CCD with a Dob?

Hi Roger,

Roger Hamlett wrote:
When the idea was mentioned, I knew I had seen something similar being
done. However there is a very real problem on the clocking, since the rate
will depend on the angle the scope is aimed above the celestial equator.
Though the sky rotates at a constant rate (hopefully!...), the movement
seen by a scope, is zero at the celestial pole, through to this rate at
the celestial equator, varying as the cos of the angle above the celestial
equator. So doing this right, would involve adding a knowing the latitude
of the observing site, and the the angle the scope if pointing, to
calculate the clock rate required. Also, while the paths approximate to a
straight line as you near the celestial equator, as you go north/south,
the curvature will be significant. Combine this with needing to align the
scope so that the axis of the CCD, is perpendicular to the polar axis, and
the whole thing seems to me, to be getting harder than just using a barn
door tracker...
I don't think this approach will every be 'simple'.

Ah, but that's what computers are for! And you need one for imaging
anyhow. These are exactly the types of problems that computers make
simple for us every day. It's just a matter of information (the scope
would need to have RA/Dec encoders) and coding. I can imagine an
algorithm to solve the problem of aligning the CCD automatically: after
all you are reading out continuously, which could be used in a feedback
loop as a star trails down the chip. It seems to me that the most
complicated/expensive part would be that the CCD housing would need to
rotate unless you restricted imaging to the meridian. But even the
latter could be a cheaper bare bones version of the same system.

The problem with tracking platforms is that while they work fairly well
for smaller Dobs, as you go to larger/heavier scopes they become
cumbersome and prohibitively expensive. At the same time, as you go to
larger aperture the drift scan idea becomes more practical.

What we need is for some ATM with electronics/CCD/software experience to
develop a home brew system to prove the concept.

Clear skies,
Greg

--
Greg Crinklaw
Astronomical Software Developer
Cloudcroft, New Mexico, USA (33N, 106W, 2700m)

SkyTools Software for the Observer:
http://www.skyhound.com/cs.html

Skyhound Observing Pages:
http://www.skyhound.com/sh/skyhound.html

To reply have a physician remove your spleen

"Greg Crinklaw" wrote in message
...
Roger Hamlett wrote:
Have a look in the SBIG application notes at:
http://www.sbig.com/sbwhtmls/app_notes.htm
Look at the one about 'using an ST6 for drift scan imaging'. With
this, a
'strip' of the sky is imaged, by clocking the CCD at the rate so the
movement matches the rate at which the pixels move across the image.
The
total exposure time, is the time taken for a star to cross the chip.
However it is not easy to do. You need to decide which part of the sky
you
are going to image, know the image scale of the CCD, and clock at
exactly
the right rate (remember that the physical pixel movement rate varies
according to what part of the sky you are on). You also need to have
the
camera very well aligned E-W, and accept that there will be a small
amount
of trailing away from the central 'line' across the image.

Thanks for the link Roger. It does prove the concept, at least to
point. Now imagine a CCD system designed from scratch with drift
scanning in mind... By eliminating the need to have a driven scope
imaging would suddenly become a simple reality for Dob owners, and I
suspect such a system would have the potential to sell very well.

Clear skies,
Greg
Possibly.
When the idea was mentioned, I knew I had seen something similar being
done. However there is a very real problem on the clocking, since the rate
will depend on the angle the scope is aimed above the celestial equator.
Though the sky rotates at a constant rate (hopefully!...), the movement
seen by a scope, is zero at the celestial pole, through to this rate at
the celestial equator, varying as the cos of the angle above the celestial
equator. So doing this right, would involve adding a knowing the latitude
of the observing site, and the the angle the scope if pointing, to
calculate the clock rate required. Also, while the paths approximate to a
straight line as you near the celestial equator, as you go north/south,
the curvature will be significant. Combine this with needing to align the
scope so that the axis of the CCD, is perpendicular to the polar axis, and
the whole thing seems to me, to be getting harder than just using a barn
door tracker...
I don't think this approach will every be 'simple'.

Best Wishes

Roger Hamlett wrote:
The problem is that it wouldn't be 'cheap'. An instrument rotator, that
doesn't result in the instrument tilting (or a field rotator system using
prisms), and is fairly accurate to move, is likely to run to at least
$300. Encoders for both axis, probably add up to another couple of hundred
dollars. With control hardware, and interface, you are talking something
like $700, on top of the price of the CCD camera. I don't feel the demand
will really be there, once the price is realised...

