Off Axis Guiding

[gnomus]

I've made a first foray into off axis guiding.  It did not go terribly well.  I have the Atik OAG.  I am trying this on an Edge HD 8" without reducer (the Edge is already 'flattened'.

The first adjustment seems to me to be to position the prism at the correct height.  I did this by shooting flats against my light panel.  I would see a shadow on the flat from the prism.  I would then have to strip everything down, make an adjustment to prism height and then reshoot.  Because one can only adjust the height by manually getting hold of the prism, this tear down - adjust - set back up again cycle proved extremely tedious.  I wonder why the height adjustment can't be done with everything connected to the scope.  This would save a ton of time.

I'm also not sure if my method ends up setting the prism too far out of the field of view.  Should I accept some shadow from the prism and then use flats to eliminate this?

First night out was a total disaster.  I couldn't get focus at all.  So I gave up.  During the daytime I set the rig up on a different scope and played around until I got a focussed image on both Lodestar and main camera.

Last night, I think I achieved focus.  However I suffered from a severe lack of stars.  I tried various different bits of the sky (around M51 was one target).  I located only one star.  This was bright enough but the star was 'L' shaped.  I tried Arcturus.  This time I could see possibly only two stars on screen.  These were very dim - one had an SNR of around 3 on a 4 second binned exposure.  This was again 'L' shaped - indeed it looked like three blobs arranged in an L formation.  

Are there any clues in the above account which might give folks an idea as to what it is I am doing wrong this time?

This hobby keeps making you feel like a total beginner.

 

 

[michael8554]

Interesting quote from the OAG instructions:

".....if an SCT is used you’ll probably want to lower the prism as much as possible, to get better star shapes. "

Hmm,  the Edge is supposed to have a flat field without coma, but I wonder what the star shape is like with the prism lowered........?

Regarding finding guide stars, you need to generate a frame of your imaging camera FOV on your planned target in a planetarium, and guestimate where the OAG FOV is, and see what stars might be available.

Michael

[Davey-T]

Hi Steve, I use the Atik OAG on my 10"SCT, I think the funny shaped stars are normal, I get banana and seagull shaped ones but it doesn't seem to bother PHD.

Ive got the prism on the long side of the sensor and it doesn't need to be sticking down very far to catch guide stars.

Some parts of the sky are surprisingly short of stars especially this time of year trying to image little galaxies out of the galactic plane so you may have to inconveniently  rotate the whole thing.

Have you tried letting PHD auto select a guide star ? it appears to find stars invisible to my eyes

Dave

Are you using the 383? I'm only using 314 so I've got a bit more leeway with the prism position.

[gnomus]

7 hours ago, Davey-T said:

Hi Steve, I use the Atik OAG on my 10"SCT, I think the funny shaped stars are normal, I get banana and seagull shaped ones but it doesn't seem to bother PHD.

Ive got the prism on the long side of the sensor and it doesn't need to be sticking down very far to catch guide stars.

Some parts of the sky are surprisingly short of stars especially this time of year trying to image little galaxies out of the galactic plane so you may have to inconveniently  rotate the whole thing.

Have you tried letting PHD auto select a guide star ? it appears to find stars invisible to my eyes

Dave

Are you using the 383? I'm only using 314 so I've got a bit more leeway with the prism position.

Yes Davey the 383.  So I may need to rotate the whole OAG/wheel/camera set up to get stars?

Jings - bit of  a pain. 

[cfpendock]

I use an OAG on my C11, albeit with an SX wheel and OAG.  As Davey-T above, I arranged for the prism to be aligned with the sensor (Atik 4000 is square, so plenty of choice!).  This made a big difference - noticeably on the flats where there is no evidence of the prism at all.  I did this by looking down the scope tube - it was easy to see the prism and sensor.  I would be very surprised if you cannot rotate the OAG against the wheel. 

Normally the lodestar is super sensitive and I never have a problem finding the stars.

