From: "James Lerch" 
To: "ATM List" 
Cc: 
Reply-To: "James Lerch" 


Greetings All,

First off, let me paint a mental picture of my current intellectual state.
In an episode of Star Trek the Next Generation, the android Data is
processing some huge amount of information, during which time he is
absolutely motionless and just keeps audibly repeating
"Processing..... Processing".  Well that's me at the moment!


I apologize for not responding to all the responses yet, I will, and I
promise, as there's an incredible wealth of information to explore!

Further more I really appreciate the level of feed back and suggestions! 
In the words of Carl Zambuto this afternoon "James, you are getting
replies from some serious 'Heavy Hitters'!"  Consequently I humbly
appreciate the work that goes into all the various replies, THANK YOU!!!

NOW, for some interesting data that just happens to relate to Michael
Peck's most recent post by pure coincidence.

This morning's test setup:

#1 Replaced the $1K digital camera and its crappy 5bit Digital to Analog
converter used for its Analog video output with my "Normal" $50
8mm Cam Corder (Btw, this isn't a web cam, just a 1993 era 8mm hand held
video camera) (PS, I have thoughts of using the High resolution
capabilities of the 'Fancy' digital camera, however it will result in first
saving the image to the camera's memory card, then downloading it into
Robo, which will effectively reduce my 30 frame per second video feed down
to about 1 frame per second, if I'm lucky..)

#2 Converted the Slitless setup to a Slited light source by placing a second
razor blade over just the light source.  (Thank you Nils!)

    A)  I can't begin to estimate the slit width, but it definitely makes a
visible change in the Foucault images. (however it doesn't really change
the results by much..)

    B) Initially the slit was so narrow, I couldn't get the LED bright
enough to get a decent signal thru the camera.  After burning up my old
trusty Blue LED by pumping too much current thru it, I replaced it with a
'Super Bright' White LED running at 110% of its maximum current.

    C) At this point I STILL couldn't get enough light into the camera, so I
carefully adjusted the slit width until I was reading a pixel intensity
value of ~240, which was used for all the test results below.

    D) Using this slitted light source, and fully illuminating the mirror by
backing out the knife edge, I got a 'fairly' uniform light distribution
across the mirror.  By 'Fairly' uniform (with knife edge on the right) the
left zone intensities all read near 250, while the right zone intensities
all read near 245.  (No idea why, but I hope it won't cause too much of a
problem...  Later experiments will explore the possibilities of doing a
'Flat Field')

    E) With the mirror fully darkened by cutting the knife in all the way,
all zone intensities went to ZERO while using a 30 frame running average
(that's got to be a good thing I hope )

#3  For this series of test runs, the only variable was the "Shade of Gray"
value used in the software.  Which, as Michael Peck confirmed, is in
essence setting the lateral position of the knife edge with respect to the
optical axis.  (PS, one item I failed to document was the Lateral position
measurement for each zone null, something I should do the next time I run
this experiment..)

The results of the test were interesting, and did not match Michaels'
predictions exactly, but this could be a result of noise or Air turbulence.
In addition, due to time constraints, I only ran one set of knife edge
readings per experiment, so I can't quote a standard deviation for each of
the test runs, something else I'll need to do a better job of during the
next experimental run!  (I'd give strong odds that the StdDev is in the +/-
0.001" range or better, based on previous test apparatus accuracy in
repeatedly finding the zone nulls.)

So, what are the results?
http://lerch.no-ip.com/atm/gray.xls (29KB excel file)

For those that don't want to open an Excel file (who could blame you!)
Here's a synopsis for total knife travel from Zone1 to Zone7 for each of
the "Shade of Gray" test ranges.  (SoG = Shade of Gray range, and
Max pixel intensity possible was ~245ish on fully illuminated mirror)

SoG = 20-40 = 0.130"
SoG = 40-60 = 0.128"
SoG = 60-80 = 0.132"
SoG = 80-100 = 0.130"
SoG = 100-120 = 0.131"
SoG = 120-140= 0.135"
SoG = 140-160 = 0.133"
SoG = 160-180 = 0.142"

What I find interesting about these results is there is almost a 'Rhythm'
to them.  I wouldn't be surprised to find I may have actually measured the
minute peaks and valleys of a diffraction wave form..

So, wanna see some pictures? :)  Below are the images obtained from each of
the test runs, all are a little less than 900KB, and the title indicates
the SoG that was used during the test..

http://lerch.no-ip.com/atm/20-40.zip
http://lerch.no-ip.com/atm/40-60.zip
http://lerch.no-ip.com/atm/60-80.zip
http://lerch.no-ip.com/atm/80-100.zip
http://lerch.no-ip.com/atm/100-120.zip
http://lerch.no-ip.com/atm/120-140.zip
http://lerch.no-ip.com/atm/140-160.zip
http://lerch.no-ip.com/atm/160-180.zip

BTW, the smudge near the center of the mirror is a "Doggy Nose
Print" (Carl, I promise I'll wash it off before I ship the mirror
out!)  However, the nose print doesn't appear to interfere with the test
results.  Also, notice the almost Ronchi band like apparitions that now
appear as a result of using the Slitted light source (very interesting in
my opinion..)

If you made it this far, congrats!  I know its a lot of data, and I sure
have gotten long winded.  However I pretty much find myself not getting
much sleep lately as I battle with hopefully finding the 'problem' with
Robo.

After repeating the above test using multiple knife edge readings per test
and documenting the lateral position of the knife edge for each null zone
found, I have one more 'Trick" to try with Robo.

This trick goes like this:

When Robo proclaims a Null zone via its current routine, a new subroutine
will be fired off:  This new routine will do a 'Box' search that goes like
this:

#1 For each classic Robo zone null, the software will explore +/- 0.015" in
longitudinal travel (in 0.001" increments) from the proclaimed null position.

#2 At each  0.001" of longitudinal travel, the full range lateral travel
will be recorded so that both zones start with a pixel intensity of 0 and
the lateral travel will continue until both zones exceed an intensity of
200

#3 once all this data is recorded for all 7 zones, a comparison will be done
looking for the Longitudinal position with the least amount of intensity
deviation over the 0 - 200 intensity range.

If I thought this thru correctly, this should simulate the
"Blinking" version of the classic Couder mask (I hope!)

Any Thoughts??????

Take Care,
James Lerch
http://lerch.no-ip.com/atm (My telescope construction,testing, and coating site)

--- BBBS/NT v4.00 MP