Well, I obviously disagree. I have looked into outfitting my Dob for
imaging using what is currently available and doing so is less practical
and *much* more expensive than what you outlined above. The drive
platforms have serious limitations and for my heavy Dob are
prohibitively expensive. I'm talking 20-inch Dobs here. But I already
sort of said all that. We'll have to agree to disagree.

Like I said, I'd buy one.

Clear skies,
Greg

--
Greg Crinklaw
Astronomical Software Developer
Cloudcroft, New Mexico, USA (33N, 106W, 2700m)

SkyTools Software for the Observer:
http://www.skyhound.com/cs.html

Skyhound Observing Pages:
http://www.skyhound.com/sh/skyhound.html

To reply have a physician remove your spleen

"Greg Crinklaw" wrote in message
...
Hi Roger,

Roger Hamlett wrote:
When the idea was mentioned, I knew I had seen something similar being
done. However there is a very real problem on the clocking, since the
rate
will depend on the angle the scope is aimed above the celestial
equator.
Though the sky rotates at a constant rate (hopefully!...), the
movement
seen by a scope, is zero at the celestial pole, through to this rate
at
the celestial equator, varying as the cos of the angle above the
celestial
equator. So doing this right, would involve adding a knowing the
latitude
of the observing site, and the the angle the scope if pointing, to
calculate the clock rate required. Also, while the paths approximate
to a
straight line as you near the celestial equator, as you go
north/south,
the curvature will be significant. Combine this with needing to align
the
scope so that the axis of the CCD, is perpendicular to the polar axis,
and
the whole thing seems to me, to be getting harder than just using a
barn
door tracker...
I don't think this approach will every be 'simple'.

Ah, but that's what computers are for! And you need one for imaging
anyhow. These are exactly the types of problems that computers make
simple for us every day. It's just a matter of information (the scope
would need to have RA/Dec encoders) and coding. I can imagine an
algorithm to solve the problem of aligning the CCD automatically: after
all you are reading out continuously, which could be used in a feedback
loop as a star trails down the chip. It seems to me that the most
complicated/expensive part would be that the CCD housing would need to
rotate unless you restricted imaging to the meridian. But even the
latter could be a cheaper bare bones version of the same system.

The problem with tracking platforms is that while they work fairly well
for smaller Dobs, as you go to larger/heavier scopes they become
cumbersome and prohibitively expensive. At the same time, as you go to
larger aperture the drift scan idea becomes more practical.

What we need is for some ATM with electronics/CCD/software experience to
develop a home brew system to prove the concept.

Clear skies,
Greg
The problem is that it wouldn't be 'cheap'. An instrument rotator, that
doesn't result in the instrument tilting (or a field rotator system using
prisms), and is fairly accurate to move, is likely to run to at least
$300. Encoders for both axis, probably add up to another couple of hundred
dollars. With control hardware, and interface, you are talking something
like $700, on top of the price of the CCD camera. I don't feel the demand
will really be there, once the price is realised...

Best Wishes

Michael Kreuzer wrote:
Agreed, but would add the sun (with a filter) to the very short list of
possibilities.

Regards, Michael

Polaris can be added as well for thos in the northern hemisphere ...
and will allow a longer exposure.

--
John Oliver

John Oliver wrote in news:zPsWc.14315$nk.2491
@okepread05:

Michael Kreuzer wrote:
Agreed, but would add the sun (with a filter) to the very short list of
possibilities.

Regards, Michael

Polaris can be added as well for thos in the northern hemisphere ...
and will allow a longer exposure.

Except that polaris moves, just like any other star. It just moves in VERY
SMALL circle.

"Greg Crinklaw" wrote in message
...
Roger Hamlett wrote:
The problem is that it wouldn't be 'cheap'. An instrument rotator,
that
doesn't result in the instrument tilting (or a field rotator system
using
prisms), and is fairly accurate to move, is likely to run to at least
$300. Encoders for both axis, probably add up to another couple of
hundred
dollars. With control hardware, and interface, you are talking
something
like $700, on top of the price of the CCD camera. I don't feel the
demand
will really be there, once the price is realised...

Well, I obviously disagree. I have looked into outfitting my Dob for
imaging using what is currently available and doing so is less practical
and *much* more expensive than what you outlined above. The drive
platforms have serious limitations and for my heavy Dob are
prohibitively expensive. I'm talking 20-inch Dobs here. But I already
sort of said all that. We'll have to agree to disagree.

