James,
Thanks for the GREAT INFO...
It's going to take me a while to digest all that and check out the
links but it's nice to have on file here to refer back to!
OK on your diff pump size... I thought from your photo of the mobile
coater that was about the size you had but it could have also been a
4 inch pump in which case I might have been able to help increase
the size of your system a bit.
I'm curious why you are sticking to a filament above the glass and
taking a risk of sputter damage or as before, drops of molten metal
hitting the glass?
My tank is 24 inches diameter and approx 30 inches tall. There's a
photo in my junk folder and I'm still not into a house yet so have
made NO PROGRESS in regards to getting anything done. Maybe by
Christmas???
Again, thanks for the info for the group!
Ken Hunter
--- In VacuumX@..., "James Lerch" <jlerch1@t...> wrote:
----- Original Message -----
From: "Ken Hunter" <atm_ken_hunter@y...>
Hi James...
Hi Ken and All <g>
Have you done any "fall-off" experiments to see just how much ONE
FILAMENT will coat?
When I first started playing with my chamber I had one filament
about 10" above
a 10" diameter mirror. It took two consecutive evaporative runs
to get an
opaque coating all the way to the edge. With just one run, the
center ~30% was
opaque and the last 1/4" of edge made for a nice beam splitter
(maybe 75%
coated.)
After the "catastrophic drip" incident, and subsequent re-
grinding, polishing,
and figuring to "erase" the drip damage (It's soo nice that glass
is so
forgiving!) I re-built the chamber to a multi-point evaporative
setup (with the
filaments out by the edge, so if another drip happened it would
miss the
mirror!)
To help me figure out the filament layout, I wrote a near field
evaporative
simulator. I've had great results playing and working with it.
If you haven't
seen it you can download a copy from here:
(9KB Exe only)
The above code can simulate mirrors upto 24" in diameter, and have
3 various
radius evaporative rings, with as many filaments as you'd like on
each ring.
(Careful! this can turn into a serious CPU Hog!)
BTW, my simulator is based on the work of Jim Hill who wrote the
code to figure
out the layout for the In-Situ aluminization of the 6.5m Mt
Hopkins MMT mirror,
See link for some Fascinating reading!! <g>
www.harvard.edu/cfa/oir/MMT/www.mmto.org/MMTpapers/pdfs/tm/tm03-8.pdf
I ran my near field sim on their data (Scaling everthing down),
and if you refer
to the above figure 3, they report a 1.6nm deviation, and my sim
shows a 1.8%
deviation, which turns out to be 1.69nm with a coating thickness
of 94nm (their
target). OH, the reason for the slight difference? I model the
mirror as a
sphere, since in usual ATM hands the difference in Path Length
between a sphere
and a parabola is really small :) Here's a screen cap of my
simulation results:
(119KB)
I've also seen 24 inch mirrors done with the (one) filament
offset
half way and the mirror rotating to reduce the filament
requirements.
You can use my simulator to simulate this arrangement. Enter the
following
Mirror Radius = 300
Mirror Roc = "whatever your test subject is"
Ring 1
# of evap points = 10 (or more, depending on how long you want to
wait for an
answer)
Evap point Rad = 150mm (50% radius)
Height of EVP (height above mirror center)
Ring 2 & 3, set # of evap points to zero.
BTW, I played around with this, and to make it work reasonably
well means your
going to need a fairly tall chamber!
That is what I'll probably do in my chamber for the
larger sizes and just try centering everyting for the smaller
sizes... What have you tried? What works, what doesn't?
I know for certain a single point doesn't work to great unless
your chamber is
really long.
So far, my 6 points with a radius of 140mm about 120mm above
mirror surface,
works really good on 10" F/5 there abouts.
What's interesting is if I change the height difference (move
mirror closer to
filaments) I can actually add a "little" extra correction to a
mirror. Even
more amazing is Robo is able to measure the change in figure, and
it agrees with
the near field simulations (were talking a VERY small change, like
a 2nm or so!)
How big is your Diff pump?
6" inside diameter, with 4 electric heaters at the bottom, of
which only 3 of
the heaters work (I burned one out thinking the pump was 220V,
when it was only
110v (ooops!). Still pumps my 12"id x 11" tall chamber down in
under a minute
(once the rough pump is done) I may have to replace that bad
heater once I get
the Larger chamber finished, but I'll wait and see :)
Take Care,
James
