[Eden (Brett)]

Hi Greg,

Thanks for sharing. You've mentioned some really good points here, on a subject which always deserves discourse.

Originally Posted by gregbradley

One question I have is time delay. I noticed Martin Pugh mentioned once he uses a 1 second delay after corrections in his autoguiding. I don't know if he does that all the time or just some of the time.

Anyone noticed any improvement in guiding from doing that? I have done that when using shorter guide exposures and if its windy or bad seeing.

I guess it would help more on a mount that had a bit of backlash or was struggling a tad as it would give the mount a chance to settle.

I've never used this feature, but it makes sense that it would help in situations where the mount has significant backlash or is under heavy load.

The first thing I learnt about autoguiding is its not a solution to bad polar alignment. It tends to be as good as your polar alignment and balance.
So I spend a fair bit of time on getting the polar alignment as good as possible.

This is something that can't be stressed enough and these are probably the two most critical factors in successful auto-guiding. Since I don't yet have an Obs, I usually spend one night setting up, refining polar alignment and auto-guiding, and then spend the following night imaging. It always pays to go that extra mile with the polar alignment.

I thought of a potential good accessory someone could make. That would be a polar alignment scope with a CCD camera in it so you could plate solve the image when its pointing at the SCTParea and adjust until the reticle matches exactly. Just like you do manually but without the crouching down and the dim SCP Octans stars.

This is a really good idea and could be easily implemented. Dion over at Astronomy Shed demonstrated a basic CCD (I think it was a modified Phillips TouCam) connected to an EQ6 polar scope. Assuming the required source images are available at that focal length, incorporating plate solving would be simple enough. The real killer app would be to have motorized alt/az adjustments which take directions from the plate solving output!

I'd like to hear some opinions about Min/Max move and what aggressiveness you use under what conditions.

A few points I use are if the seeing is not so good it makes sense to use longer exposure times for less often corrections. It would also make sense to set the max move lower so you are not trying to correct the seeing variability.

If the minimum move were too low then the mount would be correcting mostly for seeing and the correction would be meaningless as we only want to correct for drift and periodic error of the mount.

Some good advice right there. Extending your guide camera exposure time can help "average out" the effects of seeing on the guide star.

I never really saw much gain from adjusting the min/max values, until I started using AO. Because AO usually relies on fast guide camera exposures, the star will obviously bounce around a lot more. Increasing the minimum movement threshold instructs the AO to ignore the high order atmospheric scintillations, since we can't correct for those, even at 10Hz. Neglecting to do this when the AO is running at something like 10Hz results in a constant clatter of the actuators, as the unit desperately tries to correct for every minor change in the guide stars appearance.

To see how the min/max values would influence the operation of the SXV-AO, I set the guide camera exposure time to 0.01s and took a single 30 minute exposure of Antares, whilst using it as the guide star via the ONAG. In preparation, I increased the min movement variable until the AO actuators stopped going berserk and then tweaked it so that significant changes in the guide star, such as the effects of low-order seeing, would be corrected. It addressed the issue of "chasing the seeing"; Antares itself and surrounds were very well defined, but this diminished in an almost radial fashion towards the edge of the field (1.162 arcsec/pixel, 0.9 degrees field radius).

Brett

Edit: Apologies, Greg. Didn't mean to jump in before you'd finished editing!