Chris

[gnomus]

2 hours ago, cfpendock said:

I use an OAG on my C11, albeit with an SX wheel and OAG.  As Davey-T above, I arranged for the prism to be aligned with the sensor (Atik 4000 is square, so plenty of choice!).  This made a big difference - noticeably on the flats where there is no evidence of the prism at all.  I did this by looking down the scope tube - it was easy to see the prism and sensor.  I would be very surprised if you cannot rotate the OAG against the wheel. 

Normally the lodestar is super sensitive and I never have a problem finding the stars.

Chris

The Atik OAG is bolted to the Atik Filter wheel so it is fixed.  But there is a simple mechanism for rotating the camera against the wheel.  I could therefore achieve the 'prism aligned with sensor'.

[ollypenrice]

14 minutes ago, gnomus said:

The Atik OAG is bolted to the Atik Filter wheel so it is fixed.  But there is a simple mechanism for rotating the camera against the wheel.  I could therefore achieve the 'prism aligned with sensor'.

That's right. The prism does need to 'intrude' on the long side of the sensor in order to go as deeply as possible into the light cone. I used the flats test to get it as far in as I could. Are you using a dark in PHD? If not it might help to do so. After that, longer guide subs? The Mesu is fine with 5 second subs for me.

Olly

[gnomus]

I'll have another play then.  Any ideas how many suitable stars I should expect to see in the OAG typically (if not pointing at the Milky Way that is)?

[ian_bird]

6 hours ago, ollypenrice said:

That's right. The prism does need to 'intrude' on the long side of the sensor in order to go as deeply as possible into the light cone. I used the flats test to get it as far in as I could. Are you using a dark in PHD? If not it might help to do so. After that, longer guide subs? The Mesu is fine with 5 second subs for me.

Olly

Five second guider subs! Blimey. So that's where I have been going wrong!

Cheers Olly!

(Heads off to shoot some longer Lodestar darks.) ;-)

[ollypenrice]

19 minutes ago, ian_bird said:

Five second guider subs! Blimey. So that's where I have been going wrong!

Cheers Olly!

(Heads off to shoot some longer Lodestar darks.) ;-)

Hang on Ian, it depends on the mount. I'm talking about the Mesu. I use 1.5 seconds on the Avalon, for instance. The Mesu has very low PE so it can manage happily with long subs. However, it is always worth trying all sorts of guiding parameters. Be aware of one thing though: if you go for ultra short subs you might get a lovely guide trace because as soon as it deviates from the guide star image the mount will get a correction. The problem is that ultra short subs mean the guide star image is being moved around by the seeing, so the mount is following the guide star image very nicely but the guide star image isn't following the real star very well at all!

Experiment is the only the way.

Olly

[ian_bird]

Thanks Olly. The thought had occurred to me - Mesu v CGEM.

Just shooting a bigger library of darks so I can play around - and as you say - experiment! Not that we get much time to experiment here in the UK. Now if there was only somewhere I could go that had clear, dark skies.  (I guess this smiley is doing pretty much what a guide star does!)

Cheers.

Ian

[gnomus]

Incidentally, I use PHD2's bad pixel map feature rather than a dark library.  Which is best?  There's only one way to find out ....... (Apologies to H. Hill).

[ollypenrice]

7 hours ago, gnomus said:

Incidentally, I use PHD2's bad pixel map feature rather than a dark library.  Which is best?  There's only one way to find out ....... (Apologies to H. Hill).

I doubt there's much in it. I confess to using neither but Will Do Better!

Olly

[opticalpath]

On 06/05/2016 at 03:50, ollypenrice said:

I doubt there's much in it. I confess to using neither but Will Do Better!

Olly

With the Lodestar and 3 sec. guide exposures I found that I got quite a few hot pixels that were occasionally troublesome.  Dark calibration solved that, and probably improves the SNR enough to make fainter guide stars usable, so worth it I would say.

Mind you, with 5 sec. guide exposures, I don't suppose there's any shortage of guide stars!

Adrian

[ncjunk]

I've been using 5 seconds with my mesu as well for the same reason Olly says.

I've just ordered a lodestar to see if it can pick up a few more guide stars than the qhy6 i have (plus i'm suffering from a deep hatred of qhy cameras and drivers).

