drift polar alignment procedure Accurate Polar Alignment

[MarsG76]

Hi All,

Now that my observatory is finished, I wanted to do a as accurate polar alignment as I humanly could do for my Astro Imaging and so started a adventure.

I ended up getting a decent polar alignment after a few tries using different methods and wanted to share my results for anyone out there who needs help in polar aligning their EQ mount with the simplest instructions I can put to words.

I want to start by saying  that unfortunately it is time consuming and there is NO SHORT CUT if you want accurate PA. The more time you spend on it the more accurate your scope will be aligned to the pole.

The below procedure worked for me in the southern hemisphere, but it should work just as well, except change the celestial equator from north to south for the azimuth adjustment.

I got a copy of AlignMaster and with the original file, I was very limited to the stars that I was able to use for the alignment, this was due to my house, the observatory hut. Another problem was that the alignment using AlignMaster never coincided with the alignment when I used the Polar scope and/or Celestrons "All Star Polar Alignment feature" (ASPA), not even close. Usually out by a factor of 6-7 DEGREES when checking using celestron polar alignment feature to measure the polar alignment (procedure described below). 

I tried to run the procedure a number of times, again with limited star selection and the polar alignment compared to ASPA was still way off. I determined that perhaps there are not enough southern stars in the sterne.txt list so I added more stars, carefully made sure that they were J2000 down to arc second detail, all southern biased, all below 15 degrees DEC, and yes I had more stars to chose from in Alignmaster but the result was the same as before, 6-7 degrees OFF!! 

At this point I spent 2 nights playing with Alignmaster, not getting any closer to my dreamt off perfect polar alignment so I determined that either I'm using it wrong (honestly I doubt it since it is very straight forward) or, most likely, Alignmaster is not as accurate as I thought... either way, I decided that the next few nights I'll revert to drift alignment.

I spent the next three night drift aligning the mount using the below described procedure and at the end of it I allowed the mount to drift on a eastern star and than a northern equator star for 30 minutes and both instances the star was within the illuminated line marking in the eyepiece, no drift, so I dare to say that I'm very close to target. 

 

ASPA Procedure:

1. Roughly align the mount on the celestial pole using the polar scope.
2. Turn on the mount and do a 2-star alignment followed by 4 calibration stars, preferably with a illuminated reticule eyepiece.
3. Press ALIGN >> POLAR ALIGNMENT >> Display Align, this will display the true error of which the scope is out to the pole.
4. To reduce the error, do an ASPA routine and then turn the mount off and repeat from step 2.

NOTE: Steps 2 and 3 have to be performed to get a accurate reading of the error. When I do a ASPA the error os display as 0 00' 00" in both axis, obviously that not right and only after the star alignment in step 2 can the computer in the mount workout the true error with respect to the set position of the ALT and DEC on the mount.

 

DRIFT ALIGNMENT PROCEDURE:

1. Do a rough polar alignment using either ASPA or a polar scope.
2. Align the illuminated reticule so that a star follows along the left/right line, referred to as the RA line when slewing the scope in RA. Find a star and center it in the eyepiece.
3. Facing east choose a star low in the eastern horizon, 30 degrees is a good latitude since it is low and not too much effected by the atmosphere. Allow the mount to track unguided for the chosen amount of time and adjust Altitude (up/down) to bring the drifted star back to the RA line. At first you can overshoot the adjustment until the star starts to drift in the opposite direction or there is no drift.
4. Facing north use a star as close to the celestial equator as possible. Again allow the mount to track unguided for the chosen amount of time and adjust the Azimuth (Left/right) to follow star direction drift back to the RA line.

NOTE: The star will move left/right along the RA line when the Azimuth is adjusted and it will need to be chased using the RA buttons on the hand controller as the star moves off the FOV, but as the azimuth knobs are turned the star will move closer to the RA line.

DO NOT TOUCH THE DEC BUTTONS SINCE THAT WILL CHANGE THE DRIFT AMOUNT OF THE STAR AND THE NORTH STAR DRIFT WILL HAVE TO BE PERFORMED AGAIN.

