25" White-light Solar Newtonian,solar spectroscope, 4.5" f/16 Schupmann Medial refractor, 14 Stellafane awards 7 in optics
Engineering = Taking what you have and making what you need.
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
"One can also filter the light from a CFL light bulb to obtain a monochrome view "
==========================
That's what I used, though unfiltered. My camera shots are fairly good, but not what really is present with naked eye.
==========================
"If you view the fringes on an angle there will be false curvature introduced."
Yes, as I noted above. Practice makes perfect in such a situation.
================================
Nice light box, Mike! Maybe you & Kevin will come up with something big enough to place in there!
M.
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
I like to see the colors, personally. If you note the last pic, what normally borders the pink as you get a thinner airspace is yellow, to green. What's visible nearest the center, is an orange, and the very center; almost clear. This signifies that the parts are getting very close. If you fool around and press indescriminately, you can contact the testplate to the part. If I recall, the attraction is about 3x atmospheric pressure. You'd never get them separated by hand. These "first series" colors are always very stark and obvious. Simply the colors alone can be used to judge the accuracy of the fit. I remember the tolerance exibited when the colors just disappear, and it's clear prior to contacting- 1/10th wave or better if I recall correctly.
M.
--------------------
So many projects, so little time!
|
kfrederick
professor emeritus
Reged: 02/01/08
Posts: 669
|
|
here is my grain bin fringe tester
|
kfrederick
professor emeritus
Reged: 02/01/08
Posts: 669
|
|
here is the light source
|
kfrederick
professor emeritus
Reged: 02/01/08
Posts: 669
|
|
just a joke i started seeing fringes in the windows after building the bin
|
DAVIDG
Carpal Tunnel
Reged: 12/02/04
Posts: 1990
Loc: Hockessin, De
|
|
As one knows the spacing bewteen fringes is 1/2 the wavelength of light used. With pure monochrome light one knows the exact wavelength so one can determine accuractly the difference in radius between the test plate and the piece being tested. In Mike's image, he is using Sodium light at 589nm. If one uses a multi-wavelength source, what wavelength do you use for the calculation ? It's not the average of the different wavelengths since if it was one would only see one color in the test, but your seeing multiple colors. For example there is no pink emission from a monochrome source so this color is some combination of wavelengths but which ones ?
Also the shorter the wavelength of light used the more accurate the test. For critical flat work the Helium emission at 471nm is preferred over longer wavelengths like Mercury at 546nm or HeNe at 632nm. One can see finer departures from flatness since the wavelength is shorter.
By filtering the light coming from a CFL bulb by using a simple filter, like an eyepiece filter of piece of colored plastic (red, green or yellow), one will improve the accuracy of the test by defining the wavelength of light used and increasing the contrast of fringes so errors are more easily seen.
- Dave
--------------------
Homemade 'scopes 8"f/7,6" f/5", 6"f/4, 4.25" Schief. 60mm Coronagraph,60mm H-alpha system, 4.25" White-light Solar Newtonian,solar spectroscope, 4.5" f/16 Schupmann Medial refractor, 14 Stellafane awards 7 in optics
Engineering = Taking what you have and making what you need.
|
Mike I. Jones
Carpal Tunnel
Reged: 07/02/06
Posts: 1572
Loc: Fort Worth TX
|
|
Quote:
Mike's box with the beam splitter window is a very nice feature since it allows one to view the fringes straight down and on the optical axis. If you view the fringes on an angle there will be false curvature introduced.
A regular piece of 1/4" plate glass can be used for the beamsplitter, but the upper and lower faces have the same reflection brightness, creating two images of the fringes, one shifted above the other slightly. Using the 50/50 window glass with the coated side down essentially eliminates any ghost image visibility. It really makes a difference in fringe contrast. I got two pieces from the local PPG store in Fort Worth, one for me and one for Walt Goldys. Walt uses mercury lights and a green filter rather than the sodium light I'm using. Both Newtons give nice contrasty fringes.
Multicolored fringes have much lower contrast, as the interference periods overlap each other. Fluorescent lights can give very faint green fringes when no other light source is available. The wavelength used for analysis is simply whatever the dominant color of the fringes you're using. Low contrast multicolored fringes are OK for visual work but not for CCD-based fringe processing.
I like either the 589nm or 546nm lines best for the Newton interferometer. Without the green filter, Walt's Hg Newton does show the violet fringe patterns along with the more easily visible green fringes, but the green is so much brighter you don't notice them visually. A DSLR sees the violet fringes clearly though, and you have to use a green filter if you plan to photograph them. The Na fringes have very high contrast with no filter needed.
Mike
|
Mike I. Jones
Carpal Tunnel
Reged: 07/02/06
Posts: 1572
Loc: Fort Worth TX
|
|
Kevin, that window must be flat to within about 1/4 grain bin! 
I like the phase shift in the grainbin fringes where the window is cracked.
Mike
|
Gary Fuchs
scholastic sledgehammer
Reged: 05/22/06
Posts: 867
Loc: Easton, PA, USA
|
|
Mike,
Those are really nice vivid contrasty fringes. And a lot safer, better and easier setup than a salt flame. Could you give us a link to the light and ballast?
Thanks,
Gary
|
Biff
Carpal Tunnel
Reged: 09/04/05
Posts: 2372
Loc: Courtice, Ontario
|
|
I think I'll have to find a smaller light, smallest Ive come across is 150W. We have some of those at work but it appears it'd be miles too much.
--------------------
Ryan
Antares 200mm f/6 Dob & 130mm f/5 Travel Dob.
Projects on the go...
- a couple 80mm SS refractors on the back burner.
- a few small mirrors awaiting polishing
- 260mm f/7.15 mirror... still polishing
Member of DRAA
My house.
DRAACO
|
kb4mxo
member
Reged: 07/17/06
Posts: 58
|
|
Is the glass you are using a LowE coating? It the glass 6mm thick? I think the place I work for makes a glass like it. What it will do is let light pass with a loss of 1% and will reflect it back if it hits the coating. I am using this in my indirect passive solar room.
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
Dave if you'll pay a closer note, the 2 basic colors exibited are a greenish blue, and a violet. You can assign the matching values for the wavelength when doing the calcs. This doesn't apply with the true Newton's fringes, or the primary pattern that applies with near perfect or convex fits. This is where you'll see -ALL- the colors highly reinforced; and they can be used directly to assess errors in the surface fit. It takes practice/experience to do the latter accurately. When you can do this without scratching either piece, you might have enough!(and a -GOOD- thing, too!) 
Mark
--------------------
So many projects, so little time!
|
DAVIDG
Carpal Tunnel
Reged: 12/02/04
Posts: 1990
Loc: Hockessin, De
|
|
Mark,
The issue is that both the greenish blue and violet fringes are not pure monochrome wavelengths but are caused by overlapping fringes. The eye combines the two colors to make the colors you see in the fringes. The problem with this that the fringe spacing is no longer 1/2 the wavelength of light. The width of the fringe is now a combination of the two or more wavelengths. Fringe width at a given air gap varies with wavelength. So if you used 632nm red light you would only see a few fringes compared to the same air gap and you used 471nm blue light. You don't know what percentage of the width of the fringe is coming from one wavelength vs the other, so there is big uncertinity in the accuracy of the measurement. This is why an interference test is done in pure monochrome light with a known wavelength. CFL bulbs have around 7 to 15 different emmission lines which cause overlapping fringes.
My drawing hopefully illustrates my point that the blue-green fringe in your pictures can be made up of the emmission line at 546 and one at 471nm but the width of the line is not 1/2 wavelength at 546 or 471 and is not a 50/50 split between the two wavelengths. By simply filtering the light source so one only sees one wavelength all this uncertainty goes away.
- Dave
--------------------
Homemade 'scopes 8"f/7,6" f/5", 6"f/4, 4.25" Schief. 60mm Coronagraph,60mm H-alpha system, 4.25" White-light Solar Newtonian,solar spectroscope, 4.5" f/16 Schupmann Medial refractor, 14 Stellafane awards 7 in optics
Engineering = Taking what you have and making what you need.
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
I have a couple more pics of the last one in the series. This one's pretty good.
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
The last pic. Also, of the polisher that did the work on the three.
--------------------
So many projects, so little time!
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
Nice, round bullseye, good bands out to the edge, etc. The way it should turn out! As a group, all are certainly easily passable. There is still a very slight edge issue, and 2 ROCs evident between the center and outer ring. One fringe convex fit.
Dave if you want to split hairs with calculating whether it's 5430mm or 5435.102mm, knock yourself out. I'll gladly leave it to you. In this case, it's not necessary, has to be about .5% difference of the primary's curvature, if I recall the mfr's specs correctly. (I'm within 1/2-2/10%~) You should -NOT- make the calculation on anything but the primary fringe pattern where the surfaces are closest for best results. If you learn what all those colors mean with CFLs, you'd add a handy tidbit to your testing bag of tricks!
Later, and enjoy,
Mark
Had to edit the tolerance,sorry. I think I'm getting the SWINE FLU! Actually, kissed the wife when she had a headacke a couple days ago. I'm paying for it now.
--------------------
So many projects, so little time!
Edited by Mark Harry (10/07/09 04:03 PM)
|
Mark Harry
Post Laureate
Reged: 09/05/05
Posts: 3128
Loc: Northeast
|
|
Been setting on the table about 2 hours as I left it this afternoon. Oh, yeah!
--------------------
So many projects, so little time!
|
Previous
Index
Next
Threaded
Edit
Reply
Quote
[Re: 