I'm trying to understanding how to further fine tune my guiding with PHD and to get a better understanding what the PHD graph should look like, would much appreciate if you can post:
1. Model of your guide scope
2. Model of your mount
3. Your PHD graph set to 100 iteration
4. Your PHD settings
Looking at your specs assuming you guide with a Q-guider (QHY5) your image scale is approx. 2.7arcsec / pixel so your guiding is in the vicinity of +/- 2.6 arcsec. Not bad at all on an HEQ5. Your RMS is very good, your OSC a bit low though. 0.3-0.4 is more the norm?
Looking at your specs assuming you guide with a Q-guider (QHY5) your image scale is approx. 2.7arcsec / pixel so your guiding is in the vicinity of +/- 2.6 arcsec. Not bad at all on an HEQ5. Your RMS is very good, your OSC a bit low though. 0.3-0.4 is more the norm?
Yes it's a QHY5 guide camera, thank you for the advise however, how do I improve my OSC.
I'd try up the Hysteresis from 10 to 20 and see if that improves things.
Secondly once you've optimised Hysteresis try varying the minimum move before guide command issued between 0.1 - 0.8 in say 0.05 increments to see what works best for you. You could be chasing the seeing!
Yes it's a QHY5 guide camera, thank you for the advise however, how do I improve my OSC.
Cheers
Eddie
Everytime I got a low (<0.2) or high (>0.5) OSC I was guiding close to the celestial equator. 90% of the time it was a balance issue (west too heavy or too well balanced with high backlash in RA) or poor polar alignment. The rest 10% was the agressiveness in PHD. Lowering it to 60% smoothed things up. Seeing can make your guiding turn to .... real quick too, so something to consider. Bad seeing, longer guider exp. (2-3s). Normally 1s works nicely for me in good seeing. Your RMS is very good though. I typically get 0.3-0.4. May be because I dither and bump the mount all the time though.
To improve your OSC. Be east heavy all the time, get your polar alignment spot-on so you hardly correct in DEC, guide every 1-2s minimum. Pick a bright enough guide star. I found picking the right star makes a hell of a difference. Not too bright but not to dim either. Problem is that PHD will guide on really faint stars too, but the guiding IMO is not as good?
I'd try up the Hysteresis from 10 to 20 and see if that improves things.
Secondly once you've optimised Hysteresis try varying the minimum move before guide command issued between 0.1 - 0.8 in say 0.05 increments to see what works best for you. You could be chasing the seeing!
Thank you, I'll give Hysteresis a shot
Quote:
Originally Posted by multiweb
Everytime I got a low (<0.2) or high (>0.5) OSC I was guiding close to the celestial equator. 90% of the time it was a balance issue (west too heavy or too well balanced with high backlash in RA) or poor polar alignment. The rest 10% was the agressiveness in PHD. Lowering it to 60% smoothed things up. Seeing can make your guiding turn to .... real quick too, so something to consider. Bad seeing, longer guider exp. (2-3s). Normally 1s works nicely for me in good seeing. Your RMS is very good though. I typically get 0.3-0.4. May be because I dither and bump the mount all the time though.
To improve your OSC. Be east heavy all the time, get your polar alignment spot-on so you hardly correct in DEC, guide every 1-2s minimum. Pick a bright enough guide star. I found picking the right star makes a hell of a difference. Not too bright but not to dim either. Problem is that PHD will guide on really faint stars too, but the guiding IMO is not as good?
Good points, thank you, will give balancing more to the east and Hysteresis a shot.
I take it the whole idea of the graph is to flat line it as much as possible.
Yes, a flat line is the target, however this would essentially mean 0 guide corrections/tracking errors... With a HEQ5, a flat line is unlikely... Providing your tracking errors and guide corrections are lower than your imaging resolution you can be assured that your images will come out just fine...
I don't have screenshots of my Maxim guide graph, however I run 20 ~ 30 min subs day in day out with my HEQ5 and my guide graph looks much the same as yours, oscillations are perhaps a little smoother, but I think Matt's suggestion about increasing the hysteresis should sort that out for you.
If you flat line at your image scale you will be well under +/1 arcsec/pixel
Quote:
Originally Posted by AlexN
Yes, a flat line is the target, however this would essentially mean 0 guide corrections/tracking errors... With a HEQ5, a flat line is unlikely... Providing your tracking errors and guide corrections are lower than your imaging resolution you can be assured that your images will come out just fine...
I don't have screenshots of my Maxim guide graph, however I run 20 ~ 30 min subs day in day out with my HEQ5 and my guide graph looks much the same as yours, oscillations are perhaps a little smoother, but I think Matt's suggestion about increasing the hysteresis should sort that out for you.
Ok I get the idea now with the graph, from all the playing around I've done so far it looks like I need to zero in on the Hysteresis, Min motion (pixels), and balance more to the east (as my current setup is perfect balance all round).
I'd never expect a flat line - for one seeing and scintillation would always produce errors - you just have to try and adjust PHD not to respond to these errors where possible.
Craig is considering multi-star guiding - I've dropped him a few articles, notes and excel files comparing guide performane from 1 -> 20 stars (averaging each star's local gude error can significant improve guiding out drift - not seeing). Hope will get it before Feb 2010 - as even a 5 star guide should halve guide errors based on the MaximDL Multi-star guide add on results.
Hi Marc
I am still a bit confused about "east heavy" when referring to DEC movement since this is in the north south direction. How is "east heavy" achieved to assist DEC guiding. Cheers Peter
Hi Marc
I am still a bit confused about "east heavy" when referring to DEC movement since this is in the north south direction. How is "east heavy" achieved to assist DEC guiding. Cheers Peter
East heavy is for RA. In DEC a slight imbalance is also recommended so you push the load, unless you have zero backlash and no drift. DEC drift will affect your RA guiding regardless.
Thanks Marc - got it now! Yes I normally have my balancing east heavy so that RA has a load to work against but have usually had my DEC balance about neutral, so I will change this to hopefully fix the mysterious jumps I have been seeing in my PHD DEC guiding (as mentioned in another thread). Cheers Peter
I did a Google search to find out the meaning of osc and rms in PHD and ended back here. Some good information.
Scope: WO M72
Mount: EQ6 Pro
I have also used PEC which probably helps with the guiding. I tried to increase the osc but anything I did decreased the value. I think the guiding is pretty good.
I did a Google search to find out the meaning of osc and rms in PHD and ended back here. Some good information.
Scope: WO M72
Mount: EQ6 Pro
I have also used PEC which probably helps with the guiding. I tried to increase the osc but anything I did decreased the value. I think the guiding is pretty good.
Rob that's excellent guiding, thank you for posting, can you advise how you did and updated your PEC recording to the handset.
Rob that's excellent guiding, thank you for posting, can you advise how you did and updated your PEC recording to the handset.
Cheers
Eddie
Hi Eddie,
I used Pempro to measure the PE and create the PEC. You can then send it to EQMOD or to the mount. The program works very well with the Orion Starshoot so should work with the QHY5 as I believe they are the same camera. A very useful function in Pempro is the polar alignment tool. You can try it for 60 days for free.
I did a Google search to find out the meaning of osc and rms in PHD and ended back here. Some good information.
Scope: WO M72
Mount: EQ6 Pro
I have also used PEC which probably helps with the guiding. I tried to increase the osc but anything I did decreased the value. I think the guiding is pretty good.
Please Post your PHD Graph and Settings
