ASTRO: NGC 1169

"Rick Johnson" wrote in message
...


Richard Crisp wrote:

I spoke to a friend that works at a company that competes with SBIG about
your weird triangular shaped star issue. He wonders if your window to the
camera is being distorted by the shrinking aluminum in the extreme cold?

A similar potential issue could be the coverslip over the sensor: the
ceramic package shrinks when it gets cold and the TCE of the coverslip
may
be different than the ceramic so it could potentially distort the
coverslip
and introduce astigmatism

lastly the way the sensor is attached to the cold finger may be
mechanically
binding and as it gets even colder it could mechanically distort the
ceramic
package and its coverslip

the distortion would normally cause astigmatism and if it is a
rectangular
arrangement as the coverslip shape is, then you could see some odd angles
of
the astigmatism

I'd expect it on two axes for example.

I doubt it is a mechanical problem. Only happens with saturated stars.
It gets less at 2x2 and nearly vanishes at 1x1 binning. It would
appear worse at larger image scales if it were a mechanical problem. It
has to be due to saturation but that normally causes a downward spike
not one to the right. I should call it a snow cone (on its side) image
rather than triangular.



Ok that is useful information

you most likely have a situation of the horizontal shift register
overflowing...

the behavior from the binning and the snowcone tail is whatr clinches it for
me

when you bin 3x3 you are dumping the charge of three pixels in to a single
horizontal shifr register pixel and then you are dumping those three
horizontal shift registers pixel contents onto the floating diffusion
connected to the output source follower. binning 2x2 puts less charge in the
horizontal shift register pixel. Binning 1x1 puts even less.

Can you bin 3x1? If so that would separate the issue being a floating
diffusion issue versus horizontal shift register. If it looks the same
binned 3x1 as 3x3 then it is the horizontal shift register rather than some
sneak path back from the floating node to the last cells in the horizontal
shift register.


The way Kodak designs these sensors the charge capacity in the horizontal
shift register cells is not 5x the capacity of a pixel's well. In fact it is
139K e- versus 91K e- capacity for the vertical register

so if your pixel is saturated, when you bin it is going to overflow the
horizontal pixel.

that would cause the right-facing tail if the shift register shifts to the
left. It is trailing charge.

SBIG can change the clock levels a bit and affect the trailed charge
quantity, that's a design parameter under their control. But if the vertical
register is filled, then it will overflow the horizontal one if you bin

That doesn't explain but one point of the triangle. A mechanical distortion
of the window, coverslip or sensor package causing astigmatism, can explain
the other two points of the triangles you observe

Richard Crisp wrote:

"Rick Johnson" wrote in message
...


Richard Crisp wrote:


I spoke to a friend that works at a company that competes with SBIG about
your weird triangular shaped star issue. He wonders if your window to the
camera is being distorted by the shrinking aluminum in the extreme cold?

A similar potential issue could be the coverslip over the sensor: the
ceramic package shrinks when it gets cold and the TCE of the coverslip
may
be different than the ceramic so it could potentially distort the
coverslip
and introduce astigmatism

lastly the way the sensor is attached to the cold finger may be
mechanically
binding and as it gets even colder it could mechanically distort the
ceramic
package and its coverslip

the distortion would normally cause astigmatism and if it is a
rectangular
arrangement as the coverslip shape is, then you could see some odd angles
of
the astigmatism

I'd expect it on two axes for example.

I doubt it is a mechanical problem. Only happens with saturated stars.
It gets less at 2x2 and nearly vanishes at 1x1 binning. It would
appear worse at larger image scales if it were a mechanical problem. It
has to be due to saturation but that normally causes a downward spike
not one to the right. I should call it a snow cone (on its side) image
rather than triangular.




Ok that is useful information

you most likely have a situation of the horizontal shift register
overflowing...

the behavior from the binning and the snowcone tail is whatr clinches it for
me

when you bin 3x3 you are dumping the charge of three pixels in to a single
horizontal shifr register pixel and then you are dumping those three
horizontal shift registers pixel contents onto the floating diffusion
connected to the output source follower. binning 2x2 puts less charge in the
horizontal shift register pixel. Binning 1x1 puts even less.