 

[ollypenrice]

2 hours ago, ncjunk said:

 

I've just ordered a lodestar to see if it can pick up a few more guide stars than the qhy6 i have (plus i'm suffering from a deep hatred of qhy cameras and drivers).

 

Let's see if you get in as much trouble as I did for saying that, Neil!!!

Olly

[gnomus]

I am beginning to wonder if my difficulty in finding guide stars might be due to the fact that I have had to introduce ~6mm of spacing between the camera and the filter wheel because of an internal reflection issue (discussed elsewhere).  Would having the camera sensor further away from the prism than it is designed to be, mean that I cannot drop the prism down far enough, without it being visible by the sensor?  

[ollypenrice]

Trying to get my head round this. It means that the OAG is picking off its sample from the light cone when the cone is wider. Does this mean the density of the beam is lower? I think it might but I tend to get this kind of thing wrong... My poor brain says the OAG gets more light the further down the cone it goes, but what do better brains say?  Where's Steve Richards?

Olly

[Davey-T]

Just a thought Steve, have you had a look inside the OAG, when I got mine the prism was loose, it's only held in by a plastic screw and mine had rotated a bit.

Dave

[ncjunk]

4 hours ago, ollypenrice said:

Let's see if you get in as much trouble as I did for saying that, Neil!!!

Olly

Which bit the lodestar bit or the i´d like to set fire to the QHY and burn the eyes out of the people who write their drivers? I know I´m sitting on the fence a bit.

 

As Dave says above there is a possibility the prism can rotate a bit on some OAGs so it´s worth checking it looks like it´s lined up ok.

 

Olly, I see what you´re saying with the light cone. Would it be the case that the majority of the mirror would be ok as its collecting the same amount of light that is partially focused but that at the mirror borders some of that light is lost? A very bad explanation there for what I´m thinking but I hope you get the cut of my jib. 

[ollypenrice]

Thinking of a cross section through a light cone, further up the cone you get an O and further down you get an o. It's the same beam with the same number of incident photons so there are more photons per square mm in the o than the O. So more photons per pick-off prism in o than in O.

Olly

[gnomus]

Sorry Olly, but I've not had my medication yet today.  I agree that there would be a greater density of stars in 'o' than in 'O'.  Is my OAG dipping into the 'o' or the 'O'?  I think it is dipping into the 'o'.  But doesn't that also mean that the shadow it casts on the sensor is more likely to be problematic?  Is it possible that the adjustment I need to make to keep it out of the 'o' is finer than is possible using the rather crude adjustment mechanism in the Atik OAG?  If so, a very slight upward movement away from the 'o' could be carrying it well away from the 'O'.  Indeed, maybe it is nearly out of the 'O' altogether, hence the lack of guide stars.

If the forecast is accurate, I may try to get some screenshots tonight.

I think there may be sufficient thought disorder in the above to justify having me committed!

[gnomus]

No wait.  It's the opposite of what I said.  So my prism is in the 'O' and so less photons falling on the prism there?

I think I'm going to have to find a better soluion to my internal reflection problem.

[ncjunk]

2 hours ago, ollypenrice said:

Thinking of a cross section through a light cone, further up the cone you get an O and further down you get an o. It's the same beam with the same number of incident photons so there are more photons per square mm in the o than the O. So more photons per pick-off prism in o than in O.

Olly

Yes but i think at the point we're taking off the light all of the stars light is hitting the mirror, those doughnuts you get when......actually now i don't agree with myself any more...shouldn't there be some loss?

I've got Os circling my brain.

I have some stars on my oag and i've decided its Witch craft.

[gnomus]

8 minutes ago, ncjunk said:

Yes but i think at the point we're taking off the light all of the stars light is hitting the mirror, those doughnuts you get when......actually now i don't agree with myself any more...shouldn't there be some loss?

I've got Os circling my brain.

I have some stars on my oag and i've decided its Witch craft.

You could try doing what I did - breathing in and out of a brown paper bag.  That helped quite a bit, although I don't think I'll ever be quite the same again.