Going Crazy!!! Drift Alignment

[Tamtarn]

As discovered on other threads others have had problems with drift alignment.

We have MAJOR porblems.

We have a Skywatcher 8' Newt on a EQ5 mount.

We've tried for the past three nights doing the RA drift. The star is moving north in the eyepiece and has been doing so for the three nights and we keep adjusting but still no luck.

We centre the star in the EP and after about 10 mins it drifts north. According to the instructions this means the polar axis is pointing too far east so we move the star west in the EP with the azimuth knob and centre the star again with the hand controller. The star stays in the centre about 10 mins by 20 mins it moves north again. We repeat the procedure again. OK for 10 mins another 10 mins it moves again, one azimuth knob is screwed out 9mm the other 5mm.

Do we keep turning ?????????? How far can we go ????????? Or are we doing something wrong ?????????? Do we have trouble with the mount or motor drives ???????????

HELP PLEASE :help: :help: :help:

We're going crazy :screwy: :screwy: :screwy: :screwy: :screwy:

[DavidH]

Yes, drift alignment = craziness.

I spent quite a few nights trying to get it right. I found that I could adjust the azimuth knobs until I almost had it right and then I would run out of travel on the azi adjuster.:mad2:

So my technique now is that when adjusting azimuth, I move the whole mount (tripod and all) in small increments (say 5 degrees or so) around to the east or west until the stars reverse their drift direction. That way I know that the correct polar alignment lays between the last two positions, and any adjustment will be within the travel of the azimuth adjuster.:D

If, however, you are experiencing variable declination drift, as opposed to constant drift, I would be looking at the stability of the polar axis, as this must be shifting somehow.:screwy:

HTH,
David.

[sheeny]

G'Day Barb and David,

My guess is you are probably doing the right thing, just not enough of it...

Drift aligning can be very frustrating to learn, because most people worry about adjusting too much. I know I did. But the key is to start by making big adjustments, then when the star drifts the opposite way, you know you have gone too far... but you also know that proper alignment is in between your last two adjustment points!

Then you can start making smaller adjustments! I usually try to apply something like a binary search method to achieve alignment quickly.

I'll describe how big my initial adjustments are to give you some idea. I usually align with a star diagonal and a 10mm illuminated reticle EP, so for my scope I use the Right Hand Rule: if the star drifts in the direction of my right thumb, I adjust the azimuth to push the star in the direction of my right fingers. Since you are using a newt, you would apply the Left Hand Rule (star drifts in direction of LH thumb, adjust in direction of LH fingers).

For my initial adjustments, I adjust the azimuth till the star is nearly out of the FOV, then bring it back onto the Dec cross hair with the controller, adjust again, and bring it back onto the Dec cross hair... and keep repeating this until the star is on the Dec cross hair at the edge of the FOV. Then I use the controller to centre the star in the middle of the cross hairs again and check for drift.

Once I go too far (i.e. star starts to drift the other way), I adjust the other way by half this amount - i.e. till star is 1/2 way between centre and FOV of the Dec cross hair. Try to make each adjustment from there on half as big as the previous one.

My guess is though, that since you reckon your tracking is good for 10 minutes, but not for 20, then you should be pretty close. Maybe try an adjustment 2 or 3 time bigger than you've been doing just to prove you are as close as you think. If it doesn't drift the opposite way, you just need to go further.

Another thought too... since you appear to be close, have you gone back to check your latitude adjustment? It might be worth doing that at this stage, before you try for final azimuth alignment. It might be out enough to affect your alignment in az.

Hope this helps!

Al.

[Astroman]

10 minutes without movement hell I'd be happy with 2 minutes. Looks like you are very much in the ball park. EQ mounts or any mount for that matter will not track perfectly, there WILL be movement no matter the cost or mount you use. The only true way of keeping an object in field for LONG periods is to guide.

[JohnG]

Are you checking your elevation immediately after checking you azimuth? You should be doing both together? making smaller and smaller adjustments.

What is said above is correct, make large adjustments till the star is going the other way then correct back.

If both your azimuth and elevation are correct, the only drift you will get is in RA, this is normal, it is the Perriodic Error in your worm.

10 minutes without movement is quite reasonable.

Cheers

JohnG

[matt]

Yep. What the other guys have said.