Can you bin 3x1? If so that would separate the issue being a floating
diffusion issue versus horizontal shift register. If it looks the same
binned 3x1 as 3x3 then it is the horizontal shift register rather than some
sneak path back from the floating node to the last cells in the horizontal
shift register.


The way Kodak designs these sensors the charge capacity in the horizontal
shift register cells is not 5x the capacity of a pixel's well. In fact it is
139K e- versus 91K e- capacity for the vertical register

so if your pixel is saturated, when you bin it is going to overflow the
horizontal pixel.

that would cause the right-facing tail if the shift register shifts to the
left. It is trailing charge.

SBIG can change the clock levels a bit and affect the trailed charge
quantity, that's a design parameter under their control. But if the vertical
register is filled, then it will overflow the horizontal one if you bin

That doesn't explain but one point of the triangle. A mechanical distortion
of the window, coverslip or sensor package causing astigmatism, can explain
the other two points of the triangles you observe

Yes I can bin 3x1 or is it 1x3? I'm at the wrong computer, the imaging
one isn't on line. If clear tonight I'll give it a try. Supposed to
hit -35C tonight so won't have a problem hitting -50C! Could try -80C
for that matter but that would likely frost over the outside of the CCD
chamber. The problem is minimal at -35C where I normally image except
in summer where I drop to -25C. The cooler in my camera is usually
good. I can easily drop 45C with air cooling and still have only 80%
current draw. Since the nights only get cooler and warming only means
clouds rolled in, I usually set 80% as my limit. I only gain a couple
degrees in that top 20% anyway.

You get the FIT I posted?

Rick
--
Correct domain name is arvig and it is net not com. Prefix is correct.
Third character is a zero rather than a capital "Oh".

I should add that there is a pot I can adjust to increase the ABG
activity reducing the saturation point but it also reduces over all
quantum efficiency. SBIG suggest it be used only for downward blooms.
They didn't know if it would work for this problem though they haven't
gotten back to me since my last email late Friday. I'm sure that hit
after closing for the weekend.

The bloom appearance is similar to their photo but to the right rather
than down.

http://www.sbig.com/sbwhtmls/abg.htm

Rick

--
Correct domain name is arvig and it is net not com. Prefix is correct.
Third character is a zero rather than a capital "Oh".

"Rick Johnson" wrote in message
...


Richard Crisp wrote:

"Rick Johnson" wrote in message
...


Richard Crisp wrote:


I spoke to a friend that works at a company that competes with SBIG
about
your weird triangular shaped star issue. He wonders if your window to
the
camera is being distorted by the shrinking aluminum in the extreme cold?

A similar potential issue could be the coverslip over the sensor: the
ceramic package shrinks when it gets cold and the TCE of the coverslip
may
be different than the ceramic so it could potentially distort the
coverslip
and introduce astigmatism

lastly the way the sensor is attached to the cold finger may be
mechanically
binding and as it gets even colder it could mechanically distort the
ceramic
package and its coverslip

the distortion would normally cause astigmatism and if it is a
rectangular
arrangement as the coverslip shape is, then you could see some odd
angles
of
the astigmatism

I'd expect it on two axes for example.

I doubt it is a mechanical problem. Only happens with saturated stars.
It gets less at 2x2 and nearly vanishes at 1x1 binning. It would
appear worse at larger image scales if it were a mechanical problem. It
has to be due to saturation but that normally causes a downward spike
not one to the right. I should call it a snow cone (on its side) image
rather than triangular.




Ok that is useful information

you most likely have a situation of the horizontal shift register
overflowing...

the behavior from the binning and the snowcone tail is whatr clinches it
for
me

when you bin 3x3 you are dumping the charge of three pixels in to a
single
horizontal shifr register pixel and then you are dumping those three
horizontal shift registers pixel contents onto the floating diffusion
connected to the output source follower. binning 2x2 puts less charge in
the
horizontal shift register pixel. Binning 1x1 puts even less.