5. Repeat steps 2 and 3 with increasing the tracking/drifting time until satisfied with the drift amount. Drifting time will need to be increased to see the drift direction as the polar alignment gets closer to the SCP. Starting with drift time of 5-10 minutes is enough to show drift since the star will drift faster when the mount alignment is further from the pole

Once there is no or little drift in 30 minutes than that is very close to PA, MUCH less than an arc minute and good enough for very long guided exposures with no field rotation.

Generally I start with 10 minute drift, adjust and do 20 minute drifts than as a final check do a 30 minute drift alignment a few times until ideally the star doesn't move from the RA line at all.

Make sure that the drifting time is the same between adjustments, this will allow you to determine if the drift is getting larger or smaller by the distance that the star drifts.

I alternate adjustments between east and north (Altitude and Azimuth) alignment as one axis adjustment can slightly alter the other, especially at the start of the drift procedure.

Drift alignment is time consuming, for example, for my permanent setup I spent 3 nights drifting to get PA as perfect as I can, ideally i would like to be adjusting the mount until there is no drift for 60 minutes (the 60 minute drift alignment I will do next 2 nights. I figure that while the moon is out it is the perfect time to get PA perfected, so surely 5 night spent on it will be near zero error).

Of course the more time that is spent will result in a more accurate alignment but spending this much time on a portable setup is impracticable, so drift aligning can be shortened to accommodate the imagers needs, even a basic ASPA and a 10 minute drifting session a couple of times over will allow 10-15 minute subs when guided, my plan here was to have subs of over 40 minutes, even a hour if need be.

AGAIN, this drift alignment procedure was in the SOUTHERN hemisphere, so some aspects might need to be adjusted for the northern hemisphere, such as you can use Polaris, no such option down-under and your celestial equator star will be in the SOUTH.

ALSO, If I was drifting on a western horizon star, than, instead of moving the star back to the RA line in the eyepiece, I would have to move it FURTHER away than re-center the star using the hand controller. This is where keeping the drift amount times the same to determine if your drift is improving or getting worse comes in handy, if your star drifts more after a adjustment than you know that you need to turn the AZ/ALT knob the opposite way, SO using a star in the east is simply less troublesome.

I hope this info will be useful to someone who need help with polar alignment.

 

Next project... to program the CGEM with PEC using a CCD and PHD to record a few runs of the mounts error and upload the averaged recorded waveform.

 

Clear skies.

 

[LightBucket]

Get a polemaster, It will take under 10 mins...  simples

[MarsG76]

1 minute ago, LightBucket said:

Get a polemaster, It will take under 10 mins...  simples

how accurate?

 

[LightBucket]

1 minute ago, MarsG76 said:

how accurate?

 

Extremely perfect for long exposures with no guiding..

[MarsG76]

Just now, LightBucket said:

Extremely perfect for long exposures with no guiding..

hmm.... worth a look by the sounds of it...

 

[LightBucket]

It is worth a look, but it’s the price that puts people off, as if you are going to have a permanent set up then it will only get used Maybe once, perfect though if you have to set up each session...

they are about £280 over here, but probably cheaper for you....down under

[newbie alert]

 

Polemaster is great for speed but not as accurate as drifting..

I set up and tear down as I'm not on a pier and sometime venture out so I can't spend hours and hours drifting so I PA with polemaster first..then load  up sharpcap  and align with that....you will need to  slightly adjust..if polemaster was 100% You wouldn't need to adjust  anything.. In phd not that I've got it working as it should yet but my PA errors are low..when I've run the guiding assistant it's given me readings of 0.02 in RA..0.03 in DEC 

 

[Demonperformer]

Maybe it's my fingers, or the mount (NEQ6), but I always find the limitation with polemaster is the bolts that adjust the azimuth. Any software that can putperform the mechanical bits (again, either me or the mount) is good enough as far as I am concerned.

[BigRD]

Plus 1 for Sharpcap and its Polar align tool. Aligned in under 10 mins with excellent results every time........and it's £10 or free with new Altair Astro cameras.

[LightBucket]

16 minutes ago, BigRD said:

Plus 1 for Sharpcap and its Polar align tool. Aligned in under 10 mins with excellent results every time........and it's £10 or free with new Altair Astro cameras.

