Noisy PHD guiding - opinions please!

[NigeB]

Dear All,

I'll preface my question by saying that I'm aware of the need to do more diagnostics on my setup to identify issues affecting guiding. My polar alignment is also off by a few arcmins (horizon obstructions are making drift alignment tricky - a work in progress). But what I'd like to understand is: what is reasonable? And how do others see the effects of seeing impact on their guiding, particularly in the UK where we don't have Barbados-quality skies so often...With few clear nights to play with, I'm reluctant to spend time chasing down the last few % of performance. But I don't really have a good feel for where I am on the scale of Abysmal ---> Nirvana for this climate.

I'm using a TOA-150 on a Mesu-200 (Servocat version), Lodestar (version 1) autoguider, QHY8 imager, a TS Off axis guider, and PHD2.

Calibration at the beginning of each run takes place without any problem - I find PHD typically uses about 30 steps to nail this down before switching into guide mode.

I'm not seeing any evidence of periodic error, but I'm finding that both RA and Dec exhibit apparently random fluctuations, shown in the attached screenshot from last night's session. At a focal length of 1100 mm and a Lodestar pixel size of 8.3 microns, the plate scale is 1.56 arcsec per pixel. Guiding last night was at 1 second intervals, but I've seen similar behaviour at 2 and 4 seconds.

As the screenshot from PHD Log Viewer shows,the total RMS variation in the star position on the Lodestar is 0.71" or slighly less than half a pixel. Dec is the main contributor (0.61", 0.39 px). Excursions can be up to ~1.1 arcsec. This was a run imaging NGC7331, whose altitude ranged from about 60 to 70 degrees in the time I was out. I've only shown a small section of the graph, but the rest of the run is similar.

The seeing looked to be "fair" with occasional moments of "good" based on the Pickering scale - I gauged this from visual observation through the telescope at a magnification of 275. The observatory floor is isolated from the pier, and I wasn't in for most of the time. Wires are all tied up so there should be no snagging etc.

I'd appreciate views and opinions!

Thanks

Nigel

[glowingturnip]

looks pretty good to me - imaging scale of 1.56" per pixel and guiding error of 0.6", so you're guiding to better than pixel accuracy in the final image.

Your Dec corrections are mostly all in the same direction, suggests slight polar misalign, as you say, and accounts for the greater RMS in Dec

I see you don't have darks done - you should do a library in PHD, doesn't take long, then you won't have the risk of any hot or cold pixels falsely moving your centroids.

[r3i]

I'd agree with all that Stuart says.  When guiding with an OAG at longer focal lengths, the PHD graph does get choppy, so I wouldn't get too hung up about it.  Instead, look at the resultant images.

It might be worth playing around with the PHD settings to see if you can make any improvements.  For example the general consensus is to get PHD to calibrate in around 15 steps (give or take a few) so maybe increase your calibration step size.  Also try longer guiding exposures - I know you've said that even at higher exposures you're still getting rough plots but you could be 'chasing the seeing conditions' with a 1 second exposure and a longer exposure might help to smooth things a bit.  You have a high quality mount so you shouldn't need to be fighting against any mechanical inconsistencies (where a faster exposure rate can help).

[NigeB]

Hi Mike, Stuart,

(Sorry, have been away for work - currently stuck in Schiphol airport looking at some pretty ominous clouds...)

Many thanks for your replies: this is very helpful and gives me some comfort that things are probably not too bad. The images so far are a long way from the spectacular examples that are posted on SGL, but I've not had any dropped subs and the stars appear round (I need to work on my focus as well as the polar alignment though...)

I'll definitely work on step size adjustment, guide integration time and the dark library issue - good tips!

Thanks again for replying - much appreciated.

Nigel

[Zakalwe]

I wouldn't obsess too much about the graph, to be honest....it just shows that the guiding is working!   The final image is king...if you are getting tight, sharp stars then it's all good.

A bit of polar misalignment can be beneficial, especially on a mount with some backlash (which the Mesu has none of!)....it means that DEC guiding will be in one direction only, so you can run the DEC guiding in one direction. Plus you get a natural dither effect (to what effect that has I am not sure). As long as you are not getting field rotation (which bad polar alignment can bring) then I wouldn't obsess too much about it.  PHD2 does have a handy drift align tool if you want to tweak the PA a bit.

I think that it's probably worthwhile increasing your guide-camera exposure length. 1 second exposures might mean that you are "chasing the seeing"...a longer exposure will smooth that out. I am getting near the end of my new Mesu installation and it seems very happy guiding on 3 or 4 second sub exposures.