Jump to content


Photo

SSAG/PHD Questions

  • Please log in to reply
19 replies to this topic

#1 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 22 February 2013 - 07:50 PM

Hello again,

Can someone help me with two questions? 1. What is causing the red line to spike? 2. Is it more inherentlly difficult to guide in a light polluted sky? The highest SNR number i can get is around 17 and even at that i am warned of loosing connection. I hope this makes some kind of sense.

Thanks, Victor :)
Los Angeles, Calif.

Attached Files



#2 ghataa

ghataa

    Viking 1

  • -----
  • Posts: 625
  • Joined: 20 Jun 2011
  • Loc: Central, NJ

Posted 22 February 2013 - 08:04 PM

Try adjusting your dec setting systematically from higher to lower
to see if you can get a smoother response.

George

#3 ghataa

ghataa

    Viking 1

  • -----
  • Posts: 625
  • Joined: 20 Jun 2011
  • Loc: Central, NJ

Posted 22 February 2013 - 08:05 PM

Delete


#4 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 22 February 2013 - 08:06 PM

Will do,

Thank you, Victor

#5 powerstroke01

powerstroke01

    Viking 1

  • -----
  • Posts: 531
  • Joined: 19 Jun 2007
  • Loc: Western Sierra Nevada Foothills

Posted 22 February 2013 - 08:22 PM

What kinda mount? Looks like to ME that youve got alot of backlash or something. really wierd.

#6 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 22 February 2013 - 08:27 PM

ASCG5
Victor

#7 zerro1

zerro1

    Fly Me to the Moon

  • *****
  • Posts: 5871
  • Joined: 02 Aug 2009
  • Loc: Smokey Point , 48.12°N 122.25°W Elevation:512 ft

Posted 22 February 2013 - 09:07 PM

That scale is really hard to interpret, can you get a screen capture in the 100 range instead. CG5's have a history of developing stiction in the DEC. If you have good PA(and are certain of it) then I would think that the DEC is hanging up a bit. Try increasing the MAX DEC to 500ms or even higher. The reason I say to increase is that by your graph> corrections are being sent but they are not making enough of a difference. Could be backlash? Could be stiction? You want the corrections to be enough to correct before it causes drama in the image you are trying to capture.

#8 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 22 February 2013 - 09:30 PM

Thanks Robert,

I will also try increasing MAX DEC, and will double check polar align.

Victor

#9 Orion58

Orion58

    Viking 1

  • -----
  • Posts: 914
  • Joined: 28 Oct 2011
  • Loc: Southern Wisconsin

Posted 23 February 2013 - 09:08 AM

Do you know which way your drift is? It is desirable to only guide in one direction. For example: if your drift is to the south make your OTA slightly south heavy and select (guide) north in the brain setting of PHD. Guiding both ways in DEC is an invitation to backlash. If you are guiding in one direction only with your OTA heavy to the opposite side you will minimize backlash.

And you may have to increase your Max Dec setting as you mentioned.

Good luck

#10 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 23 February 2013 - 08:13 PM

I"m back,

Ok, here" what i did.

1. Adj"d left azimuth knob 1/8" clockwise, i was off,
2. made east heavy by 1/2" more up the counterweight bar,the cold temperature effects the balance on this mount.it"s sticky somewhere,
3. guided in Auto and all was fine.
4. guided with Dec guiding off and was able to get 5 min exposures w/o trails. although 3 mins is my max here and that"s with a lp filter.
i heard that if your polar alignment is good enough, it"s not neccessary to use Dec guide. Don"t know if that"s true or not.

Thanks Bruce, Victor

#11 Orion58

Orion58

    Viking 1

  • -----
  • Posts: 914
  • Joined: 28 Oct 2011
  • Loc: Southern Wisconsin

Posted 26 February 2013 - 09:23 AM

i heard that if your polar alignment is good enough, it"s not neccessary to use Dec guide. Don"t know if that"s true or not.


Yes, that is true - the trick, of course, is to get it "good enough". You'll only know when you reach that point by trial and error - an alternative, which I described before is to guide in only one direction.

If you have to set up your scope every night, I doubt you will want to spend the time getting your polar alignment to the point that you don't have to guide in DEC - I think you'll find that if you get it "close" then guide (DEC) in one direction, you'll be imaging a lot sooner.

Good luck!

EDIT: and remember that you need to balance your OTA in DEC also, not just RA. You were correct in making it slightly east heavy in RA but DEC is just as important. As I mentioned before, if your polar alignment shows a slight south drift make your OTA slightly south heavy (because you will be guiding north) and vice versa if you have north drift.

#12 michael hester

michael hester

    Viking 1

  • -----
  • Posts: 640
  • Joined: 28 Nov 2008

Posted 26 February 2013 - 03:48 PM

There are multiple things happening in that graph.

The rough spikes are the result of the minimum motion setting. That's the .15 in the middle of the screen. This means that the star can't deviate more than 15/100ths of a pixel away from where it started. To put it mildly that is EXTREMELY accurate. More accurate than most mounts can handle and certainly more accurate than you need to shoot with your DSLR. Change this to .5 or 1.
The reason is your mount's dec backlash. When the spikes go positive PHD will try to correct southward until the star is .15 or less away from where it started. When the spikes are negative PHD will try to correct north until the star is .15 away. Your backlash is large enough that the star actually drifts back to where it should be before it reaches the other side of the backlash. Once it drifts south phd will correct to the north. Since your mount is on the north side of the backlash PHD ends up generating a spike in movement. It scales its correction based on how far away the star is but it will overshoot and now the star is back north again. It starts guiding south and the cycle repeats.

