What is my PHD graph telling me?

[vanwonky]

Hi All, I have attached two screen shots of PHD from last night hoping someone could tell me if the graphs are looking OK or not as far as the guiding goes. Having said that the star in the 30 minutes between the two screen shots did not move from beneath the wire of the 10mm reticle so I think things aren't too bad. However is there an ideal waveform to be trying to achieve and can anybody see where I might make improvements? To me it seems a bit wonky despite the good result.

Unfortunately time ran away and I didn't get any long exposure shots to really check results.

Cheers
Dave

[cventer]

Looks like you are over correcting every RA move. Guider is moving mount too far then moving too far back again.

Try reducing aggressiveness even more or and dialing the Max RA down as well

Also make sure the backlash compensation on your mount is off and you are slightly weight biased to theeast side of the mount to keep the worm gear from bumping against the direction it it moving which can also cause these kinds of oscillations.

Also what focal lenght is your guide camera ? Depending on answer to this your min move may be too low as well.

[mithrandir]

The vertical grid spacing on the graph is 1 pixel on your guide camera. You need to work out the ASP for both the guide and imaging cameras.

For example I might get 0.79 ASP on the image camera at f/10 and 3.81 ASP on the guide camera. So I need to keep within 0.25 of the centre line to stay within 1 pixel on the image camera.

If I change imaging scopes I get 1.69 ASP at f/7.5, so then I only need to stay within 0.5 of the centre line.

[vanwonky]

The guider is an Orion 80mm . 400mm FL. Backlash comp on the mount? Is that in EQmod? Will look at weight bias. Is reducing aggressiveness etc. something that you can do whilst guiding or do you need to have stopped guiding first?




Sorry Andrew - "ASP"?

[mithrandir]

arc seconds per pixel
Download a copy of CCDCalc (http://www.newastro.com/book_new/camera_app.php)
If your scopes and cameras are not in its database you will have to add them.

An Orion ST80 is aperture 80mm and focal ratio f/5.
You didn't say what the other bits are, but for example an SSAG or QHY5 (same horse, different colour) is 1280 x 1024 pixels and the pixels are 5.2 x 5.2 uM.

Then choosing Orion ST80 for the scope and SSAG for the camera it tells me that would be 2.68 ASP.

Hmmm, maybe I should switch guide cameras. That would be closer to my image scale and the SSAG is USB2, and I guess I don't really need a cooled guide camera.

[cventer]

You can change aggressiveness and min move on the fly while guiding. You can see the settings below the graph.

I'm am not across eqmod so don't know where you set backlash compensation for your setup.

Tell us what your auto guider is and your main imaging camera and scope.

This is so we can help work out what your mimove settings needs to be to give you round stars.

So provide the following

Main scope: focal length and aperture
Guide scope : focal length and aperture
Imaging camera: chip pixel size in um eg 5.5 um
Guide camera: chip pixel size in um


The relationship between your guide scope and main scope is important to help get guide settings right.

As mentioned before ccdcald is a great tool to work out your arc sec per pixel for your guider and your main imaging scope.

[vanwonky]

Thanks for that info guys. Appreciated. It is another technical step that I haven't even read about yet so that is fantastic. I will look at CCDcalc and see if I can get my peabrain around it.

Meanwhile: Main Scope - Sharpstar refractor. Guide Scope: Orion StarShoot AutoGuider + 80mm Short Tube

Main scope: focal length and aperture - 690mm , F/6.5
Guide scope : focal length and aperture - 400mm , F/5.0
Imaging camera: chip pixel size: 6.42μm
Guide camera: chip pixel size: 5.2μm

Cheers
Dave

Edit: Hmmm - found this site (http://www.trackthestars.com/ccd-calculator.html) with this calculator.

[cventer]

Ok Using CCD Calc I get the following numbers for your setup:

Guider 2.68 arc sec per pixel

Camera: 1.92 arc sec per pixel


Good news is this guider and camera combo should theoritecically allow you to guide at around 0.6 arc sec per pixel accuracy. Reality depends on your mount, camera connections. ie how tight your guide scope and camera and main camera and scope are coupled. How tight your guidescope is coupled to main refractor etc... Double check all of these things and tighten evrything up as much as possible.

So depending on your local seeing conditions. Lets assume moderate seeing of 2.5 arc sec.

You need to find guide settings that keep movement to within the limits of seeing. ie within 2.5 arc sec

Given that 1 pixel on your guide camera sees 2.68 arc sec of sky if you can keep your guiding to within 1 pixel on PHD you should have satisfactory results.

from your graphs you posted you can see your osciallation index is at 0.41. This is saying there is a 41 % chance that at the next exposure you take on guide cam will likley cause a change direction in RA. Try get this down to around 30%

Your RMS is .29 which is acualy not too bad and most of your data is falling within + or - half a pixel which meats your typical seeing conditions of 2.5 arc sec.

Other than the fact you are getting so much oscialltion from exposure to exposure things ar elooking pretty good. This osciallation wil effectively just blur the image a little and wont be as sharp as you could get. Your stars will be round though but just a little smeared.

Try tame the osciallation by turning off bakclash comp if its on, reducing the agressiveness and max RA. I would try half the max RA first and see how you go. then try reduce egressivenes if this has no results.

Also your minimum move is quite low. You have it set at 0.15 of a pixel. This means you are telling PHD to make an RA drive correction every time the star deviates by more than .15 of a of a pixel. Remember you seeing is only probably about 2 - 3 arc sec. so with this low a number your mount will be chasing the seeing all over the place which can also account for the oscillation. try increase your min move to something like 0.4 instread of .15

In summary:
1. Set your RA Aggressiveness lower to to reduce errors
2. Set your Maximum RA move as low as possible
3. Set minimum move to ignore seeing variation. ( I think pushing this up to .4 or .5 will remove your osicllation)

if you want some theory here is a good arcticle http://www.wilmslowastro.com/tips/autoguiding.htm

[vanwonky]

Thanks for the effort of that detailed explanation. Really good help for a beginner to get. Looking forward to getting out again and trying with these settings although that might not be till the weekend. Had a quick install and look at CCD Calc. Another awesome free program!
Cheers
Dave

[CarlJoseph]

This thread has been extremely to me - someone just thinking of getting started in guiding.

I found this article which others might find useful to ... http://www.wilmslowastro.com/tips/autoguiding.htm It might help to explain the various numbers/maths described here.

Cheers,
Af.

[vanwonky]

Nice work Carl but you can't take all the credit! The link was mentioned 2 posts up ;). I did think as well that this thread would be a good resource for newbies to guiding. I am now trying to get my head around the whole 'arcsecond / G-Ratio' thing. It is funny how you can go out and just take an image with the basics and get some results but then when someone mentions the deeper stuff you just have too learn it. It is great that people will go to the trouble to help.
Cheers
Dave

[CarlJoseph]

Damn it! Beat me by this much! ;) Actually, beat me by heaps. Totally missed that link hiding there. :D

[bartman]

Thanks for posting all this info guys.
Last week I had a few goes with the guide scope and PHD, bit more playing than anything serious, but to get my head around all those weird settings took for ever. Youtube didnt help to much.
But this info will help me along that next rung!
Cheers
Bartman

[John K]

So just a quick question on this as I am trying to do some deep sky stuff myself, if my camera setup is 0.74 AS/P and my Guidescope is 3.85 AS/P, does that mean really that I need a Guidescope set up with a lower AS/P value to better match my camera setup?

In essence, I am trying to guide a 12.5" f/5 with a Canon 400D and an 80mm refractor 300mm FL with an SPC900 webcam.