Can you bin 3x1? If so that would separate the issue being a floating
diffusion issue versus horizontal shift register. If it looks the same
binned 3x1 as 3x3 then it is the horizontal shift register rather than
some
sneak path back from the floating node to the last cells in the
horizontal
shift register.


The way Kodak designs these sensors the charge capacity in the horizontal
shift register cells is not 5x the capacity of a pixel's well. In fact it
is
139K e- versus 91K e- capacity for the vertical register

so if your pixel is saturated, when you bin it is going to overflow the
horizontal pixel.

that would cause the right-facing tail if the shift register shifts to
the
left. It is trailing charge.

SBIG can change the clock levels a bit and affect the trailed charge
quantity, that's a design parameter under their control. But if the
vertical
register is filled, then it will overflow the horizontal one if you bin

That doesn't explain but one point of the triangle. A mechanical
distortion
of the window, coverslip or sensor package causing astigmatism, can
explain
the other two points of the triangles you observe

Yes I can bin 3x1 or is it 1x3? I'm at the wrong computer, the imaging
one isn't on line. If clear tonight I'll give it a try. Supposed to
hit -35C tonight so won't have a problem hitting -50C! Could try -80C
for that matter but that would likely frost over the outside of the CCD
chamber. The problem is minimal at -35C where I normally image except
in summer where I drop to -25C. The cooler in my camera is usually
good. I can easily drop 45C with air cooling and still have only 80%
current draw. Since the nights only get cooler and warming only means
clouds rolled in, I usually set 80% as my limit. I only gain a couple
degrees in that top 20% anyway.

You get the FIT I posted?



No i didn't get the FIT

you can send it to my FTP site:

I am not sure if it is 3x1 or 1x3 binning. I thought the first number was
the vertical direction and the second was the horizontal

here's how to access my inbound ftp site

and here's a free FTP client that has been verified to work

http://www.coreftp.com/download.html

use the COREFTP LE

go ahead and FTP the FIT data to my site and then send me an email telling
me that it is complete. I am really curious about what this looks like.

If you can zip up files that show it and others that dont as a function of
binnign that would be great

"Rick Johnson" wrote

Rick,

Are you sure it's not an issue with the telescope or the attachment of
the camera to the scope? At -50C the mechanical components holding the
optics must be shrinking. Whenever I here "triangular stars" I think
"pinched optics". Perhaps there is some reflection off the filters?

As for star saturation: I've done some experimenting with factoring down
data before applying de-convolution (which I understand that you don't use).
I need to play around more with the gradient reduction routines in IRIS and
work on getting full 16-bit data into the program. With your dark sky I'm
surprised that you have any gradients at all except under moonlight.

AIP4WIN 2.0 will add sub-exposures forever. There is no upper limit on
the values in uses. Unfortunately, I've not re-loaded that program since the
last time my laptop crashed.

George N

George Normandin wrote:

"Rick Johnson" wrote

Rick,

Are you sure it's not an issue with the telescope or the attachment of
the camera to the scope? At -50C the mechanical components holding the
optics must be shrinking. Whenever I here "triangular stars" I think
"pinched optics". Perhaps there is some reflection off the filters?

As for star saturation: I've done some experimenting with factoring down
data before applying de-convolution (which I understand that you don't use).
I need to play around more with the gradient reduction routines in IRIS and
work on getting full 16-bit data into the program. With your dark sky I'm
surprised that you have any gradients at all except under moonlight.

AIP4WIN 2.0 will add sub-exposures forever. There is no upper limit on
the values in uses. Unfortunately, I've not re-loaded that program since the
last time my laptop crashed.

George N

It's not mechanical, I ruled that out long ago. SBIG agrees but has
given me no help so far other than keep the camera warmer.

No bloom at 1x1 binning, some bloom at 2x2 and severe at 3x3. It's got
to be something in the camera that can't take the cold electrically.

Rick
--
Correct domain name is arvig and it is net not com. Prefix is correct.
Third character is a zero rather than a capital "Oh".

"Rick Johnson" wrote
......
It's not mechanical, I ruled that out long ago. SBIG agrees but has given
me no help so far other than keep the camera warmer.

Image "warmer"? LOL.......

I guess those California guys don't understand winter in the north.

George N