Also, it's been posted many times before and no doubt will be posted many times more, but this is pretty much the best drift align sim you can find: http://www.petesastrophotography.com/tutorial/guidingsim.html

Dead easy, and very realistic to what you'll see when you're out at the (crosshair) eyepiece:)

People can explain the process over and over again, but there's nothing like actually "doing" it, to understand how it all works

[Tamtarn]

Thanks guys a combination of all your input has been successful :thumbsup::thumbsup::thumbsup:

We have just managed to keep the star central for an hour in a 12.5 mm EP so we're very happy :D:D:D

[spearo]

well done
such a frustrating experience especially at first. But it becomes so much less daunting once you've done it a few times
frank

[sheeny]

Cool!:thumbsup::thumbsup::thumbsup:


Al.

[Tamtarn]

Frank & AL thanks for your encouragement

Even though we can hold a star central in the 12.5mm eyepiece for an hour our max. exposure time seems to be 90sec for a reasonable shot.

Have just purchased a 12 mm illuminated reticle eyepiece with matrix grid, so we can see more exactly what is happening with any movement and hopefully correct it and improve the length of time we can shoot.

David

[sheeny]

G'Day David,

The best way I know to troubleshoot those sort of things is to use the drift explorer in K3ccdtools... I notice you don't list a webcam in your signature, so I suppose you don't have one, but if you could lay your hands on a webcam and K3, the graphs in the drift explorer will tell you a) how good your drift alignment really is (very quickly) and b) whether you have any periodic error or tracking rate errors. Just a suggestion;).

If want to give that a go, and haven't used K3 before, we (myself, 1ponders, and probably some others) can try to help to steer you through the drift explorer by remote control... it might not be as easy as showing you in person, but we could have a go!

Al.

[JohnH]

David, Barb,

Ah so it's astrophotography that you want is it? Hmmm welcome to the dark side brouhahhahahahha :P !!!!

If you can get 90 sec unguided shots with an 8" Newt on an EQ5 with no trails reliably that aint bad at all, in fact I would say this is good!

You do not say what camera you are attempting to use but I assume you are at prime at F5 or so and using a DSLR so the trails will be noticeable if tracking is off by >2" over the course of your shot.

If your polar alignment is good (and it sounds like it is) then the remaining error will mostly be due to the periodic error (PE) in your RA drive - this can be as much as +/- 30" over the period of the worm gear (typically 8 mins) and may not be even thus you can get jumps and/or flats spot that ruin even a short exposure shot. This depends on the mount build quality and condition.

You can verify if the trails are PE by observing the orientation of the trails you see in your image with the scope axes. DEC drift is mostly alignment error, RA is a combination of alignment and PE and normally dominated by the latter.

To see why I think 90s is a good reult lets do a horibbly approximate sum:
The average error in your mount will be:

60" (peak to peak error) / 480s (worm cycle) = 1/8" per sec

If my numbers are about right then the error is 90/8" or approx 10" for your 90s shot. This is worst case as the RA error will vary as the COS of DEC of the target so this assumes a target on or close to the ecliptic but you get the idea.

For me working at f9 on a Vixen Sphinx I cannot shoot unguided reliably much above 30s however 30s at iso 1600 with a DSLR gives you plenty to work with before you get into more $$$ and the fun that is autoguiding...

[Tamtarn]

Thanks John

I guess we could be asking a bit more than our equipment is capable of doing as you suggest.

We haven't had a chance to try with the reticle as yet, maybe we can creep up a few more secs :) If we can bump up to 2mins we'll be extremely happy, if not we'll be happy with 90 sec.

We're using a Pentax ist DL. - Prime Focus at F5

Hey !! Astro imaging is really frustrating :confuse3:

[Tamtarn]

Hi Al

Thanks for your offer of help with Drift Explorer.
We have don't have a webcam or K3CCD tools, might take you up on your offer at a later date though as it sounds like a good idea. :thumbsup:


David

[Scorpius]

An interesting topic .. I have an Illuminated Reticle EP coming some time next week. These notes will come in useful when I venture into the world of drift alignment.

An interesting comment Roger G mentioned to me was: Do not rely on the elevation scale on your EQ mount for setting your latitude position, Its only a disk held on with double sided tape and can be several degrees out. I use a builders inclinometer, available at any good hardware store. I originally got it for setting up satellite dishes.

