MaxIm not reporting correct value for MultiStarGuiding [workaround & bug filed]

[Samuel Lising]

Hi.

I've just begun setting up ACP and scheduler at my remote setup at Sierra Remote Observatories and have run into some issues with ACP and guiding.

My hardware setup:
AP 1600
AP 130 GT
SBIG STT-8300 with FW8G-STT

Software:
APC 7.2a
Scheduler 3.7a
APCC
M1 Weather Watcher
Maxim DL 6.07
Focus Max 4

The issue I have is that guiding seems to only work intermittently. With the targets that I have entered into Scheduler, early in the evenings ACP seems to be able to guide properly, both before and after a meridian flip. However, for some reason my target later in the evening don't seem to be able to acquire guide stars properly and the plan fail. It's been rather bad weather over the last few weeks and I have only had a few nights to test but I believe what I am noticing in Maxim's guide log is that the guide stars are drifting for the targets that fail. It's strange that guiding would work on both sides of the meridian for some targets and then drift in the wrong direction later on in the evening for other targets.

this is the log message:
09:52:51 Updating pointing...
09:52:51 (taking 5 sec. exposure, Clear filter, binning = 3)
09:52:51 (using Raw readout mode)
09:52:51 (starting exposure)
09:52:59 (exposure complete and image downloaded)
09:53:00 Image finished
09:53:00 Plate-solve pointing image.
09:53:00 722 image stars found
09:53:00 886 catalog stars found
09:53:01 Solved! 229 stars matched.
09:53:01 Average residual is 0.4 arcsec.
09:53:01 Pointing error is 0.266 arcmin @ angle 275.55
09:53:01 True focal length is 862.5 mm.
09:53:01 True binned plate scales (arcsec/pix): H = 3.87 V = 3.87
09:53:01 True image center (J2000): 07h 04m 13.9s -11� 18' 21.46"
09:53:01 Imager sky position angle is 31.6 deg.
09:53:01 [sync] telescope synced
09:53:01 Within max error tolerance, no re-slew needed.
09:53:01 Target is now centered.
09:53:01 (long exp, no orbital tracking, trying to autoguide)
09:53:08 (guide star SNR=97.8 X=165 Y=126)
09:53:08 (initial exposure interval 2)
09:53:13 (guide star SNR=49.4 X=165 Y=126)
09:53:13 (final exposure interval 2)
09:53:17 (guide star SNR=44.3 X=165 Y=126)
09:54:32 (guide star SNR=79.0 X=297 Y=28)
09:54:32 (initial exposure interval 2)
09:54:37 (guide star SNR=70.7 X=297 Y=28)
09:54:37 (final exposure interval 2)
09:54:41 (guide star SNR=59.3 X=297 Y=28)
09:55:57 (guide star SNR=46.8 X=171 Y=102)
09:55:57 (initial exposure interval 2)
09:56:02 (guide star SNR=38.4 X=171 Y=102)
09:56:02 (final exposure interval 2)
09:56:06 (guide star SNR=43.0 X=171 Y=101)
09:57:21 (guide star SNR=42.3 X=172 Y=99)
09:57:21 (initial exposure interval 2)
09:57:26 (guide star SNR=40.1 X=172 Y=99)
09:57:26 (final exposure interval 2)
09:57:30 (guide star SNR=45.1 X=172 Y=99)
09:58:40 **Autoguiding failed.
09:58:40 (**AG failed for aggregate exposures and external guidescope)
ACP console log closed 24-Dec-2014 09:58:40 UTC

I've tried 2 methods for calibration - first with ACP set to "rotated off axis pickoff" with the guide sensor rotation set at 0 degrees and the calibration executed through ACP. The second method I just calibrated Maxim and had ACP set to "no rotator (all types)". With both methods I calibrate pointing to the east. I have found that with the second method I can't calibrate from ACP because the ACP calibration process slews the telescope to find a guide star, but it is based on the field of view of the main imaging ccd and not the guiding field of view. But as I understand it - as long as maxim is calibrated properly, it shouldn't make a difference in ACP. In both cases I have had the same results, being able to guide on my targets early in the evening and then failing with targets later in the evening.

