AO mount bump seems to wander

[GoogleCodeExporter]

I have been imaging for the last couple nights, and have noticed that when 
mount bumps occur, the AO tends to wander (as viewed on the AO graph).

The current code picks a direction and steps towards it, and it seems like that 
might not be the right algorithm.  It might be better to have a high water 
mark/low water mark system, where every step it moves towards the center (not 
to where the center was when it was decided to bump), and to keep bumping until 
it gets "close enough".

I don't think this is a "show stopper" for first release, but I wanted to log 
it as a reminder for the second release.

Original issue reported on code.google.com by br...@boneheads.us on 1 Dec 2013 at 6:47

[GoogleCodeExporter]

I agree, bumping towards the center until "close enough" would be an 
improvement. We could use the averaged position vector (not the actual position 
vector) as the basis for corrections to avoid sending oscillating corrections 
to the mount.

Original comment by andy.gal...@gmail.com on 1 Dec 2013 at 6:17

• Changed state: Accepted
• Added labels: Type-Enhancement, OpSys-All
• Removed labels: Type-Defect, OpSys-

[GoogleCodeExporter]

Original comment by andy.gal...@gmail.com on 10 Jan 2014 at 11:00

• Changed state: Started

[GoogleCodeExporter]

Here is a patch that implements the always-bump-toward-center behavior. It was 
very straightforward, and I have tested it in the simulator and it seems to be 
an improvement.

I'm reluctant to commit the change without trying it in the field, but it is 
cloudy here for the next few days.

In the meantime, if anyone with an AO wants to give it a try, please do!

Original comment by andy.gal...@gmail.com on 11 Jan 2014 at 8:00

Attachments:

• bump_towards_center.patch

[GoogleCodeExporter]

This issue was closed by revision r684.

Original comment by andy.gal...@gmail.com on 15 Jan 2014 at 6:09

• Changed state: Fixed