Sunday, March 29, 2015

150329 TimC on isolating problem causing scratches

Tim C  8:17 AM (2 hours ago) to Jerry, Mike, Richard, Tim, me, Tom:
   Of course I am very sensitive to this. Let's check Tuesday. I am so sorry to hear of these scratches. I am very confused by all of these scratches. It seems to me we have (as a group) never had so many show up in finishing stages. Coincidence? I am not sure. It only takes one grain of grit to do the damage. The question is, " where is the likely source of this contamination?" How does one try to isolate the source?
As I have told many of you, in the dental lab, when we have a problem ( not necessarily this kind of problem, but similar) we try to list all the components in the production to the point of the problem, then one at a time change a component till the problem ceases.
With the problem at hand, if it cannot be fixed, the problem occurred late in the polishing/ figuring stage. To me, it calls for looking at :
A. The environment the problem occurred
B. The cerium oxide and container it is in
C. The surface you are working on
D. The pitch / pitch lap
E. Cleaning and storage materials and containers
Are there more I have not listed?
If it were me, and boy it has been, I think I would group the pitch and cerium into one group. Otherwise changing just one perpetuates the problem mercilessly, while replacing both and fixing the problem still isolates the cause greatly. I do have a question here though. That is, is it really necessary to change these out now? Can Mike go back to 500 then come forward with the same materials, especially the cerium, as it did no damage for a long time? Is the pitch now contaminated or is it just a matter of cleaning the lap carefully?
If it is likely that this was a random piece of grit that has now been expelled from the working environment, it is just a matter of being careful with work surfaces and all materials.
Again Mike, I am so sorry to hear this and I'm obviously sympathetic. As you know, I am going to go back a sixth time to regrind the donated mirror. I'm going to complete another 8" I have simultaneously. I want to know more of if one piece of Pyrex can be different from another. I still want to know if the cause that persists with this one mirror can come from the blank and not all the other standard components. I did change out all of the others and the scratches persisted. If my methods hold up it seems the problem point to the Pyrex. As Jerry suggests this may be highly unlikely. So that leaves me. Because I work in a very abrasive environment everyday, I may be the source. I may be bringing in contaminants under my fingernails, on my clothes ( which I change before each workshop) or hair.
One last component is possible. Randy, from Astrosystems told me he had to put a plastic sheeting above the work area he was in. The acoustic ceiling tiles were a source of contamination for them. As many in the workshop have commented in the past, "it's too bad we do not have a dedicated place we can keep our projects and even a dedicated clean room to polish."

Saturday, March 28, 2015

150328 MikeC mirror status -- scratches?

---------- Forwarded message ----------
From: Mike C
Date: Sat, Mar 28, 2015 at 5:32 PM
Subject: Re: Am I close enough or should I tempt fate making my mirror more perfect
To: Richard B
Cc: Tim C, Jerry, TomT, Tom W

Latest update:
I managed to get the mirror back to a sphere due to the over correction. We were about to think about parabolizing.   However, at the last meeting we seemed to have a contamination problem and as a result I got one big scratch and a pit on my mirror.  After that I tried to polish out the scratch I brought it over to Tom's house to pick up more cerium oxide.  However, we noticed many more scratches. This Tuesday Tom and Jerry will determine if I have to go back to grinding out the scratches.  This might set me back six months depending on how much work I need to do.  At the very least I'll have to chuck the old pitch lap and polish, but I'm tending to think that I need to go back to 500 grade or worse.  That will mean I'm back to square one which is where I was over a year ago.

On Sat, Mar 14, 2015 at 5:14 PM, Mike C wrote:
Thanks for getting back guys.
I'll be making the Ronchi eyepiece and masks but won't be able to use them until I get back to Santa Barbara as the cloud cover is solid and might not be gone until a day or so.  I might go far as to make another offset circle with a 3.5 inch circle.
In the ;meantime I need to get serious with a more permanent way of holding the optics.  My setup is a handful to keep pointed when the angles are too high and too low.  That's because the board used for elevation needs to be adjustable.

On Sat, Mar 14, 2015 at 11:22 AM, Richard B wrote:
Hi Mike--

Have you ever done a star test on a "perfect" mirror? If you haven't, an easy way to do so is to make a circular off-axis stop about 2 inches diameter for your 10 inch. Use a magnification 1/5 as great, and run a star test. The errors over the small aperture will be small compared to the full miror, and you'll see what a "perfect" image looks like.

The effect you describe -- hard edge one side of focus with soft edge the other side -- is very common. However, the description is qualitative, and that makes it hard to know whether you can ignore what may be a 1/20-wave error or be worried about a 1/2-wave error. I used to take a Ronchi eyepiece (an eyepiece housing with a 100-line Ronchi inside) to star parties, and I looked a lot of mirrors. In general, it's best best to slip in the Ronchi, do your test, and not say anything to the guy who owns the telescope!

I suggest you make a ring-shaped stop with a diameter of 9 inches. Make sure the edge is nice and clean, such as an Xacto-blade cut edge in file-folder cardboard. Just for fun, you could make a mask that cover only 180 degrees of the mirror's rim. If you have a turned-down edge, it probably won't be as much as 1/2 wide, so you see what the star test shows with a non-turned edge.

Another useful diagnostic is to place a Ronchi screen at the focus, and view the mirror surface in the light of a star. With four or five bands, they should be perfectly straight. If the bands bow in or out, the mirror is over- or under-corrected.

