Hi Rider,
Thanks for the follow-up which is appreciated.
Quote:
Originally Posted by rider
Thanks Gary,
I'll try your method to see if I like it.
In most sessions, I tend to concentrate on a particular part of the sky and if I don't stray too far from the selected (1 star) alignment star, I get good accuracy. if I do switch to a new part of the sky, it only a matter of selecting a new alignment star. -very quick.
|
This is also a good strategy and one we also recommend.
What is happening here is that in the face of polar misalignment and
mount errors, the pointing residual will be smallest in the neighborhood
of the alignment star. So re-aligning on some star when you move to a new
part of the sky exploits this fact.
Once you get more familiar with TPAS, you can use it in a similar quick way
but with the added benefit that the system can help refine the pointing model
so that the pointing improves across the whole sky. Rather than align on
a new star, SAMPLE the position of the star, DIAL up SETUP MNT ERRORS,
COMPUTE ERRORS and accept them all in the model. This is all easy in practice.
You simply find yourself hitting the ENTER button a few times in rapid succession.
What is neat about this approach is that after you SAMPLE and COMPUTE
on each new star, after a while you will often find that when you move to some
other part of the sky, you don't need to do it again for the rest of the night as
the pointing improves across the whole sky.
Quote:
|
Brisbane's sky has not been cooperating this year, and if I do get a good night, I don't want to waste any of that short window between dark and dew adjusting equipment.
|
You have got your priorities right. Never let aligning the system get in the way
of observing.
Quote:
|
I'll take on board your comments about the negligible nature of sender drag, however I did some experimenting last night, and the result was that if I un-balance my scope without the Dec sender belt on, it will pivot on the axis under its own weight, but with the belt attached it stays (firmly) where it is. The same happens with the RA axis cog disengaged, so there must be some drag. Static load only perhaps?
|
OK. However, it is not really a good experiment when it comes to demonstrating
the load a motor 'sees' on a telescope when it comes to determining the motor's
torque requirements.
You might be familiar with or heard of a concept known as "moment of inertia".
It is a measurement of how hard it is to change the rotational motion of a body around
a given axis. On a scope, the further out the OTA and counterweight from the axis
of rotation, the larger is the moment of inertia and hence the larger the effort
required for the motor to initially move it.
Rather than think in terms of pushing on the OTA or counterweight, which is what we
tend to do as humans so we have some mechanical advantage, think in terms
of the motor attempting to rotate these two masses at the very axis itself.
It is like when one pushes open a door. Normally we push the door near the handle
and it opens easily. However, if one pushes it close to the hinges, it can take
considerable effort to push it open. Much of the mass of the door is some distance
away from the axis of rotation (which runs up through the hinges). This mass
distributed some distance away from the axis increases the doors moment of
inertia.
The moment of inertia the motors "see" when they rotate the OTA and counterweight
will be orders of magnitude larger than the moment of inertia that would be required
to rotate the encoder shaft. Indeed, once in motion, the encoder also provides a
constant light static drag (torque) as do the bearings in the mount, but compared to the
initial effort required for the motor to spin the scope from a point close to the axis, it
is an absolutely tiny amount.
Quote:
|
Regarding accuracy of any type of goto or push to device, my opinion is that if the object is in the ep field then the unit has acheived its function.
|
Absolutely and in the Manual and whenever I give talks on pointing
analysis, we talk in terms of an individuals "pointing goal".
For example, landing the object in the FOV of an eyepiece of choice is
a commonly articulated pointing goal. Others have more demanding goals,
such as blind pointing to land an object onto a small CCD array.
Quote:
|
Realistically its a bit too much to expect ultimate levels of accuracy out of portable mass-produced amateur equipment (such as Eq mounts). With an accurate polar alignment prior to starting, both Argo and Synscan acheive this. I recon they both do a good job.
|
The good news is that often the primary source of pointing errors in some
GEM's is not entirely the fault of the manufacturer at all but Dec to optical
axis non-perpendicularity that was in play by the user. Sometimes something
as simple as not having collimated the scope or not mounting a dovetail squarely.
What is surprising is how a handful of terms in a pointing model can
so often dramatically improve, in a very real quantitative sense, the pointing
performance of scopes such as these. As long as the source of the errors
is systematic rather than random (no system can compensate for random errors)
a good pointing analysis system can transform commodity mounts into even
greater value performers.
Thanks again for the post and I hope your weather in Brisbane is better than
the weather here in Sydney tonight.
Best Regards
Gary Kopff
Wildcard Innovations Pty Ltd.
