If I had any hair left i'd be ripping it all out by now! (aka Now Whats Wrong?)

[Tim]

</insert: half an hour of Mutley style grumbling, moaning, and half spoken expletives/>

As I have mentioned before, I am very keen to get some decent images at the native focal length of my C9.25. Just an hour spent on the Crab Nebula showed its potential, and I hope to add some "close up" shots to my growing scrapbook.

But getting nice stars over 5 mins is proving very difficult to me. I spent the whole of last night gathering pics of star trails it seems, and the night before.

As per recommendations, I have drift aligned the mount, and the star stays put in DEC for 10 mins at least, I havent tested longer.

Using PHD I think I must have tried every combo of settings last night to try and get it to hold the shot still.

RA aggressiveness tested from 50% to 150%. From the look of it it made no difference whatsoever to the trails.

The movement is in RA (at least I think so, its in the direction the stars move if you press LEFT/RIGHT on the hand controller)

The odd thing is, on PHD the stars dont seem to move around the screen at all, but there's a gradual creep down the screen (same as pressing RIGHT button on controller) on the photo.

Its not PE, as that shows up as something else, little wobbles, and its on every photo. In fact, since I drift aligned, its almost gotten worse.

I tried the following in PHD.

Turn off DEC guiding.

Pause between guide frames 0 - 5ms

I even turned off guiding all together and tbh the result was about the same as the guided shots. Targets near to and far from the NCP were all affected. I'm wondering if the guiding makes any difference whatsover. Move the mount via the controller though and PHD starts complaining.

I downloaded maxim, to try that, but it refused to find my DSI. The temperature really dropped at around 3.30 last night, and the sky was beautifuuly clear, and no moon. Such a waste of a night.

Anybody got any ideas?

Yours desperately :help: :)

TJ

[CELESCOPE]

Cant offer any help yet Tim ,this was aproblem i used to have before i set up my Dome and a fixed Pier, its so darn annoying getting a super image on and after about 5 mins the stars start going off shape ,i gave up then , but have not tried since , i now have the C8 and intend doing some c8 imaging soon, will the same problem arise now i have it all fixed , we will see ,if it still remains , then i will be with you trying to sort this problem out, like u i have polar and drift aligned , and can run 2/3 mins unguided on the EQ6 which is good in my book , that of course is with any of the Equinox 120 /80 and 66 , so we shall see .

Surely ppl like Rob and Dave ( CENTROID ) can offer some guidance here . or is it just our Mounts . ???? I did notice that Helen is doing very well with her set up with the C8 .also Martin B could offer some help here as well .

Rog

[steppenwolf]

TJ this sound so much like 'flexure'. Flexure occurs when the imaging telescope system and the guiding telescope system don't move exactly with one another. Notice I say 'system' - more in a minute.

You may think that the two are bolted together so all will be well but this is not always the case I am afraid as there is always the risk of flexure in the overall system. This flexure will cause the 'creep' in RA that you are seeing and there are several avenues to explore.

1. check the tightness of all mounting bolts - any looseness will result in 'sag' as the weight shifts as the system moves in an arc whilst tracking

2. sounds obvious but check that both CCDs are properly seated and clamped in the focus tubes

3. if you are using 'T' mounts, ensure that the three grub screws that hold the camera specific bayonet to the 'T' adapter are tight - these often work loose and you will then get a change in the optical path as the system tracks across the sky

4. Mirror flop - if the primary moves at all, you will get a trailing effect even though tracking is smack on

Hope that helps.

[SteveL]

I was going to say flexure as well, but then saw Steve say the same thing. If your polar align is accurate and still seeing trailing, its either flexure or mirror flop.

[Robd]

TJ

I have had all sorts of guiding problems myself. HEQ5 Pro, SC1 mod webcam and PHD. I think I got it sorted a couple of nights ago - the web cam was 180deg out :oops:

Two of your comments make me think that you might have done the same thing.

The movement is in RA (at least I think so, its in the direction the stars move if you press LEFT/RIGHT on the hand controller)

The odd thing is, on PHD the stars dont seem to move around the screen at all, but there's a gradual creep down the screen (same as pressing RIGHT button on controller) on the photo.

I would have expected that movement down the screen in phd would be DEC hence up down buttons not right/left.

Sue and I spent ages the other night trying to decide which way stars should move when each button is pressed, which way the start will drift if tracking is turned off and finally what difference it makes when pointing in different directions.

After rotating the webcam, and then discovering that dec guiding was off, so turned it on, I managed 3, 10min exposures with no drift.

Sounds rather basic, but hey it's the stupid things that usually catch us out.

Might not help you but at least will be a warning to others.

Rob

[Tim]

I 'think' mirror flop can be ruled out, the focus stays where it was, and at F10 even the slightest shift seems obvious, but is that the test?

As for flexure, hmm, I dont know how things could be any tighter, but I will check.

There is a plate mounted on to the EQ6 with a clamp that holds the rail on the C9.25, that is solid.

On the other end is the WO72 in guide rings. I align the two so that the DSI and the DSLR are both pointing at exactly the same thing. Now that I think about it, there does sometimes appear to be a bit of shift in the centralising of the target after slewing the scope. Is this perhaps the indication of flexure somewhere in there? I may still have the old clamp I used before, in which case I may be able to rigidly mount the WO72, so that might help.

Is there a definitive way to test for mirror flop alone? and would every exposure show the same trailing if it was mirror flop?

Thanks for the help, this is really driving me mad!

TJ

[MartinB]

Agree with the 2 Steve's. The key to the diagnosis of these guiding problems is to have a decent tracking log. Maxim shows a tracking error graph and also can keep a log. I don't know if you do the same with PHD. A tracking error graph will tell you whether the trailings are due to ocscillations or continuous movement in the same direction.

The fact that you are getting big movements of the overall image makes it very unlikely to be down to the guiding control since the guidestar would also be seen to move and would eventually be lost.

Mirror flop is mainly a problem crossing the meridian as the mirror changes balance point.

You could always confirm flexure/mirror flop by getting a Celestron off axis guiding gizmo and see if the problem goes away.

[CELESCOPE]

Dont think its mirror flop ,i tried all those things Tim , keeping the imaging by using one alignment star and imaging close by with the subject. where no mirror flop should happen , result same thing, as for flex Himm could be , never tried that , be interesting now with the C8 as its light and less weight on top with the 66 and everything is rock soiled , ,so will let you know how it goes the next clear night .

[Tim]

Thanks Rob. I had the DSI aligned so that when the stars drfited, they drifted along one of the axis on the bullseye overlay. It was then adjsuted in alt/az accordingly to enable the stars to not drift, either on the eastern horizon and the meridian. As far as I can see, there is no drift in DEC, in fact, just left to it, there's no drift anywhere.

Must be flop or flexure then.....both sound painful

TJ

[Tim]

Oh heavens no, dont say the OA guider is coming back out of its box

[Tim]

Martin, I was watching the guide graph last night for some time. There is a central line that runs through the middle, and the RA graph line was ticking along next to it, without massive peaks or troughs. With Dec guiding turned on there was the occasional bigger trough. the OSC index was given as 0.25, but I haven't the foggiest what that means!

Anybody happen to know why maxim wouldnt find the DSI?

Oh the other strange thing, by getting the polar alignment more accurate, it made the trailing worse than it was

[Ant]

Is it possible that the 72 just doesn't have the FL to guide the C9.25 at F10?

It is a 432mm FL scope trying to guide a 2350mm FL scope... Whats accurate in one isn't in the other. Maybe try adding a barlow in to increase the effective FL of the 72?

Just a thought?

Ant

[Tim]

Dunno Ant, works great the other way round though!!! I could try that.

[661-pete]

I use a 'cheat'. I don't have any autoguiding (and no budget to upgrade to it ) so I have a lot of PE in R.A., although my declination stability is fairly good. I get rid of the trails after imaging using Focusmagic. I know this isn't strictly the 'right' approach, there's no substitute for achieving proper stability of your images at the time of imaging, so if you can improve matters from the suggestions posted above, that's still the best.