Surely that is only as good as how well your scope is aligned with the mount..?? With polemaster you are aligning the mount and not the scope...!  Depending on scope tube ring and dovetail alignment, or in the case of using a finderscope, how well that can be aligned with the mount with the fiddly 6 adjustment screws...

[MarsG76]

3 hours ago, LightBucket said:

It is worth a look, but it’s the price that puts people off, as if you are going to have a permanent set up then it will only get used Maybe once, perfect though if you have to set up each session...

they are about £280 over here, but probably cheaper for you....down under

I was thinking that too... sure I'm spending a long time getting zeroed in on the pole, but once it's done that's it... perhaps check it every year to make sure nothing changed... 

[MarsG76]

2 hours ago, newbie alert said:

 

Polemaster is great for speed but not as accurate as drifting..

I set up and tear down as I'm not on a pier and sometime venture out so I can't spend hours and hours drifting so I PA with polemaster first..then load  up sharpcap  and align with that....you will need to  slightly adjust..if polemaster was 100% You wouldn't need to adjust  anything.. In phd not that I've got it working as it should yet but my PA errors are low..when I've run the guiding assistant it's given me readings of 0.02 in RA..0.03 in DEC 

 

I thought that 10 minutes to perfect was too good to be true... but autoguiding does help a lot... I'm already a lot closer to perfect than I was when setting up nightly and even than I could get round stars in 30minute subs.

[MarsG76]

2 hours ago, Demonperformer said:

Maybe it's my fingers, or the mount (NEQ6), but I always find the limitation with polemaster is the bolts that adjust the azimuth. Any software that can putperform the mechanical bits (again, either me or the mount) is good enough as far as I am concerned.

I had a similar issue with the CGEM. Fine adjustments were not possible since the tension built up on the friction between the eq head  and the tripod than jumped a big amount, fine tuning was impossible on the azimuth, altitude is fine.

i got some of the stick on teflon tape that is used for people to easily slide furniture around with out scratching the floor... I got sizable pieces that were adhesive on one side and cut out 2 semi circle shapes on which the CGEM sits, not the azimuth adjustment is super smooth and fine adjustment are easy, literally 3 fingers used to turn the head with great precision.

[newbie alert]

Yes I do the same regarding the Teflon sheet..i have 3 layers in between... I read about it and have used ever since..

I love the polemaster but find it's not a subtle adjustment like sharpcap is..

[MarsG76]

I think that once I'm setup to accuracy where I can turn off DEC guiding and still get up to 45-60 minute subs with round stars than I'll be over the moon. Won't need any polar alignment tools other than ASPA and drifting.

Checking PA once every year and re tweaking if need be is no issue.

i think that pole master, align master, sharp cap etc are good for quick setting up when the unit is portable but for a permanent setup,  nothing beats spending the time to get very close to zero error.

In theory, if there is any error in PA where autoguiding has to correct in both axes than as the PA offset increases than motion blur like effects are introduced into the subs.

For short subs like 10 or 15 minutes it won't matter much if the error is greater but for 30 min and longer, it does start to soften the subs.

[newbie alert]

I find the ASPA accurate enough for visual..but way out on imaging, if it was good enough for imaging then I wouldn't of bought a polemaster for £250...The handset might tell you where it thinks it is but in reality I think it's way off.. since using the method  I use now (polemaster/sharpcap) I can get double the exposure length than what I could with the ASPA.. all unguided 

[newbie alert]

14 hours ago, MarsG76 said:

I thought that 10 minutes to perfect was too good to be true... but autoguiding does help a lot... I'm already a lot closer to perfect than I was when setting up nightly and even than I could get round stars in 30minute subs.

Sub 10 mins of arc! I'm looking for under a min, even in bad transparency.. 

[ollypenrice]

18 hours ago, LightBucket said:

Extremely perfect for long exposures with no guiding..

Long exposures with no guiding are surely only possible using a mount with an accurate sky model and absolute encoders allowing it to 'guide' on itself. At lower elevations a mount which maintains 100% perfect sidereal rate will show tracking errors because the apparent position of a star varies with atmospheric diffraction as the elevation changes. This is why the 'unguided' mounts like 10 Micron and ASA need a sky model which allows them to correct for this effect. In truth they are guiding on their model and on their own encoders. One of SGL's 10M users reports taking hour-long subs without issue.