Thought of the day
What would we do without each other :)

[g__day]

And a hint from a newcommer to astrophotography and auto-guiding, the better your alignment, the less frequent and over-bearing your autoguiding commands will be.

It should be obivious that the better you're aligned the less work your auto-guider has to perform. Ideally if perfectly aligned you'd only be correcting for PE manifesting in your RA. Now auto-guiding can under or over correct, so its not a perfect solution. The less you have to do, and the less frequently you have to correct the better your longer shots will look by far.

If your auto-guider is correcting every 1/2 second - don't expect 200 second shots to look fabulous. If you only need corrections every 3 seconds (you're probably less than a 20 arc minutes off polar alignment) so your shots will look much better. If you only need corrections > 5 secs - you've probably set things up pretty well (assuming short corrections triggered by a 1/5 of a pixel movement) and you can guide off much fainter stars if corrections are needed less frequently.

So as precise as possible polar alignment and focusing are - ultra critical - keep at it and you will enjoy this immensely!

[Tamtarn]

Thanks for the tip Dave, re the elevation scale on the EQ mount and the info on the builders inclinometer.

David

[sculptor]

Hi,

I finally gave up on drift alignment and built an autoguider. But a few things that helped before that fateful day:

Align once and for all: get out the lathe and make fittings for a half-inch aluminium rod (nonmagnetic) that fits where the polar scope on the CG5 ought to go. (Make sure it doesn't get crunched by the hole in the declination axis). The fitting will also fit your EQ6 when you buy it!!! Put a spirit level and a precision ($250.00) compass about half a metre down the rod. Ignore the compass and level settings for the time being. Do your alignment. Dedicate a night to it. Now put the scope in some fixed position, eg meridian/equator on west of mount, so magnetic bits in standard relationship to compass. Park car well away! Now just record, really accurately, the compass bearing. Glue the spirit level on with an acrylic wedge and superglue. Now you can REPRODUCE your alignment within seconds! (remember that the actual absolute compass reading is meaningless, due compass error, magnetic variation, where you parked your car, etc). Also watch out - even the screws in the corners of your spectacles are enough to put the compass out by a degree. Really! No keys, watches, within a metre. No cars within 4 or 5 metres: Error goes with the cube of the distance. And always in same spot due south of scope, never east or west.

The plot thickens: When doing the alignment, get out the graph paper and plot the N/S drift rate and direction as a function of compass bearing, one measurement every 2 or 3 degrees of azimuth. Be accurate. You don't even have to have found the perfect alignment - just read where your graph crosses zero.

Mark your turf: If legal and safe, hammer three 25mm dia stainless (nonmagnetic!) steel tubes into the ground (eg just below ground level) exactly where the where you normally put your tripod. Scope goes back in same spot each time. Saves ages.

Photograph the drift: When measuring drift rate, use your camera and laptop. Expose for say 1 minute, and count pixels of drift. Plot pixels per minute vs compass bearing. Same deal for the spirit level. No need to wait 20 mins.

But after you've done all that, one day, an autoguider will somehow end up in your possession - now you are doing 3 hours of astronomy and 10 mins of setup instead of the other way round.

Regarding autoguiders - must do proportional control at tiny fraction of sidereal rate, talking to CG5 directly via hand controller, not the old style bang-bang on-off control via the stegosaurus-vintage "autoguider" port. Then, it is best to put the polar axis intentionally out of alignment by say 1 degree, so that the autoguider is always working in the same direction, no backlash.

Make the motor work for its cocoa: (b) carefully put the counterweights very slightly out, so that the drive motor is always working very very very slightly against the imbalance. No backlash. Absolute disaster, whether autoguiding or not, is to have the scope wanting to lurch west ahead of the worm-drive from time to time. Even perfect balance is suboptimal, because things catch and lurch, catch and lurch.

Agree with the other folks that you're already doing superbly with a CG5. I feel they're a great "visual" mount rather than a "photographic" mount, which requires more engineering if you want 10 minute exposures.

For those that can't do any of the above, there is nothing wrong with stacking lots of 1 minute exposures using RegiStax etc. The up-side is that you can throw out those images where the wind gusted, or a cloud came. The main down-side is that the read-out error (one-off noise on each exposure) is root(10) times worse than with one single 10-minute exposure. But read-out error is usually small compared with dark current, skyglow, and hot pixels, so not to worry.

