From: Mark Holm 
To: atm{at}shore.net
Reply-To: Mark Holm 


I have been also fighting a TDE.  I think I am making progress.  It was
pretty bad, and has taken a lot of work, but the tests slowly move in the
right direction.  I have decided that the only way to really know (for me
at least) what is going on at the very edge is to make a special edge mask
to use with the
Foucault test.  As I get close to correct figure, I use a total of 10 zones on a
204 mm clear aperture mirror.  Some may think me over zealous with the
multiple mask zones, but I have a good quantitative estimate of what is
happening to the mirror all the time instead of guessing whether that
shadow looks good, bad or indifferent.

Here are the mask zones I am using.

Zone
Rinner
        Rcenter         Router
1
0
        21.875          35.722
2
19.431
        38.426          53.538
3       35.993
        50.131          62.502
4       53.669
        63.919          73.405
5       62.502
        71.906          80.731
6
73.405
        81.316          88.860
7
80.731
        88.292          95.543
8       88.855
        92.162          95.410
9       88.860
        95.543          102.00
10
95.410
        98.732          102.00

Rinner and Router are the dimensions the mask is cut to.  Rinner is the
number to plug into the Foucault reduction program.  (I use Sixtests.)

Zones 1, 3, 5 and 7 are on one mask. Zones 2,4,6 and 9 are on another.
Zones 8 and 10 are on the third.

Notice that zones 8 and 10 are 1/2 as wide as zone 9, and cover essentially
the same territory.  The narrower zones are a bit harder to see clearly,
but have the advantage that they are not so affected by strong zonal
irregularity (turned
edge).  I found it very difficult to get a reliable zonal position for zone
9 with the turned edge present, even when it has been getting milder.

Zones 2,4,6 and 9 have exactly the dimensions resulting from Nils Olof
Carlin's web zone calculator.  (Also as calculated in my Couder mask
program.)  Zones 1,3,5 and 7 are calculated by my Couder mask program to
follow Nils Olaf's recommendations but position the zones
"between" the zones given by his algorithm.  Zones 8 and 10 were
made using the editing feature in my program to force the zone boundaries
to be where I wanted them.

With all these zones measured near the edge, I find I get a pretty good
idea what is happening there, and don't have to guess.  It takes a few
minutes longer
each time I test, but I think the higher degree of certainty is worth the
time and effort.

Mark Holm
mdholm{at}telerama.com

--- BBBS/NT v4.00 MP