I have been trying to test whether ACP is properly configured to handle the mount flip. During the test I was using a simulator for the camera, but the mount and dome were live. The scope was being positioned by generated star fields. With pointing correction these could be corrected to being dead on target.

What I have found is that ACP correctly recognizes that a flip is required and issues a slew to perform the flip, but Gemini ignores the slew. I asked the author of the new Gemini ASCOM driver to look at this. His analysis is below. To summarize he believes that before the flip the scope is positioned so close to the target that Gemini ignores the slew command intended to cause the flip.

Using a live camera on the sky will introduce real errors, but it is conceivable using "Every Image Centered" that ACP could have positioned the scope to within 10 arcsec of its intended location.

Is there a specific ascom interface command for meridian flip? That would be safer.

Rob Hawley

From: paulkccd

Sent: Tuesday, June 14, 2011 4:38 AM
To: Gemini_ASCOM_Driver@yahoogroups.com
Subject: [Gemini_ASCOM_Driver] Re: ACP initiated mount flips

Hi Rob,

I've had a chance to look a bit closer at your log. Here's what I see might be the problem, and maybe other ACP users can chime in with some help:

Whenever ACP tries to flip the mount, it always executes a slew. But, the slew is always 1 arcsecond away from where the mount is currently pointing. Here's are a couple of examples from your log:

16:51:19.527 Serial received: TID:17 [0]: :GR, 08:35:34 <--- Current RA is 08:35:34

16:51:19.828 Serial Transmit TID:17 [0]: :Sr08:35:35#:ONPC Object#:Sd-00:02:30#:MS# <---- ACP tells Gemini to slew to 08:35:35, only 1 arsecond away!

Same here:
16:55:02.974 Serial received: TID:17 [0]: :GR, 08:35:34
16:55:04.633 Serial Transmit TID:17 [0]: :Sr08:35:35#:ONPC Object#:Sd-00:02:30#:MS#

I assume that Gemini simply doesn't want to slew the distance of one arcsecond, since RA doesn't change after this command is issued. The slew command is accepted, but perhaps it's too small for Gemini to bother, or maybe it's within the rounding error due to the precision being 0.5 arcseconds.

Because of this, your mount never flips, since Gemini always assumes that 1 arcsecond is close enough that a slew isn't necessary. Perhaps there are some settings in ACP to ensure a reasonable size margin for the slew needed to do a flip. I would make this margin at least 10 arcminutes, to be sure that the flip occurs.