Another thing that is happening is the slow southward drift in the star's motion. This is error in your polar alignment (basically what you're intending to correct with PHD). If you tighten up your polar alignment by drift aligning you will minimize this drift and the impact it has on your images.

When the guider makes corrections it causes field rotation. This rotation's strength is based on how frequently the guider has to correct the aim. With a drift as strong as your graph shows you'll be limited to 5 minute exposure before the drift becomes apparent in your images.

Finally the roughness of the lines is caused by seeing and camera noise. Seeing has more of an effect than noise though. This usually doesn't bother deep sky images but if you're trying to get detail on a high magnification the seeing will matter.

#13 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 27 February 2013 - 11:23 AM

Thanks for the replies guys,

Bruce, I did have to tweak my polar align. I just learned how a VERY little adjustment of the knobs and alt screws can make a big difference. an eye opener. Fortunately i have my mount on a permanent pier and should make things a little easier. I am also getting better at balancing. Noticing how fast either the south side of the scope or the counterweights fall from being horizontal. hope that makes sense.
michael hester,I tried setting the min. motion to .5 and 1.0 but the blue line kept going down. For now, .15 seems to work best. I also will feel more comfortable with 3 minute or less exposures for now because of the "seeing" as you mentioned, my equipment, and skill level. I think my stars are tight enough for Deep Sky Stacker to except more subs.
But please let me know if there are any other adjustments i can make for now.
P.S. allmost forgot, Why does PHD lose connection during calibration? I keep getting a "not responding" message on the top of the screen and even the PHD icon will dissapear for a fraction of a secound from the task bar.

Thank you, Victor

Attached Files



#14 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 27 February 2013 - 11:34 AM

Pic no. 2
This is the 4 min pic that goes with the graph above.
Vic

Attached Files



#15 guyroch

guyroch

    Vendor (BackyardEOS)

  • *****
  • Vendors
  • Posts: 2955
  • Joined: 22 Jan 2008
  • Loc: Under the clouds!

Posted 27 February 2013 - 12:46 PM

WOW! What an improvement!

Guylain

#16 Mike7Mak

Mike7Mak

    Apollo

  • *****
  • Posts: 1277
  • Joined: 07 Dec 2011
  • Loc: New York

Posted 27 February 2013 - 01:25 PM

A couple things...

The first graph does look like PA drift combined with dec axis stiction. But increasing MAX DEC won't do a thing. MAX DEC 'limits' the ultimate size of guide corrections, it doesn't force them to be larger.

As the dec axis drifts farther and farther PHD sends ever increasing guide corrections up to the MAX DEC limit. When those corrections get large enough to overcome stiction the mount jumps back to the line (probably past). If MAX DEC wasn't already high enough the mount would never correct. In other words guide corrections overcome the stiction before getting as large as the MAX DEC setting.

Also, it's impossible to say what MIN MOTION 'should' be without some idea of the guiding scale. Guiding at 10 arcsec/pix will use much smaller 'min' values than guiding at say 1 arcsec/pix.

#17 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 27 February 2013 - 05:29 PM

Guylain, Thanks to the people here, i am getting closer.Polar alignment and balance very important.Bye the way, love your program.Sorry, couldn"t pass that up.
Mike, i will drop that Mx Dec down to 150 and see what happens, i think that is the default anyway.
Just between you and me, i have no clue how this program works. i do not understand the terminology and just going by rote kindof. i am trying to comprehend what these parameters do but i am pretty slow to begin with.

Thanks fellas, Victor

#18 Mike7Mak

Mike7Mak

    Apollo

  • *****
  • Posts: 1277
  • Joined: 07 Dec 2011
  • Loc: New York

Posted 27 February 2013 - 06:00 PM

Mike, i will drop that Mx Dec down to 150 and see what happens, i think that is the default anyway.
Just between you and me, i have no clue how this program works.

NO, no, no, that's not what I meant. You can increase MAX DEC to 1000 and forget about it. It's not really useful for tuning guide performance. It's just there to limit the duration of pulses in the event of large movements of the mount. Once PHD is calling for corrections that large the image is likely ruined anyway.

With the performance indicated by the first graph, (the slow drift down and the sudden jump back to the line), reducing MAX DEC could cause it to never get a guide pulse long enough to move the mount. It would just continue to drift down.

You may have to be a member of Stark's yahoo group to access this file, but it's a pdf called 'Guide to Guiding' and definitely worth a read. It explains a lot of what's going on in PHD.

http://tech.groups.y...e/files/PHD ...

#19 pianoman2

pianoman2

    Vostok 1

  • -----
  • Posts: 171
  • Joined: 13 Apr 2010
  • Loc: Los Angeles county,Calif.

Posted 27 February 2013 - 06:32 PM

Thanks Mike, I will bump it up to 1000.
Victor

#20 Mike7Mak

Mike7Mak

    Apollo

  • *****
  • Posts: 1277
  • Joined: 07 Dec 2011
  • Loc: New York

Posted 27 February 2013 - 06:46 PM

Thanks Mike, I will bump it up to 1000.
Victor

Or leave it at 400 if you want. It seems to be working there. My point was that MAX DEC isn't really intended as a 'tuning' parameter. It's just an upper limit for when things get squirrely.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics