RE: Shutter timing

[Tom Droege <droege@wwa.com>]

Well, I tried and did not think of anything immediately.   I am more trying 
to keep the options open.  The problem needs to be in the hands of someone 
that can solve the software problems.  This is not me.

The stamp should be able to time exposures to a millisecond if that is all 
it does.  It could also be programmed to open the two shutters for 
different lengths of time so that similar stars saturate at the same 
magnitude in the two filters.  I am not sure this is needed.

The shutters take about 100 milliseconds to open.  My guess is that this 
will be uniform to about 1% from the motor drive.  This means the center 
will always be exposed 200 milliseconds more than the edges.  The shutter 
opens like a parlor door so you can imagine how this affects the image.  I 
think everything will be highly repeatable.

Remember, you can not get at the interrupt structure from QBasic.  Not 
quite true, but by the time you go through Windows, it takes a real 
expert.  With interrups and someone that can write code for the serial 
port, there is a lot that can be done.

Tom Droege



At 08:08 AM 6/23/00 -0700, you wrote:
>Argh.  You can't do all-sky photometry if you don't
>have an accurate shutter.  This is something you
>can control; why be happy with one percent?  You
>should be able to time the signal to open and/or
>close the shutter to higher accuracy than 1/10 second
>with a PC.  You should be able to test the time
>necessary to actually finish the motion to high
>precision, and see whether it is repeatable.
>   I agree that knowing the precise midpoint of an
>exposure to the nearest millisecond is overkill
>for most applications, but knowing the exposure
>duration is important.  Exposure times will change
>depending on the application (such as looking at
>Landolt standards or bright variables), so don't
>be tied to the 100-second regime when thinking
>about exposure lengths.
>   Can you think of a reason why you would ask the
>Stamp to do something while an exposure is taking
>place?  I can't.  Why not let it time the exposure
>in the most convenient way?
>Arne