Re: [ATM] My main testing method ..

["Jerry" <wa4guu@verizon.net>]

You are using the edge of a Ronchi line as a knife edge. So why can't you
take the same measurement with the Foucault knife?

Your basic procedure is ok. Smoothing the bend of the Ronchi band is a
sensitive and good way to figure. That is what I do. It helps to measure
zones too. You can have a smooth bend to the Ronchi bands that is not the
proper bend and still have the edge and center knife measurement right.
Measuring a few zones will confirm that the rate of change of the zone
measurements is right across the curve.

I guess that you didn't trust the Foucault measurements early in your mirror
making and have worked around it, not realizing that you are taking a
Foucault measurement.

I have heard that Tom Cave said that if you could make the curve smooth by
Ronchi and get three zones right by Foucault that you could make a very good
mirror. For mirrors not too big and fast I agree with that and think that if
you have done it enough to know what the parabola looks like with the
Ronchi, that you only need to measure two zones. By knowing what the
parabola looks like I mean that the bend of the band is not uniform from
center to edge and the degree of bending at various locations depends on
distance inside or outside of the ROC. 

So overall your method is ok. What I am not sure of is your measurement of
the center and edge zones. I would trust your measurement of the center and
edge more if you used a knife rather than a Ronchi band. And I would trust
your measurements more if you had experience measuring zones with a mask. If
you had that experience of measuring by Foucault and Couder screen and
smoothing the curve by Ronchi, I would trust your ability to measure well
enough the way you are doing it.

If you are good at measuring zones you would have no reason not too. I use
much the same method as you except that I measure the zones with and without
a mask.

I put correction in and smooth it across the mirror by looking at the Ronchi
bands. I look at it and measure with Foucault. I will then put more
correction in if needed taking into account how it went before, adjusting
the stroke to see if I can get the correction to go on smoothly across the
curve so that less local correction is needed. Then Ronchi and Foucault test
again smooth any kinks in the bands. Continue until 100% corrected. I try to
get the knife movement from each zone to the next at the proper number. I
don't pay much attention to the correction from center to edge. It is the
correction from one zone to the next that I get right. If each of those is
right the total will be right too. And the Ronchi band will bend right too.

There is much on the internet that discourages ATM's from trusting Foucault
measurements. Too much. Don't let those who did not do it enough themselves
to learn it talk you into joining them in burying their heads in the sand.

Jerry





-----Original Message-----
From: atm-bounces@atmlist.net [mailto:atm-bounces@atmlist.net] On Behalf Of
vorblesnak@peak.org
Sent: Tuesday, October 02, 2007 4:32 PM
To: atm@atmlist.net
Subject: [ATM] My main testing method ..

I have been wanting to throw this out to the list and get some discussion
on it.

I use a very coarse ronchi grid, 50 lines to the inch, but I don't use it
the way the gurus of ronchi use it, I don't match curves.  I use the lines
to smooth the mirror, no zones, turned edge is identified, (actually I
could just tell you one was there, after all I polished the mirror),
regular curve from edge to center.  I also look at the sharpness of the
lines to gain some idea of the smoothness of the surface.

Now I pull it back to the ROC and look for that moment when the lines at
the center begin to diffuse into a gray zone, a zone of confusion.  I
concentrate on just the middle of the mirror and watch for that subtle
change.  That is line one.  Now I do the same thing to the edge of the
mirror on the vertical center where the lines are easiest to see.  That is
line two.  The distance between should be r squared over R.

This is very much like a Foucault test but I could never read the shadows
of the Foucault test well enough to figure the mirror.  I know, I know,
some of you will argue that I still can't read the shadows.  Using the
ronchi this way allows me to read alot of data from the mirror and still
get pretty close to a precision test.

So, kick it around a bit.  Peck holes in it and give me some opinions
please.

David Davis
Toledo, OR 97391

_______________________________________________
ATM mailing list http://www.atmlist.net/

_______________________________________________
ATM mailing list http://www.atmlist.net/