Like I said, I'd buy one.
OK. The sheer size of your Dob, obviously makes it a candidate for such a
system, but I still think the demand will just be too small to cover
developing the system. I have stepper drives here, used to move my milling
machine, which give step sizes under 1/10th thou, and can move several
hundred pounds of weight. Adapting one to move a Dob, I'd consider a much
smaller job than making the CCD system. The small number of large Dobs,
and even smaller number of people who want to image with one, just is not
likely to cover the development costs (remember that on top of the
rotator, and encoders, a custom CCD camera is going to be needed). I'd
estimate a competent engineer, ought to be able to motorise your Dob, with
more or less 'off the shelf' parts in less than a couple of weeks, while
I'd put the development of the rotator, and camera system, as perhaps 18
months work (including writing camera drivers)...
The problem is the sheer scale of development involved, takes it into
something that a person doing it, has to live while developing, making it
too larger a project for 99% of hobbyists. Conversely, the motorised
platform is a relatively simple solution (remember also that stepper
motors are commonly available from scrap equipment).
Look at:
http://www.piclist.com/techref/io/st...step/index.htm
Add a Type 23 stepper motor, and you are talking a drive that at the sort
of reduction ratio used on a platform, should provide more than enough
control and power.

Best Wishes

Whatever happened to the CWIP system that was demonstrated at
Astrofest five or six years ago? IIRC, it was a camera that had
automatic stacking of frames, done inside the camera. I think it was
for alt-az mounts without tracking. The camera output was a standard
video signal to a television. Does anyone remember this?

Francis



"Roger Hamlett" wrote in message ...
"Greg Crinklaw" wrote in message
...
Hi Roger,

Roger Hamlett wrote:
When the idea was mentioned, I knew I had seen something similar being
done. However there is a very real problem on the clocking, since the
rate
will depend on the angle the scope is aimed above the celestial
equator.
Though the sky rotates at a constant rate (hopefully!...), the
movement
seen by a scope, is zero at the celestial pole, through to this rate
at
the celestial equator, varying as the cos of the angle above the
celestial
equator. So doing this right, would involve adding a knowing the
latitude
of the observing site, and the the angle the scope if pointing, to
calculate the clock rate required. Also, while the paths approximate
to a
straight line as you near the celestial equator, as you go
north/south,
the curvature will be significant. Combine this with needing to align
the
scope so that the axis of the CCD, is perpendicular to the polar axis,
and
the whole thing seems to me, to be getting harder than just using a
barn
door tracker...
I don't think this approach will every be 'simple'.

Ah, but that's what computers are for! And you need one for imaging
anyhow. These are exactly the types of problems that computers make
simple for us every day. It's just a matter of information (the scope
would need to have RA/Dec encoders) and coding. I can imagine an
algorithm to solve the problem of aligning the CCD automatically: after
all you are reading out continuously, which could be used in a feedback
loop as a star trails down the chip. It seems to me that the most
complicated/expensive part would be that the CCD housing would need to
rotate unless you restricted imaging to the meridian. But even the
latter could be a cheaper bare bones version of the same system.

The problem with tracking platforms is that while they work fairly well
for smaller Dobs, as you go to larger/heavier scopes they become
cumbersome and prohibitively expensive. At the same time, as you go to
larger aperture the drift scan idea becomes more practical.

What we need is for some ATM with electronics/CCD/software experience to
develop a home brew system to prove the concept.

Clear skies,
Greg
The problem is that it wouldn't be 'cheap'. An instrument rotator, that
doesn't result in the instrument tilting (or a field rotator system using
prisms), and is fairly accurate to move, is likely to run to at least
$300. Encoders for both axis, probably add up to another couple of hundred
dollars. With control hardware, and interface, you are talking something
like $700, on top of the price of the CCD camera. I don't feel the demand
will really be there, once the price is realised...

Best Wishes

Subject: CCD with a Dob?
From: Greg Crinklaw
Date: 8/21/04 8:13 AM Pacific Daylight Time
Message-id:


What I'm wondering is why nobody has come up with a commercial CCD
system that works as a drift scan? It seems to me this is just a matter
of some electronic wizardry when reading the chip, and with all those
undriven large Dobs out there could be very lucrative...


In addition to the App Note pointed out earlier - also check out the image on
this page:

http://www.sbig.com/402/ST-402.htm

taken with an unguided scope using drift scan method. Also see:

http://www.sbig.com/pdffiles/ST7-9ICameraApps.pdf

The software and hardware is already available to do this.

Regards,
Michael Barber
SBIG