Adaptive optics for a small telescope

Thanks everyone. I learned a lot since posting my message on Sunday! See
my remarks below.

West Coast Engineering: The price to the user should be less than $1500.00
for an entry level
system.

It is possible for an "entry level" (F/10, 10 to 19 channels) correction
unit. Sensor camera, electronics
and computer will add a bit

Steve: Is it worth doing anything but tip-tilt in a small system (0.5
metre) ?

Yes, I think so. The right question would be - is it affordable?

Clif Ashcraft: I would be interested, assuming the cost ended up not much
more than a
good CCD camera.

You may need a good camera just to run the AO system.

Chris L Peterson. At an aperture of a meter, you can just begin to achieve
true adaptive
optics- and at that point, only with a few channels. Otherwise, there
simply isn't enough light available for feedback to achieve the
necessary correction rates. for a 25cm scope, you will at best manage a
simple first order correction.

The total amount of light ~ D^2 (D is the mirror diameter). The optimal
number of channels ~(D/R_0)^2. = Amount of light per channel ~ R_0^2.
This basically means that small system needs a small number of channels, but
the available light per channel is the same for a small and a large
systems.

Louis Boyd: Unless you give a dollar value "quite expensive" doesn't mean
much. At
$1k (unit and software, user supplied computer) they'd sell to the
masses. (I'd buy one just to play with). At $10k some would sell. At
$100k few if any who would want it would have the money.

I think $5k for a complete setup without computer.

Ian Anderson: There are two pieces of advice I can offer:
First, keep the final cost of the system under $10 000.
Second, you should try to make it work with F8 systems because most off
he shelf Ritcheys are in this range. If it only woks with F10 or
onger then you will exclude most of your prime market.

F/8 is possible, optics will be more complicated. The argument for F/10 is
simple - AO limits the field forcing F/20 to F/40 to get a good scale per
pixel anyway (AO will improve the resolution so the image must be scaled),
then F/10 looks fine.

Aidan Karley: I wouldn't waste effort trying to make the kit fit into a
1.25" form
factor - most serious amateurs (if you're talking about metre scale
telescopes, then you're talking about quite seriously committed amateurs)
will
have taken the plunge to use 2" eyepieces etc.

The field is going to be small, so 1.25" socket looks OK. There are adapters
for 2" to 1.25"

Louis Boyd: Long range surveillance cameras and spotting scopes are two
that come to
mind. The scintillation distortion looking horizontally though 1000
meters of air in the daytime is often worse than looking at stars though
100km of atmosphere vertically at night. Bigger optics don't help much
or daytime observation as the telescopes are neither limited by photon
rate or the telescope's optical characteristics.

Extended object WF sensor is a complex and expensive device. But then solar
astronomy comes to mind. You can check
http://www.noao.edu/noao/staff/keller/irao/ for some images obtained on an
adaptive solar scope using our deformable mirrors.

The discussion on liquid lens

So far the "liquid lens" technology is not very usable for the AO purposes.
Liquid crystal AO can be much more useful for astronomy as it can be
extremely cheap, but so far it still has some unresolved problems - it is
slow, polarizations sensitive, has strong chromatic aberrations.

Dan McKenna

CMOS based sensors are in a fast lane to low noise performance.
It may be possible to obtain cmos sensors with machine vision processors
integrated in the near future.

I would start by using a bimorph to correct the static mirror errors
first. You would have a knob for tip/tilt, defocus, astigmatism X any Y
etc. That would allow people like me with aging eyes to be able to enjoy
not wearing glasses at the eyepiece.

We have such a system, not for astronomy yet. Look for my name in
the 1 April 2005 (no joke) issue of Optics Letters. Email me for a PDF
reprint.

Membrane mirror is also a curvature corrector, easy to use with curvature
sensors.

Dave: This has been the trend for decades now. The professionals come up
with some
idea for their requirements. Some amateurs with enough resources to try it
for
themselves find a more economical way to do it, but there is either not
enough
interests in the possibilities or enough amateurs with the right resources
to
copy the attempt. If both of these problems are overcome, then it starts
making
significant inroads into the amateur community. Autoguiding, CCD imaging,
even
webcams which can be seen as fast data takes ala scintillation imagers (but
used
for different purposes and goals) were done via this route. Not everything
has
flowed in this direction of course, because the requirements of the
professionals don't necessarily overlap with those of the amateur, but some
have.

