From: Dale Eason 
To: atm{at}shore.net
Reply-To: Dale Eason 


--0-2070447361-1057515674=:52382
Content-Type: text/plain; charset=us-ascii


I wonder what accuracy is needed for the flip and diff method of finding
null radiuses.

I did the following test with two different mirror data sets.  The first
was my 8" f7.5 and the second was with James published robo test
mirror run AxisA_15.fig.  Using Fixurexp, what I did was to round off the
zone/mask radius to 2 digits.  Then compute the surface errors.  Next I
subtracted .1 from each reading computed the surface error and then added
.1 to each original reading and computed the surface error.  The following
table shows the results.

8" f7.5  14 zones

Unaltered RMS 1/29 Strehl .954

rounded  RMS 1/28 Strehl .953

-.1  RMS 1/27  Strehl .95

+ .1 RMS 1/30 Strehl .959

AxisA_15  15 zones

Unaltered RMS 1/12 Strehl .782

Rounded RMS 1/10  Strehl .753

Minus .1  RMS 1/7  Strehl .575

Pluss .1 RMS 1/15 Strehl .842

As you can see it did not effect the 8"  mirror much at all less than
1% change in Strehl

The larger mirror took a biger hit but not too bad for this much error.  So
I don't think that the flip and diff suffers too much from 1/2 this much
error.

I would be interested in what others think.

Dale Eason


---------------------------------
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
--0-2070447361-1057515674=:52382
Content-Type: text/html; charset=us-ascii

I wonder what accuracy is needed for the flip and diff method of finding
null radiuses.
I did the following test with two different mirror data sets.  The
first was my 8" f7.5 and the second was with James published robo test
mirror run AxisA_15.fig.  Using Fixurexp, what I did was to round
off the zone/mask radius to 2 digits.  Then compute the surface
errors.  Next I subtracted .1 from each reading computed the
surface error and then added .1 to each original reading and computed the
surface error.  The following table shows the results.
8" f7.5  14 zones
Unaltered RMS 1/29 Strehl .954
rounded  RMS 1/28 Strehl .953
-.1  RMS 1/27  Strehl .95
+ .1 RMS 1/30 Strehl .959
AxisA_15  15 zones
Unaltered RMS 1/12 Strehl .782
Rounded RMS 1/10  Strehl .753
Minus .1  RMS 1/7  Strehl .575
Pluss .1 RMS 1/15 Strehl .842
As you can see it did not effect the 8"  mirror much
at all less than
1% change in Strehl
The larger mirror took a biger hit but not too bad for this much
error.  So I don't think that the flip and diff suffers too much
from 1/2 this much error.
I would be interested in what others think.
Dale Eason
Do you Yahoo!?
http://pa.yahoo.com/*http://rd.yahoo.com/evt=1207/*http://promo.yahoo.
com/sbc/">SBC Yahoo! DSL - Now only $29.95 per month!
--0-2070447361-1057515674=:52382--

--- BBBS/NT v4.01 Flag-4