[ATM] Re: Figuring Help Request: 6" F7 - Which Stroke Next?

[Mark Holm <mdholm@telerama.com>]

1. Don't even think of putting it back on the lap until you do more 
testing.  If your numbers are to be believed, the mirror may be very 
fine indeed.

2. I ran the numbers through Sixtests just to confirm that the results 
matched those calculated by Andreas' program.  The agreement is very 
close.  I am not surprised because Andreas said he used the same 
mathematical treatment that Jim used in Sixtests.  The one caveat is 
that Andreas reports the P-V and RMS errors as the wavefront values. 
Sixtests reports RMS as the surface value.  Thus, Sixtests' result is 
half of Andreas' number.  There is nothing wrong with this, so long as 
you know how to interpret it.  In either case, these test results, taken 
alone, indicate a very fine mirror.

3. Unfortunately a single set of Foucault values in 4 zones is not 
enough to decide the quality of the mirror.

4. First, in a Foucault test without a mask, does the surface appear 
smooth, or lumpy?  A smooth surface will show the shadows grading evenly 
from light to dark, without local brightenings or darkenings.

5. Is the edge good?  The zonal Foucault test is not too good at picking 
up edge defects.  Again use the unmasked Foucault test.  The edge of the 
mirror should be uniformly bright all the way around when the knife edge 
is adjusted to null the outermost zone.  Try putting the mask on, 
nulling the outer zone, then, without moving anything else, remove the 
mask and look at the shadows.  There are some good images somewhere on 
the web that show what you do and do not want to see.  Right now, I 
can't find them.  Perhaps another list member can tell us where they are.

6. Was the mirror at thermal equilibrium?  Many atm's have reported 
different Foucault test results after letting their mirrors equilibrate 
for a couple of hours.  Put it on the test stand during a time when you 
can expect the temperature in the room to remain stable for a couple of 
hours.  Come back after two hours and retest.

7. One set of test readings usually isn't enough.  While you have the 
mirror well equilibrated, do at least four sets of readings.  Andreas' 
program will average them for you.  The average of a set of four should 
have roughly 1/2 as much random variation and should be considerably 
less likely to be thrown badly off by one outlier.  Analyze the data as 
single sets, and as the average of all four sets.  If the results for 
the individual tests are all over the map, it may imply that something 
about your test setup isn't repeatable enough.  If all four give similar 
results, and the average gives good numbers, then you can be reasonably 
confident.

8. Get a copy of Harold Suiter's book "Star Testing Astronomical 
Telescopes"  http://www.willbell.com/tm/tm5.htm   Build the mechanical 
parts of your scope.  Install the mirrors, do a good alignment 
http://www.atmsite.org/contrib/Carlin/collimation/  and do the star 
test.  You can do this before the main mirror is aluminized.  If the 
mirror passes this test, you can send it off for aluminizing with good 
confidence that it will perform well.

-- 
Mark Holm
mdholm@telerama.com

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