I have maxim set with the "pier flip" setting off and "on pier flip" set to "do not change". In all the scheduler plans I have dithering set to -1 (auto). I am wondering if one of these settings might be the problem. In my past software set up I have found that I needed to reverse x on pier flip, but as I understand it I do not need to set this when running ACP. Is this correct? Maxim is also set to use multiple guide stars. I believe I had tried in the past to just use a single guide star with the same result (I can't be certain that I tried this. The days of weather delays extends beyond my usable memory)

I also have APCC with "enable pointing correction" on and "enable tracking correction" on. I've been able to guide in maxim with these settings so I don't think they are the culprit.

I'm including attachments of the guider and imaging session log files. You will notice that the first 3 targets were acquired successfully (M45, IC434 and NGC2264) after that all the remaining logs show the same guiding error. If you look at the maxim guide log you will see that the failing targets have x and y errors that seem to be drifting in the wrong direction.

I saw an earlier thread a similar issue with intermittent guiding results and replacing the SBIG plugin in Maxim seemed to solve the issue. I've checked my plugin and it seems to be a more recent release that the one that was posted.

I'll probably try to play around with the pier flip and dithering settings when I get some good weather again, but I'd like to know if anyone with a similar set up has had a issues with guiding such as this.

Thanks.
Sam

[Bob Denny]

Sam --

The less you play with pier flip the happier you'll be. You will drive yourself crazy trying to second guess ACP. If you can't get guiding going on the east you won't get it going at all. It's unlikely bu looking at your logs that fipping polarity changes are the problem. It just won't even start guiding.

My hardware setup:
AP 1600
AP 130 GT
SBIG STT-8300 with FW8G-STT

I cannot tell what your guiding setup is from this. I can tell what you have set in ACP. Do you have a rotator or not? If not, do not attempt to use a rotated guiding mode, and the guide sensor angle is irrelevant. If you do have a rotator you must use the appropriate rotated mode based on whether you are using the 8300's internal camera or an off-axis pickoff like a MOAG or whatever. Cal in the east. Read the angle in MaxIm after this and put that into the guide sensor angle. Enjoy!

One other thing, is there really a -122 step focus change going to Blue? I assume you ran the FocusOffsets script to get the offsets? I'm asking to see if maybe the guider is not focused well? Have you looked at the guider image? Is it snappy?

Before I get on with you I'd like to suggest a "back to basics" test.. I have a feeling that your "go for broke" tests are hiding something simple...

1. Set the ACP Guiding settings for the correct configuration given your guider, minimum 0.1 sec exposure, max 10 seconds, initial 2 seconds.
2. Calibrate in the east, do not play with any MaxIm settings,and do not have MaxIm's Observatory Control connected to anything, especially the mount.
3. Set MaxIm for single-star guiding for the time being.
4. Set up a test 600 sec exposure at some random place in the east with no bright stars or deep sky object in the field, just stars, no dithering, 0-PA if you have a rotator, super bright guide star Mag 6 or so
5. Run it, verify that the super bright star is in the field, look at the guider image, watch the track box, and watch the guiding log.

Let me know how this goes. I will get on with you if needed to work this out. Don't fret, we'll get it going!

[Samuel Lising]

Hi Bob.

I haven't played with the flip settings yet and have maxim set at "do not change" for guider axis on pier flip. I also have been able to guide both in the east and west without ACP.

Late last night however, there was some good weather and I wondered if perhaps the problem was due to multi star guiding, so I tried setting maxim back to just a 32x32 tracking box. It seemed to work. ACP was able to start guiding and didn't have the huge initial x, y errors seen previously when trying to start guiding. I tried several times to get ACP to start guiding in both the east and the west with multi star guiding but with every attempt I made, ACP could not get maxim to start guiding. It seemed that every time it tried to start guiding, the beginning X and Y errors were to large to guide out and start the guiding process. I also noticed this on the logs I posted.