The purpose of testing on stars is that "you get what you see." There is no math to interpret or figures to misinterpret. The fact that you see errors in the formation of your star images says your mirror will always perform that well and no better. The Ronchi test on a star is essentially a (not very sensitive) null test; straight bands mean good. You can also use a knife edge at focus, but you also see all the air currents.

Testing at center of curvature is good because you do it indoors under controlled conditions, but you're measuring the first derivative of the figure, not the figure itself.


On 3/14/2015 8:57 AM, Tim Cwrote:
Hi Mike,

Tom or Jerry are the experts here, but, I want to give you my take- what I would generally interpret first. Then, you can check with Tom, Jerry and Richard for more advice and most likely a clearer explanation.

Before I get there though I would like to say I use a program written by Reifke to interpret my Foucault results. I do not know what algorithm he uses in his program but I like the way it shows results. The pin stick I gave you does not use the same settings for pin placements. You will have to manually measure and change these within the program. The output of this program gives you a picture of the wavefront and where to make changes. It looks like this:

The lower horizontal line is a perfect mirror. The output is the line traced from left to right representing the mirror surface. The vertical lines represent the zones. You can visually see where the correction is needed. Both Tom and Jerry ( actually several others as well) tell me you really can't trust this because the algorithm is unknown.

In Richard's book on page 212, you see the pattern in general you describe showing an over correction. To me, that means the curve is too deep. Where I get all fouled up is visualizing if this means you have too long or too short a radius of curvature. I always seem to get it backward. But, I think it means the radiuses are shorter in an overcorrected mirror. For me, I just visually see a deeper curve or dish if you will. It means we need to flatten the curve. 

What I would do, depending on where the errors are on the wavefront is to use TOT and a spherizing stroke in general. 

The hard part here is you sent some images that do not represent this exact picture. I am going to search some pictures in Suiter's book a bit and see if I can't find something a little more definitive . In the meantime, yes some Ronchi pictures might really help here. Try for 4-6 lines inside or out. The other thing is, you can wait and we'll all look at this next Tuesday. 

PS- just for fun see if the pin stick placement measurements on my pin stick are the " effective radius" you see in the picture. I believe these are the measurements you would need to manually use in Reifke's program. ( if you choose to check it out)

Sent from my iPad

On Mar 13, 2015, at 9:56 PM, Mike C wrote:

Hi Tim:

I did center over center last night as per Tom, followed by large w's to blend and got similar results. I'm not sure why things are looking exactly the same in star testing. I see a bright outer ring out of focus and then a diffused circle with bright spot inside focus.  That either means that I've got over correction or turned edge.  I;ve got to figure out again how to get some decent Ronchi grams to show.  Again what does over correction mean with respect to the curve and what part of the curve needs working on?


Friday, March 20, 2015

150320 why Ronchi lines are straight...

>     On 3/20/2015 12:46 PM, Jerry W wrote:
>>     Hello Tom
>>     At last Tuesdays workshop you asked why Ronchi lines are straight for a perfect parabolic mirror when viewing a star but also straight for a perfect sphere on the test bench.
>>     The explanation is this. The primary mirror in our telescopes is a paraboloid because that shape has the property of bringing incident light from a point source at infinity to a single point in the image plane of the telescope. Recall the image plane for an object at infinity is located at a distance in front of the mirror equal to the focal length or one half the radius of curvature. A sphere, on the other hand, has the property of bringing incident light from a point source at the radius of curvature to a single point in the image plane of the optical setup we use in Broder. In this case the image plane is located at the radius of curvature.
>>     This ability to produce a point image from a point object in these two different setups is due to the geometry of the two different surfaces. You may have studied this in geometry or algebra, wherever conic sections were covered.
>>     In terms of wave optics the two different setups involve different geometry wave fronts.  The light waves from a star at infinity has the geometry of a plane wave while that from the point source in our Ronchi setup constitutes a diverging spherical wave front. A paraboloid brings a plane wave to a point while a sphere does it for a spherical wave. After the light bounces off the mirror, if it's heading toward a single point in the image plane, it will constitute a converging spherical wave. This is the case for a plane wave bouncing off a parabola or a spherical wave bouncing off a sphere.
>>     Now when you put a Ronchi reticle in a converging spherical wave the reticle will be the same distance inside or outside of the image plane and what you see will there for be straight lines.
>>     Hope this helps.  Jerry

> On Fri, Mar 20, 2015 at 4:28 PM, President SBAU wrote:
>     Jerry
>     Thanks for the explanation.  Add this to your book of "101 optical explanations for telescope mirror makers"?
>     However, I am not totally clear on why a star is a planar wavefront, that is, why is it different from the test point source which is a diverging spherical wavefront?  Probably because the part of the sphere we see from the star is hardly different from a flat surface by the time we see it in our little telescope?  A star test (coma?) must be the ultimate check on a telescope mirror compared to Ronchi or Foucault tests and guesses.

On 3/20/2015 5:23 PM, Jerry W wrote:
> Hi Tom
> Your supposition of why a star presents a planar wave is exactly correct. Close to the star it is also a sphere, but by the time it gets to us it is a plane for all practical and intensive purposes. The presence of spherical aberration, coma, astigmatism etc are all aberrations on the simple wave front explanation I gave. They are of a secondary nature compared to the primary one of producing a true point. We can get away with ignoring them because when we are squinting through the Ronchi grating with our naked eye we can't see them.  If we were to greatly magnify the image then they would show up.
> Jerry