But Focusmagic can turn an unusable set of subs into something that at least gives you a sort of result. Most of the images I've posted on forums have had some of the 'treatment'.

You need to buy the full registered product to get to work properly with it, but at US $45 it's not a huge spend. It works as a plug-in to PS or PaintshopPro, and, using the 'fix motion blur' function it will remove trailing up to 20 pixels (that's about 20 arcsec. at my scaling). It creates some artifacts like hollow stars, maybe someone has a suggestion how to get rid of those! What I do is apply de-trailing separately to each sub (after subtracting darks) before aligning and stacking.

Don't know how much help this is, but I thought I'd put this down anyway...

[Tim]

Thanks Pete, to be fair I could doctor the images in PS and most folk would never know, except for me, and thats the whole point really. Part of the attraction for me is getting past these niggles and actually doing it properly. There's more a sense of achievement which increases the satisfaction of a decent image.

If I can only find the clamps and bits I need, I think im gonna mount the WO on its dovetail, and use a barlow. Either that or buy another C9.25 to guide it with

TJ

[Helen]

The only thing I can say from my short (ie 1 week!) guiding experience is to watch for rotation caused by loose screws in eyepiece holders and guide rings, loose diagonal/barlow/CCD nosepiece barrels, and cable friction. But I did manage to guide a C8 at f10 with a ZS66 for 4 minutes

Hope you get it sorted TJ!

Helen

[Tim]

Martin, I have found the log files for the tracking, and have included them in a zip file with some pics showing the errors at various aggression settings.

I'd really appreciate you (and anyone )taking a look and seeing if there is anything obvious I am missing, the figures mean very little to me at the moment.

Here's the link;

http://ukastronomers.com/Tim/action/download_files/fileid/68

Again, thank you for all the kind assistance folks!!

TJ

[MartinB]

I doubt it's a problem with the focal length of the guide scope TJ. Guide software now all does sub-pixel, centroid based guiding.

I'll have a look at the file Tim

[Tim]

Great, thanks Martin.

Heres a cut and paste of what PHD says about the OSC number. Mine was 0.25 - 0.29 last night as I recall.

In the lower-left of the graph, an "oscilliation index" is shown. This is the result of calculating (in the current window's worth of data), the odds that the current RA move is in the opposite direction as the last RA move. If you are too aggressive in your guiding and over-shooting the mark each time, this number will head towards 1.0. If you were perfect and not over- or under-shooting and your mount had no periodic error, the score would be 0.5. Perfect with periodic error and the score may be closer to 0.3. If this score gets very low (e.g., 0.1), you may want to increase the RA aggressiveness (and/or decrease the hysteresis). If it gets quite high (e.g., 0. 8 ), you may want to decrease the RA aggressiveness (and/or increase the hysteresis).

[Psychobilly]

Hi TJ,

Sorry to here you were havign problems lastnight...

I read this article by Craig Stark a few days ago.. probably wont tell you anything you didn't know already...

http://www.stark-labs.com/craig/articles/assets/AutoGuiding%20Craig%20Stark.pdf

Hope that betwwen you and MartinB you can get to the bottom of wahts going on and you can start some long FL imaging.

Billy...

[MartinB]

Some great stuff on Craig Stark's sit Billy isn't there?

Tim, I've had a look at the log and your guiding is obviously working. When set to 0.1 aggression the Y axis wanders off by 20 pixels (assuming the figures relate to pixels rather than arc secs) over the guiding run so I guess this must be the dec and it's measuring your drift. On the 0.9 and 0.8 runs your y axis is the most erratic but it is containing the gudie star mainly within a 1 pixel range.

Back of fag packet calculations gives your guiding sampling rate at approx 2.5 arc secs/ pixel and sampling rate of imaging camera 0.6 arc secs per pixel. So 1 pixel of drift with the guider equates to 4 pixels on your final image. Although the guiding is more erratic in Y than X there are enough big peaks in X to give you a reasonable aspect ratio - so round, slightly bloated stars (inevitable at that focal length with our seeing).