I've never owned a Polemaster but I've tested the accuracy of the alignment of Polemaster-aligned mounts twice. While the PA in both cases was not bad it certainly wasn't perfect. The easiest way to test it is to take an early and late sub exposure from the same continuous run (without flip or dithering) and align the two images on the star pattern. You can then see whether the frames of each image are still aligned or whether they are rotated. In the cases of the data sets I looked at there was significant rotation after a few hours. For all that, long guided subs were fine and there could have been user error in the Polemaster alignments.

For a permanent setup I vote for drift alignment because it is a direct measurement, software and gadget-free!

Olly

[fireballxl5]

16 minutes ago, ollypenrice said:

The easiest way to test it is to take an early and late sub exposure from the same continuous run (without flip or dithering) and align the two images on the star pattern. You can then see whether the frames of each image are still aligned or whether they are rotated. In the cases of the data sets I looked at there was significant rotation after a few hours. For all that, long guided subs were fine and there could have been user error in the Polemaster alignments.

For a permanent setup I vote for drift alignment because it is a direct measurement, software and gadget-free!

Olly

Good point Olly. I have see some drift over 3 hours but the 10 minute subs are fine.

Using 2 subs as you suggest, is there a way to measure the drift?

Regards, Andy

[ollypenrice]

Just now, fireballxl5 said:

Good point Olly. I have see some drift over 3 hours but the 10 minute subs are fine.

Using 2 subs as you suggest, is there a way to measure the drift?

Regards, Andy

The test I proposed above gives a quick visual indication of the rotation between first and last sub. If you're trying to quantify the rotation then I'd suggest the following, which we did try on one of the datasets I mentioned:

Stack the first and last sub without aligning them. (Meaning the images will be aligned on their frames rather than their stars, if you like.) If the stars are nicely aligned then you have perfect PA. In reality you'll probably find that the stars in the middle are aligned on top of each other but that, out in the corners, you'll get dual star images possibly not overlapping at all. That's what we found. You can then zoom in to pixel scale and count the number of pixels between the centres of the two stellar images. Say you find a drift of 10 pixels over 5 hours. You therefore know that your corners are smearing at the rate of two pixels per hour, or one pixel per half hour.

I don't really know what one would define as acceptable but instinct tells me that half a pixel would be as much as I'd care to tolerate, so that rig would be limited to 15 minute subs. I stress that this value is plucked out of the air!

Olly

[LightBucket]

1 hour ago, ollypenrice said:

Long exposures with no guiding are surely only possible using a mount with an accurate sky model and absolute encoders allowing it to 'guide' on itself. At lower elevations a mount which maintains 100% perfect sidereal rate will show tracking errors because the apparent position of a star varies with atmospheric diffraction as the elevation changes. This is why the 'unguided' mounts like 10 Micron and ASA need a sky model which allows them to correct for this effect. In truth they are guiding on their model and on their own encoders. One of SGL's 10M users reports taking hour-long subs without issue.

I've never owned a Polemaster but I've tested the accuracy of the alignment of Polemaster-aligned mounts twice. While the PA in both cases was not bad it certainly wasn't perfect. The easiest way to test it is to take an early and late sub exposure from the same continuous run (without flip or dithering) and align the two images on the star pattern. You can then see whether the frames of each image are still aligned or whether they are rotated. In the cases of the data sets I looked at there was significant rotation after a few hours. For all that, long guided subs were fine and there could have been user error in the Polemaster alignments.

For a permanent setup I vote for drift alignment because it is a direct measurement, software and gadget-free!

Olly

Points taken, perfect was probably the wrong word, maybe “very good”..

[MarsG76]

2 hours ago, ollypenrice said:

Long exposures with no guiding are surely only possible using a mount with an accurate sky model and absolute encoders allowing it to 'guide' on itself. At lower elevations a mount which maintains 100% perfect sidereal rate will show tracking errors because the apparent position of a star varies with atmospheric diffraction as the elevation changes. This is why the 'unguided' mounts like 10 Micron and ASA need a sky model which allows them to correct for this effect. In truth they are guiding on their model and on their own encoders. One of SGL's 10M users reports taking hour-long subs without issue.