Right


Heather ... Assumpt: Why not search for a cure to prostate problems first,
and go back to
outfitting dog sleds with J12 Pratt & Whittney engines? !?

Belongs to alt.prostate.pratt&whittney

Mike Jones wrote in message:
Just a thought: for lower cost introductory AO systems and smaller
apertures (say to 40-60cm), might I suggest a simpler system for
correcting only X,Y tilt and focus? That would help to restore a large
portion of the image quality without the complexity and cost of a
multi-actuator DM, fast processor/FPGA, and laser beacon star.
Curvature is easily sensed using an HOE and narrow-band filter rather
than Shack-Hartmann and directly scaled to generate the drive signal for
a single MEMS actuator curvature-only membrane mirror. The system could
even be partially based on low noise analog circuits to reduce
processing latency and get closed-loop system speeds up to 100 Hz or
more. WFE sensor camera rates are a problem, as the WFE camera must
image at roughly 10X the closed loop rate, or at 1000Hz or more,
reducing integration time and sensitivity. X,Y tilt is a good start,
but adding focus compensation gives noticeable improvement.


I think it's a very good concept for a 25cm telescope, but suspect that more
channels will be needed for a larger instrument. In our systems, using
selectable set of SVD-optimized modes, we have the camera rate of about 4
times the correction rate, not 10 times.

So far the discussion is very useful. Thanks again.

On a sunny day (Tue, 22 Mar 2005 01:20:17 GMT) it happened "Marc Reinig"
wrote in :

Babcock's proposed (1953) Eidophor used a mirror covered with a film of oil
scanned by an electron beam to change the local slope.

It was never implemented, however.
We had an Eidophor TV projector in the TV studio, BIG machine, BW, sixties.

On Mon, 21 Mar 2005 15:25:37 -0700, Dan Mckenna
wrote:

Curvature mirrors are being made for optical communications and as the
volume goes up the prices go down. These are low voltage pizo bimorph
mirrors and could be made cheap if you knew how. (i don't)

Yes, the flexible actuated mirrors can be made inexpensively. I did a
2", 8-zone, comb driven design micro-machined from a single silicon
wafer. It could be probably be manufactured in moderate quantities for
well under $100. In fact, the MEMS processes used were compatible with
CMOS fabrication, so including circuitry directly on the mirror would
also be possible.

But as you note, that is hardly the only piece of the system required.

And I still see this as a light starved system in almost all cases, and
(in amateur usage) limited to planetary viewing/imaging. I just don't
know how large the market is for a $1000+ accessory that has no use
except sharpening the view of three objects.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

On Tue, 22 Mar 2005 13:11:10 GMT, Jan Panteltje
wrote:

We had an Eidophor TV projector in the TV studio, BIG machine, BW, sixties.

AFAIK, that was how all projection TVs were done until the late 70s or
so, when the first high-intensity phosphor CRT projectors were
developed. I remember seeing eidophor-based projectors being used in the
80's.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

Dear Gleb,

Is the possible to design and then build an AO system for
planetary(lunar) works like AO for solar works
you referred here http://www.noao.edu/noao/staff/keller/irao/ ?

With lesser correction elements (say, 12-16 for 12" aperture) and in a
serial manufacturing, I guess, the
cost of such a system will be significantly smaller, that that 25K.
May be custom DSP will help even
more.


VD

"Martin Brown" wrote in message
...
Marc Reinig wrote:

Babcock's proposed (1953) Eidophor used a mirror covered with a film of
oil scanned by an electron beam to change the local slope.

It was never implemented, however.

? Eidophors were invented by Fischer in 1939 and the first one
demonstrated in 1943. There was one still in use in a lecture theatre at
my university in the late 1970's.

There is a picture of the prototype online at:
http://www.cinephoto.co.uk/eidophor_1.htm

Regards,
Martin Brown

Yes, Eidophors existed prior to Babcocks proposal, as did lenses, mirrors,
knife edges and slits. But I believe Babcock was the first with a serious
proposal to use the Eidophor, et al to correct for atmospheric distortion
for observations.