When I tested guiding just with Maxim, multi star guiding worked fine. The X,Y error values started between .5 and .8 and guiding started almost immediately.

For some reason when using ACP with multi star guiding turned on, it seems that in most cases the x, y error is too large to begin guiding. I believe I have read that ACP should not have a problem with multi star guiding. Is there some thing I need to set in ACP or maxim to make multi star guiding work?

Thanks.
Sam

[Bob Denny]

Oh, OK, sorry I was just making guesses based on the info you provided so I could help...

There were corrections to the errors reported by the multi-star guiding in some version of MaxIm 6. Are you running 6.08 (I think that's the latest)? If so we may need to work on this after Diffraction opens again. I too will be mostly offline till the 5th, starting this afternoon.

[Samuel Lising]

Hi Bob.

No problem - I don't think I was being as clear with my description as I could have been (too much good holiday food has made me a bit fuzzy!)

Thanks for getting back to me. Yes, I've been using the latest version of Maxim. From what I can tell, it almost seems like there is a different guide star being used from when maxim first starts the guiding process to when ACP begins receiving the X,Y error values. So the error values are too big to start guiding. It does periodically work though.

It seems like I can guide now with a regular guiding box and I will test some more tonight. No rush with solving multi star guiding. Have a great holiday!

Best,
Sam

[Samuel Lising]

Hi Bob.

I've tested this over a couple of nights with the same result. When using Multi-star guiding, for many targets, the starting error is too large to correct and imaging can not start. I'm using the latest version of Maxim DL (v6.08).

Also wondering if anyone else has experienced this with multi-star guiding? I haven't seen any posts regarding this issue.

Best,
Sam

[Bob Denny]

This sounds like a bug that they already fixed in MaxIm. I researched this on the MaxIm forum. Does this sound like your problem?

Multi-Star Guiding Initialization

[Samuel Lising]

Hi Bob.

I just read through the forum link and it's a similar problem. However, I've been able to initiate multi star guiding without the huge errors when using Maxim on it's own. I seem to only see the issue when using multi star guiding with ACP. Also, it doesn't seem like they have solved the problem yet. The thread just ends with a comment that they are still investigating.

I will post my issue to their site as well to see if they've found a resolution. Is there any interaction between ACP and maxim in regards to initiating multi star guiding or is ACP completely passive? I also noticed that in the release notes of Maxim (6.08) they added multi star guiding scripting control. Would this added scripting control require ACP to handle multi star guiding differently from regular guiding?

Thanks.
Sam

[Bob Denny]

OK, good. The problem appears when an external program (ACP, CCD Autopilot, etc) uses the guider errors via MaxIm's programming interface. That's why you don't see it when you use it internal to MaxIm. Let me know. and if needed I can raise the issue with them as well. I actually thought I had but the old ticket system is offline, replaced by the (much better) forum system.

[Samuel Lising]

Hi Bob.

There was a response on the Maxim forum today that seem to answer the issue of multi star guiding in ACP. Here's the most significant post in the thread:

What I'm saying is: End users, don't use multistar guiding with CCD Commander, and possibly ACP (I don't know how they implement guide-star selection; but unless they let MaxIm DL take the reins, it probably won't work). If you use single star guiding, it will work as it always has. Multistar Guiding has the potential to work in those programs, but it may take some work on behalf of the developers to make that happen.

For developers/scripters: our multistar guiding algorithm does not support manual selection of guidestars, and attempting to manually set a guide star is incompatible with the multistar guiding algorithm. Please refer to the help manual on how to handle Multistar Guiding via the scripting interface.

From what I understand, ACP is selects a guide star and this is probably the cause for the huge guide errors that I've seen. The guide star selected by ACP and the star selected by the multi star guide algorithm in some instances may be quite far apart causing the huge error. I'm guessing this will be quite a bit of work to implement. I've had some pretty good results in the few tests I've done with multi star guiding since Maxim 6 got release. I'm not sure if this feature is on your development road map, but I would be interested in this feature.

Best,
Sam