I've never owned a Polemaster but I've tested the accuracy of the alignment of Polemaster-aligned mounts twice. While the PA in both cases was not bad it certainly wasn't perfect. The easiest way to test it is to take an early and late sub exposure from the same continuous run (without flip or dithering) and align the two images on the star pattern. You can then see whether the frames of each image are still aligned or whether they are rotated. In the cases of the data sets I looked at there was significant rotation after a few hours. For all that, long guided subs were fine and there could have been user error in the Polemaster alignments.

For a permanent setup I vote for drift alignment because it is a direct measurement, software and gadget-free!

Olly

Good point, a confirmation of my assumption that polemaster wouldn't be perfect and that drift aligning is best for a permanent setup.

Another thing to is that unguided long exposure using a CGEM, or CGEM class mount is just not possible, the PE would kill a sub within 5 mins, but autoguiding on it when it is very near perfectly polar aligned, it is capable of very accurate guiding and HQ subs even at 2000mm FL.

[ollypenrice]

3 hours ago, MarsG76 said:

Good point, a confirmation of my assumption that polemaster wouldn't be perfect and that drift aligning is best for a permanent setup.

Another thing to is that unguided long exposure using a CGEM, or CGEM class mount is just not possible, the PE would kill a sub within 5 mins, but autoguiding on it when it is very near perfectly polar aligned, it is capable of very accurate guiding and HQ subs even at 2000mm FL.

Given that you might want to keep the RMS down to half an arcsecond there is no purely mechanical mount of which I'm aware which can deliver this. The options are encoder-guiding or guiding on a star. To be fair to the Skywatcher and Celestron mounts, they might have a PE of 30 arcseonds but this doesn't stop them from guiding out very well indeed.

Also worth noting in this thread, maybe, is that perfect PA may not be the way to optimal tracking on many mounts. Backlash in Dec is a common problem which, unlike backlash in RA, cannot simply be cured by running slightly out of balance.* Autoguiders sometimes cause the mount to oscillate across its Dec backlash and the classic fix is to run slightly polar misaligned and to disable the Dec autogiude corrections in the direction not needed to correct for this misalignment.

Olly

*An imbalance fore or aft along the Dec axis reduces to zero as the tube points to the zenith. You can use moveable weights (eg magnetic) which you relocate during the night but it sounds like a palaver.

[MarsG76]

1 hour ago, ollypenrice said:

Given that you might want to keep the RMS down to half an arcsecond there is no purely mechanical mount of which I'm aware which can deliver this. The options are encoder-guiding or guiding on a star. To be fair to the Skywatcher and Celestron mounts, they might have a PE of 30 arcseonds but this doesn't stop them from guiding out very well indeed.

Also worth noting in this thread, maybe, is that perfect PA may not be the way to optimal tracking on many mounts. Backlash in Dec is a common problem which, unlike backlash in RA, cannot simply be cured by running slightly out of balance.* Autoguiders sometimes cause the mount to oscillate across its Dec backlash and the classic fix is to run slightly polar misaligned and to disable the Dec autogiude corrections in the direction not needed to correct for this misalignment.

Olly

*An imbalance fore or aft along the Dec axis reduces to zero as the tube points to the zenith. You can use moveable weights (eg magnetic) which you relocate during the night but it sounds like a palaver.

I read the slight misalignment foe DEC to only be guided in one direction. It definitely makes sense, and that is my plan B if I can't get my PA aligned to such a level where I can completely turn DEC autoguiding OFF and only guide the RA axis.

On another point, when misaligned than on one side of the meridian DEC drift would only be in one direction no matter how the eq mount is misaligned.

[ollypenrice]

Getting to turn off the Dec guiding entirely is a nice idea and sounds OK in theory, but I don't know anyone who has ever managed it, and you need to consider atmospheric diffraction as I mentioned in an earlier post.

If running misaligned and guiding in one direction only, you do need to reverse the guide axis in Dec after the flip.

Olly