That's a long winded way of saying your guiding is ok. It is correcting PE and polar misalignment drift. I don't know how long your guide exposures were but if they are 1 second you are getting about 1 arc minute of drift over 5 minutes. That won't cause the problems seen here.

Time to compare OAG with guidescope!

[Kaptain Klevtsov]

Just so that I'm clear with what you are describing TJ. Are you saying that the guide star stays put exactly, but the imaging camera gets trailing stars? If thats the case, then you need to look at the trail length on each sub. If its mirror flop, then I'd not expect to see them all the same, whereas an alignment issue would have the same trailing on every sub.

As everybody else has already said, its either something floppy or droopy, unless your polar alignment was pants.

Hope you get it sorted soon.

Kaptain Klevtsov

[Tim]

The annoying thing is, before I altered the mount to more accurately polar align it, I was able to get 300sec subs with very little trace of trailing. At 360secs the trailing appeared.

KK, the image seen on PHD's live view does not change, the stars do not move, and they do not appear to rotate. The FOV of the DSI in the WO72 is quite similar to the dslr on the c9.25.

When I lessened the aggressiveness of the guiding, down to 50 percent say, then PHD screen did show the star start to move a little down the screen over a half hour or so. To make the stars move in the same direction as the drift I would have had to press the DOWN button on the handset. With agressiveness set to normal levels the star did not drift noticeably.

Now then, ahem. I have just been out when I got home from work, and wiggled everything. The rotation bit on the WO72 was a bit loose, and did wiggle a bit to be honest. For a completely thorough test of this setup i'll need to re-do the run with that defect catered for. I am also going to route the cables differently so that there is no chance that they might be pulling etc on the cameras.

From those figures then Martin, what areas of the setup could i look at to tweak the tracking a bit?

I may retry the OAG. The biggest problem for me here is that at F10 under my skies it is rare to get many stars bright enough to guide on that you can actually see. But now I have a more sensitive camera, and before I was trying to manually guide with a crosshair EP. If I could get this to work it would be absolutely ideal for what I really want to do, which is a dual shot for each object, one a widefield view with the WO72, showing the objects place in the sky, and then the object itself.

One other thing I cant get my head round, the exposure times for PHD. If I set that at say 2 secs, does that mean that the program will only send a guide pulse every 2secs? I get confused because I thought I read somewhere that the guiding is done on a different a sub pixel level and doesnt depend on the exposure times???

If the exposure time you select does affect the guiding, then this may be another factor. To get the calibration over and done with I usually select 0.2secs as the exposure time, but once it is guiding, I use 2 secs. As I am typing this I have the gut feeling that this might be a very silly thing to do?

There is a settings in setup though that you determine the calibration step in millisecs to be used for calibration. And this determines how long a guide pulse should be used during calibration. This is reduced for long FL scopes and increased for short FL scopes. I have it set at 1300ms for the little WO72. Im not sure if this setting simply pushes the mount in a direction for 1300ms.

I think I will re-read the PHD gumph and try a few settings inside in the warm to see what it does.

Again, thanks all, really appreciate the help.

[Tim]

I think the answer might possibly be staring me in the face. All along I have been convinced that it was something in PHD that I was doing wrong.

I think I may be confusing it. By doing what I described above, ie, giving it 0.2 secs exposures for calibrating, then altering it to 2.0 secs for guiding, am I in effect making it impossible for PHD to do its job?

Or does it not matter what you initially select as the step, and PHD automatically does the maths and adjusts accordingly. (It is called PUSH HERE DUMMY after all!) I would have thought that if changing exposure length during a session was mission critical, then at least a warning might have been made?

All this thinking makes my 'ead 'urt!

EDIT: I've re-read (twice) the phd instructions. They do say to choose exposure at the start of the session before calibrating. They dont say NOT to alter it. Any ideas ?????

[Psychobilly]

Now thats interesting because i often adjust the exposure during an evenings session. I normally only calibrate once as well. I have read thats its worth recalibrating whenever you move to a new target. and to use sufficently long exposures to allow the "seeing" to average out.

Billy...