Marc Reinig
UCO Lick Observatory
Laboratory for Adaptive Optics

For visible light, having subaperatures smaller than 12" makes little
difference. So for a 12" aperture, you would spend a lot for very very
little. However, a big improvement can be had by simply correcting for
tip/tilt at 50Hz.

Marc Reinig
UCO Lick Observatory
Laboratory for Adaptive Optics

"Vader" wrote in message
oups.com...

Is the possible to design and then build an AO system for
planetary(lunar) works like AO for solar works
you referred here http://www.noao.edu/noao/staff/keller/irao/ ?

With lesser correction elements (say, 12-16 for 12" aperture) and in a
serial manufacturing, I guess, the
cost of such a system will be significantly smaller, that that 25K.
May be custom DSP will help even
more.

Dear Marc

It depends on seeing, place and generally acceptance of what is "the right
thing". Professionals work at good sites and assume that seeing is similar
everywhere. This is one of the reasons why they are so sceptical about AO
for small telescopes. Another reason can be that they just know more ).
Yet another possibility can be that it's a kind of commonly accepted
mistake.

I never managed to find any mention of adaptive optics with a 10" telescope,
although everyone knows that it will be not useful, while assuming R_0=5cm
we get to a sensible AO system with 25 actuators! Something should be wrong
here.

To check for myself, I've done a couple experiments with a 25cm Newton in
Delft - which is probably the worst place you could find - a lot of light
pollution from greenhouses and very strong winds from the Norths sea - and
figured out that AO will help even with a very small telescope. Focusing
camera to the input pupil reveals strong turbulence with a scale much
smaller than the mirror size. Star images have clear boiling speckle
structure.

Of course the number of correection modes must be selectable so if tip-tilt
is the only significant term, the system will work on it, but if there is
more and the light is sufficient, then the system must be able to correct
more aberrations.

I assume that amateur astronomers are distributed uniformly over places with
reasonable to good living standards, and these places may have very bad
seeing conditions as these conditions are generally not taken into account
when people choose where to live.



"Marc Reinig" wrote in message
. ..
For visible light, having subaperatures smaller than 12" makes little
difference. So for a 12" aperture, you would spend a lot for very very
little. However, a big improvement can be had by simply correcting for
tip/tilt at 50Hz.

Marc Reinig
UCO Lick Observatory
Laboratory for Adaptive Optics

"Vader" wrote in message
oups.com...

Is the possible to design and then build an AO system for
planetary(lunar) works like AO for solar works
you referred here http://www.noao.edu/noao/staff/keller/irao/ ?

With lesser correction elements (say, 12-16 for 12" aperture) and in a
serial manufacturing, I guess, the
cost of such a system will be significantly smaller, that that 25K.
May be custom DSP will help even
more.

I agree. And funny enough, browsing the net I was amazed how many
amateurs are busy with making photos of exactly these three objects!


"Chris L Peterson" wrote in message
...
On Mon, 21 Mar 2005 15:25:37 -0700, Dan Mckenna
wrote:

And I still see this as a light starved system in almost all cases, and
(in amateur usage) limited to planetary viewing/imaging. I just don't
know how large the market is for a $1000+ accessory that has no use
except sharpening the view of three objects.

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com

On Wed, 23 Mar 2005 19:51:44 +0100, "Gleb" wrote:

I agree. And funny enough, browsing the net I was amazed how many
amateurs are busy with making photos of exactly these three objects!

g

But seriously- how do you get feedback for correcting a planetary image?
Analyzing a point source for distortion is one thing; analyzing the
wavefront of an extended object is something else altogether.

Also, you have to view your market in terms of imagers and viewers. The
former can get equivalent results with a webcam very inexpensively. So
the real question is how many visual observers will shell out some
serious money to sharpen up a few planets? I don't know... maybe a lot
will, just a question (the question, no doubt, that you are asking
yourself).

_________________________________________________

Chris L Peterson
Cloudbait Observatory
http://www.cloudbait.com