Going long with PHD

[g__day]

I am trying to get PHD tuned so it will allow me to take longer than 10 minute using a 127mm MAK at 750mm focal length into a Meade DSI to guide shots on a 2.3 metre focal length C9.25 into a Canon 400D. At last attempt I was seeing about 1 arc second un-corrected RA drift every minute - so I lept trying and tweaking variables.

Last night I stumbled to the fact setting my RA aggressiveness in the range 50 - 80% on 2-3 second guide intervals wasn't having much effect, so I change the setting to 90% and star trails halved! Trying again with it set to 100% produced first a 10 minute then a 15 minute shot with no star trails!

Moving targets and trying again procudes a box effect (maybe stars were binary and too close) cause I ened up with a strange yo-yo producing a 4 of everything in a tight box effect.

It was around 3:30am so I packed it in but I am definitiely going to investigate this more (hopefully not just with brute force testing).

Whilst normal I guide with settings like:

RA aggressiveness 80% (tried 60 - 80%)
RA Hystersis 10%
DEC guide auto
Min Pixel 1.20 (normally use 1.1 - 1.3)
Pulse guides 1.5 - 2.5 secs

My best results before yo-yo's came back was obtained on a faint star
with settings

RA aggressiveness 100%
RA Hystersis 10%
DEC guide auto
Min Pixel 0.20
Pulse guides 2.0 secs

I guess once flexure is minimised everywhere (the Losmandy side saddle and rings are as tight as I can get it by hand) - its seeing, polar alignment, PEC and paramter tuning for PHD - oh the fun.

So at 15 minutes I get 4 shots and hour to test settings on - slow going!
Can anyone think of a way to speed up my testing (other than vary a setting - try at 300 secs pass / fail - try at 600 - pass / fail - try at 900 pr more seconds?

Thanks Matt

[RB]

Thanks for the info Matt as I've just bought my DMK to use as a guider with PHD, so I'll keep an eye on your findings.

:thumbsup:

[Alchemy]

can only say i use mine DMK it works fine with phd on standard default settings on the g11 .... i have found 2 things though, first settings on mount to 1/2 sidereal adjust rate helped not 1x or 2x, and dont use big fat stars it seems to jump left right left right etc.

[g__day]

Agreed,

With big fat stars I assume PHD does a statistically wieghted centroid calculation - so turbulence could affect the shape of the star if its say 4 * 4 pixels wide and PHD has to guess which pixel - down to a tenth of a pixel it should judge the centroid from.

So I try and aim for the smallest stars that I can fine that don't dim out.

When I see stars jumping that much in a frame by frame shot I wonder what chance PHD (or any guide software has and think either my MAK needs collimination - and urgently) or the DSI / MAK combination just isn't really good!

I still wonder why PHD is issuing such a large move that it takes the guide star out of the search frame?

Possibly the star jumps say 1 pixel left by seeing then PHD issues a 1 pixel left correction - sees no change so does it again - then the seeing reverts and the star is 1 pixel to the right - but if this is the worse case then the star should only be 3 pixel out of position - not 12 or 14.

Amazing how frustrating this is!

[Terry B]

I think that it moves much more than this because it is correcting for backlash as well. i.e. it thinks there is 0.5 sec of backlash plus 1 pixel to move so it moves 0.5 secs at the speed it detected in its callibration plus 1 pixel further. This would be OK if your backlash is constant but this is not always the case.

[JohnH]

You might review my PHD guiding struggles here....

http://www.iceinspace.com.au/forum/showthread.php?t=18092

Som of the tips that worked for me, sort your mechnical and aligment issues out first then in no particular order...keep you correction speeds below siderial, turn backlash compensation off in the controller and software, calibrate in the region of sky you are going to guide in, do not saturate your guidestar image (can cause centroiding issues), consider slight defocus of guidestar, guide in one direction only for DEC (opposite the residual drift), unbalance your mount a bit to the E and rebalance after meridian flip...keep your guidelogs and review after each session.

[g__day]

John,

Thanks - yes I am really trying to nail polar alignment first, then I need to sort DEC backlash - will read your posts with great interest thank you.


I have turned DEC and RA backlash off on the hand controller. Also as my alignment gets better I tend to re-do PEC, using PHD set at 0.5 secs intervals 100% correction and no hystersis.
Matt