Hope these tricks help a bit.

[sculptor]

Hi,

I finally gave up on drift alignment and built an autoguider. But a few things that helped before that fateful day:

Align once and for all: get out the lathe and make fittings for a half-inch aluminium rod (nonmagnetic) that fits where the polar scope on the CG5 ought to go. (Make sure it doesn't get crunched by the hole in the declination axis). The fitting will also fit your EQ6 when you buy it!!! Put a spirit level and a precision ($250.00) compass about half a metre down the rod. Ignore the compass and level settings for the time being. Do your alignment. Dedicate a night to it. Now put the scope in some fixed position, eg meridian/equator on west of mount, so magnetic bits in standard relationship to compass. Park car well away! Now just record, really accurately, the compass bearing. Glue the spirit level on with an acrylic wedge and superglue. Now you can REPRODUCE your alignment within seconds! (remember that the actual absolute compass reading is meaningless, due compass error, magnetic variation, where you parked your car, etc). Also watch out - even the screws in the corners of your spectacles are enough to put the compass out by a degree. Really! No keys, watches, within a metre. No cars within 4 or 5 metres: Error goes with the cube of the distance. And always in same spot due south of scope, never east or west.

The plot thickens: When doing the alignment, get out the graph paper and plot the N/S drift rate and direction as a function of compass bearing, one measurement every 2 or 3 degrees of azimuth. Be accurate. You don't even have to have found the perfect alignment - just read where your graph crosses zero.

Mark your turf: If legal and safe, hammer three 25mm dia stainless (nonmagnetic!) steel tubes into the ground (eg just below ground level) exactly where the where you normally put your tripod. Scope goes back in same spot each time. Saves ages.

Photograph the drift: When measuring drift rate, use your camera and laptop. Expose for say 1 minute, and count pixels of drift. Plot pixels per minute vs compass bearing. Same deal for the spirit level. No need to wait 20 mins.

But after you've done all that, one day, an autoguider will somehow end up in your possession - now you are doing 3 hours of astronomy and 10 mins of setup instead of the other way round.

Regarding autoguiders - must do proportional control at tiny fraction of sidereal rate, talking to CG5 directly via hand controller, not the old style bang-bang on-off control via the stegosaurus-vintage "autoguider" port. Then, it is best to put the polar axis intentionally out of alignment by say 1 degree, so that the autoguider is always working in the same direction, no backlash.

Make the motor work for its cocoa: (b) carefully put the counterweights very slightly out, so that the drive motor is always working very very very slightly against the imbalance. No backlash. Absolute disaster, whether autoguiding or not, is to have the scope wanting to lurch west ahead of the worm-drive from time to time. Even perfect balance is suboptimal, because things catch and lurch, catch and lurch.

Agree with the other folks that you're already doing superbly with a CG5. I feel they're a great "visual" mount rather than a "photographic" mount, which requires more engineering if you want 10 minute exposures.

For those that can't do any of the above, there is nothing wrong with stacking lots of 1 minute exposures using RegiStax etc. The up-side is that you can throw out those images where the wind gusted, or a cloud came. The main down-side is that the read-out error (one-off noise on each exposure) is root(10) times worse than with one single 10-minute exposure. But read-out error is usually small compared with dark current, skyglow, and hot pixels, so not to worry.

Hope these tricks help a bit.

[Scorpius]

Amazing... Last night I tried out my shiny new Orion Illuminated Reticle 12mm Plossl and did my very first Drift alignment at our Lunatics meeting. Roger G was there and he adjusted the elevation adjustment screws on the EQ6Pro whilst i called out the positions. We had Saturn in the cross hairs. after establishing a fairly stable situation I went to M42 and landed smack bang on the cross hairs of the finderscope. the eyepiece was a tad out. Left it for over half an hour and M42 had barely moved at all.

So migrated in the opposite direction to the moon. Dead center on target. To be honest this was the first time I have managed to get the EQ6 to do what I wanted.

On top of this I only did a single star alignment of Sirius and initially moved the whole tripod round to get within cooee and finally adjusted the polar trim knobs.

We had quite a heavy dew start forming so decided against getting the camera out. But a most satisfying result and evening. Yes a large spirit level cost $3 from the Big B and a $10 compass from Aust Geo shop

Thought of the day
If I can do it anyone should be able to do it.:whistle: