All Sky polar alignment kept switching off

cpl42

I attempted to use the All Sky PA function in v2.1 last night - my SCP was completely obscured by a large tree. What I found was that when I slewed to a target sky (a bit west of the meridian and 70deg altitude) and started PA, it immediately started the standard PA protocol. On checking the status of the All Sky setting, it had switched off. I tried switching on All Sky both before and after slewing to my target sky with exactly the same results. (It resulted in a slowly rotating field over the next 3 hours of imaging.)

Has anyone had a similar experience, or know what I am not doing right?

Thanks in advance,
Paul.

cpl42

I checked to see if ZWO has released an update to their ASiair manual, but no, it's still at version 1.2.

So, I decided to do some testing. I selected some stars not too far from the Meridian on the western side (Canopus, Sirius). Both failed. I then selected stars closer to the western horizon (Betelgeuse, Rigel) and in the northern hemisphere - this was on the assumption that being in the southern hemisphere, All Sky might decide to rotate eastwards, instead of westwards: both failed. I then chose several stars in the eastern sky starting near the eastern horizon (e.g. Spica, Alphad, Gacrux), but again all failed. It appears that All Sky just doesn't work in the southern hemisphere!!?

I noticed that the PA button in the v2.1 app is labelled "PA SCP" with the "SCP" in smaller lettering and below the "PA" label. And that is precisely what you get and nothing more. ZWO: what the heck is going on here? Please advise me if I'm doing something wrong!

w7ay

cpl42

Paul,

I don't use the ASIAIR "All Sky" plate solve. But I wonder if your problem is related to plate solving.

Make sure that you can plate solve (gain, exposure time, focus, etc) in the Preview mode before you start the polar alignment process. Try to point away from a bright star when you do this test. Find a region of the sky with lots of stars with about equal brightness. Make sure the brighter of the stars do not saturate -- this is why you want to avoid having a very bright star in the frame, since there is not enough dynamic range to see other stars and still keep the brightest star unsaturated.

For the "first star," try to position the OTA so that it does not point to a bright star. Try moving close to the Meridian towards the north to stay away from east and west (which would yield larger alignment errors).

So, in particular, make sure that you do not start at Beetle Juice (:-), Rigel, Spica, or any named stars since those are way too bright, and may cause problem plate solving. Just choose a random RA and declination. The ASIAIR is not looking for the position of a star, it is looking for the center of the photographic plate.

Now, do a test plate solve there to make sure it suceeds. If Plate Solve in Preview does not succeed, neither will it succeed during polar alignment

Next, go turn on the All Sky toggle switch before you enter the "PA" mode.

Only then, go to the "PA" Mode and push that button.

Chen

cpl42

Many, many thanks for your suggestions, Chen. I really appreciate it as I was pulling my hair out over this one.

Unfortunately, we're due to have a week of cloudy weather, so I can't check this out for a while, yet. As soon as I can, though, I'll out there, and I'll let you know. It's a bit curious, but the app doesn't seem to attempt a plate solve at all as it just immediately switches to SCP PA and checking the status of All Sky shows it turned off. If it had flagged that dynamic range issue (assuming that this is the culprit) I'd have a lot more hair!

BTW. Does it really have to be a northern star field? I'm at 32 deg south and I have lots of southern sky to choose from.

Again, thanks buckets! Paul.

w7ay

cpl42 Unfortunately, we're due to have a week of cloudy weather, so I can't check this out for a while, yet

Hi Paul,

I haven't done a single session since Fall :-). There were a couple of days in WInter where it was clear, but too cold to be worth going outdoors. It is now the rainy season, and there won;t be any break until Spring.

As far as hair goes, I still have them, but mostly gray now :-).

Does it really have to be a northern star field?

Well, in some ZWO instructions, it says no direct east, and no direct west, and I thought that you have obstuctions at South (otherwise you would have used the regular polar alignment), so I just winged it and suggested north :-). But it does look like you'll do OK on the south too.

By the way, the futhest south I had lived (in my secondary school days in the 1960s) was in K.L. at 3ºN, and I still miss the beauty of the Southern Cross -- I can barely remember it nowadays by listening to Victory At Sea :-).

Chen

cpl42

Chen, I commiserate with you. I live on the north NSW coast in Australia. Our weather is rather variable throughout the year but even with a wet year just passing, I could usually expect 3-5 clear nights a months, or better. Dry years are even better than this, but with the semi-tropical, coastal area here one has to expect a good bit of variability, making it difficult for the weather forecasters, particularly for cloud cover. As case in point: tonight was predicted to be 10-30% cloud cover, but it rained at 17:00 and by 18:00 was still 90% cloud cover. I'll stop moaning as, at least, I don't have to face monsoon conditions!

I'm afraid that my hair reference was a bit of a dig at myself. I'm now around 85% bald, and that is definitely grey!

North or South: one YouTube video presenter on this topic used a southern star field whilst sited definitely well north of the equator, and had no problems. So, I expect that I will be OK but will try a number of different fields to test All Sky out.

I'll keep you posted as, and when, I can get a clear night.

Paul.

cpl42

Well, the weather gods smiled on me tonight and I was able to try out your recommendation, Chen, sadly to no avail.

My procedure was: a) from the Park position, use the manual slew controls to move to various fields; b) click the Preview button to check that I had a reasonable number of not-too-bright stars and no prominent one; c) click the Solve button to ensure the ASiair knew where it was - worked perfectly each time, solving at a steady 0.6sec; d) checked that the All Sky PA was turned on; e) click the PA button. On each occasion the PA routine selected the SCP method (telescope model showing dotted lines to a highlighted SCP, and instructions consistent with the normal polar alignment routine). Checking each time showed the All Sky option was still turned on. I didn't bother to proceed with the PA (been there, done that, waste of time - too far from the pole message).

I chose fields on either side of the Meridian, in both the Northern and Southern hemispheres. I also tried one field about half-way to the horizon on the South-East side. All performed exactly the same.

For the record, my mount was as Losmandy G11 running Gemini 1 software. Now, the Gemini is getting technically old and might just be an issue here, so I want to re-test your method with my new EQ6-R mount. However, please feel free to comment, as I would appreciate any suggestion.

Paul.

w7ay

cpl42

Paul, if you did everything right, the remaining thing is a bug in ASIAIR for users in the Southern Hemisphere (or at least close to NSW's latitude).

Are you sure the regular polar alignment does not work for you? The OTA does not have to point at the pole. The polar axis eventually has to, but the OTA can point up to 30º in declination away from the pole (minus the FOV of your OTA, which is probably a couple of degree at most). And the starting point does not require the OTA to be top dead center ("counterweight pointed to ground). It simply requires two spots at (RA, some declination less than 30º from pole) and at (RA-4 hours, same declination) where it can plate solve. The 4 hours is that 60º the polar alignment slews. Check a sky atlas like SkySafari (especially easy if you have already mapped your "horizon" for Sky Safari) and see if you can find two spots that meet the criteria. It does not require the sky to be visible anywhere else except for those two spots.

Take a look at post #11 (around there) where I had drawn a diagram to show that the pole itself need not be visible. It just needs to be visible at "Start" and "End" of Polar Align.

https://bbs.astronomy-imaging-camera.com/d/11580-polar-alignment-without-polaris/11

There can even be a tree or chimney in between "star" and "end" and it wouldn't matter.

Notice in the drawing that I had positioned Start to be 2 hours east of the Meridian, and end to be 2 hour west of the meridian, to maximize the amount the OTA is above the horizon/roof when it is making plate solves. But that start and end can be anywhere as long as declination is the same, and they are separated by 4 hours (60º).

The RA +/- 4 hours is for the Northern hemisphere. If the ASIAIR is rotation counterwise (with mount between you can the pole), then you will need to reverse the signs for the RA above.

When you do this 30º in declination from the pole, the accuracy is not as good. I forget it now, but I think my simulation runs showed something like 2 arc minutes off.

I think you had mentioned being 30ºS latitude. This lets you point the OTA a whopping 60º above the southern horizon, so there is a good chance that you can see some parts of the sky that you can plate solve.

Now, if the ASIAIR's bug is because it cannot slew your mount, the usual polar alignment scheme won't work either.

BTW, there is a known flaw in ASIAIR polar alignment -- it has a strict requirements that the slew rates of the mount is set to the factory default. If you had used the mount for visual and had set the rates to something else, you need to set them back.

Chen

cpl42

Many thanks for this detailed reply, Chen. I appreciate your patience with me and my problem. I now have a better understanding of the flexibility built into the PA routine in ASiair.

Firstly, I performed a standard SCP PA just 2 nights earlier with the same mount and OTA configuration, without having a problem. As this was the first time I had done a PA with this mount, I was a little slow, but I got there with <1' error in both axes. Guiding resulted in ₀.7" error (360mm guidescope) and gave very pleasing round stars over 600sec exposures. I conclude from this that the ASiair and G11-Gemini 1 were working together satisfactorily. (I got about the same guiding error with the same guidescope using my EQ6-R mount.)

I've checked my mount in my SCP-blocked location (my backyard - it's fairly free of interfering lights). The large tree and shrubs I'm trying to work around block a good bit of sky to the South and South-East, with a smaller amount to the SSW. To miss the top of this tree requires >40deg Dec - not a goer - and it is too wide to choose 30deg Dec to the East. However, I did find one possible part of my southern sky that might just work. I need to have RA at 1.5hrs westward and Dec at 25deg westward. This would allow the SCP PA routine to rotate 4 hours to 2.5hrs eastward (PA rotates the RA anti-clockwise looking at the pole), both end points being free of interference. Now, can the PA routine accommodate both off-axis adjustments simultaneously?

You raised the question as to slew rates: the Gemini defaults to Sideral when it starts up. However, rates for Slewing, Centring, and Guiding are installed in its memory, and currently my Slew rate is 4deg/sec.

If the weather gods are kind enough, I may be able to find a window of cloudlessness to try out this method. I'll let you know.

Paul.

w7ay

cpl42 This would allow the SCP PA routine to rotate 4 hours to 2.5hrs eastward (PA rotates the RA anti-clockwise looking at the pole), both end points being free of interference. Now, can the PA routine accommodate both off-axis adjustments simultaneously?

That should work fine, Paul.

Remember that Polar Alignment is to make sure your polar (RA) axis points to one of the celestial poles. There is no camera at the polar axis (unless you have installed a PoleMaster).

The start and end points, on the other hand, have nothing to do with the pole. They simply provide two measurement plates that have the same declination angle from the pole -- that is why only the RA motor is driven that 60º. The declination motor has to be absolutelly stationary, otherwise the two-point measurement method won't work. That 60º movement is done by slewing the RA axis; it is not done as a GOTO (and thus no Meridian flip either as the slew crosses the Meridian).

ZWO probably don't understand the math, and copied the code from somewhere, otherwise they could have done 3 points with the same declination, and get good accuracy even more than 40º in declination from the pole.

When the ASIAIR reports polar alignment accuracy from the pole, that is not the actual error. It could say 0" accuracy, but if you are using 30º in declination, it could be off by a lot (2 arc minutes, if memory serves -- I can look for my old measurements again in my files, if you are interested). (I personally use 89.5º declination and starting with the OTA tilted 60º to the east. After the 60º rotation, the OTA is top dead center, and give me the best clearance for me to adjust the altitude and azimuth bolts and locks.)

With 180 second type exposure times (you don't need more than that unless you are in Bortle 4 or better), each subframe will not show any field rotation even with 2' off the pole unless your focal length is really long and/or your plate is really large. So, you should be OK. The subframes are allowed to be rotated from one another, since a good stacking program will handle that (not the stacker in ASIAIR -- the ASIAIR stacking only handles Galilean transformation, not a full transformation matrix).

Good luck.

Chen

cpl42

Chen, I've just returned home and found your reply.

Thanks for confirming that my combined RA + Dec adjustments should work. It's not looking good for tonight - there's both low-mid level stratocumulus and high level cirrostratus cloud systems at the moment, and the prediction is for >80% cloud cover. And here am I champing at the bit to try this method - if it works, it would be a good substitute for All Sky. I'm coming around to agree with you that there might be a bug in the All Sky function code concerning South Pole alignment.

Also, thanks for the technical background. I was aware of some/most of this and realized that ASiair was using plate solving at two rotation locations to determine using circular/spherical coordinate maths how far off the centre of rotation was from the known/calculated polar position, and offering a simple tool to help you negate/minimize this offset. While the standard SCP/NCP PA routine actually includes the polar region in its shots, you now advise me that this is not a necessary condition, but any field within a 30deg radius of the pole will also work with the usual proviso that the start and end fields are clear of interference. (I notice that where SCP/NCP PA uses just 2 shots, All Sky uses 3 shots!)

I am lucky in that I am photographing from a Bortle 3 site, so 600sec exposures are reasonably clear of light pollution effects. My pole-visible site, however, has street lights that are mildly troublesome making it more a Bortle 4 location. Again, I can get decent results from 600sec exposures as the background gradient is not too severe. And I'd be quite happy with 2' off-axis error - it should be well within the capabilities of my guiding system (it would become a nuisance, though, should guiding be interrupted by cloud - an automatic field re-centring routine built into the guidance system would be useful). I use Pixinsight for most of my image processing and it quietly sorts out frame differences, including rotation, in its image registration process - I'm usually totally unaware that rotation has occurred, and am not really sure that I have seen it!

BTW: I like your idea of having the standard PA routine finish the 2nd rotation at or near the top of the RA axis. I'll keep that in mind for my next session. And, I'll keep you posted as and when I can try out your method. I'd like to nail M83 soon as it's well placed for my backyard/"dark" site allowing me around 7 hours imaging at the moment.

One other thing: it would appear that All Sky should work if it has to/decides to cross the Meridian, since it seems to be following a similar procedure to your method. Am I right, here?

Clear skies, Paul.

w7ay

cpl42 I am lucky in that I am photographing from a Bortle 3 site

No kidding?! I am too old to find dark sites -- I don't even go to a dark site for meteor showers anymore. There are other aspects of this hobby that appeal to me more than taking photos, in any case.

So, are you out in the outback by a bilabong?! :-)

I like your idea of having the standard PA routine finish the 2nd rotation at or near the top of the RA axis.

Yeah, it definitely keeps the cables and such clear of the bolts that you need to adjust and then tighten down (especially since my RST-135 is very compact -- there is no other strain wave mounts that are as small yet).

Too often, people read what ZWO writes in some abbreviated "documentation," about starting at top dead center, and don't think of the spherical geometry problem by themselves).

By the way, "top dead center" actually corresponds to the OTA pointing to East or West horizon when the Declination is set to zero, believe it or not.

Hope it works. The only time it doesn't is when someone is imaging from a south facing balcony (or in your case, someone imaging from a north facing balcony).

Once you find a starting location that works, write the declination angle and the LST minus RA (which, ta da, is the hour angle of the mount) in your note book. In the future, subtract that hour angle from the LST, and you have the RA. Declination does not change from session to session, of course.

Chen

cpl42

Well, here in Australia, our population is concentrated in a number of densely inhabited cities along the seaboard in most cases, with a few modest-sized inland centres. We are a "clustering" population and have the reputation of one of the most urbanized countries in the world. Consequently, you don't need to move far from small towns before you get reasonably dark conditions. I live in a modest village of ₇-8,000 spread in pockets over a 10x3 km area. Where I live, apart from the aforementioned street lights, the nearest concentration of light pollution is over 2 km away, making it easy to get Bortle 3 away from the street lights. I don't need to move too far (max. 5 km) and I have Bortle 2! It's not bad for being able to live in comfortable conditions with most amenities and services readily available, and fairly good services only a half-hour drive away. Our astronomy club is about to develop a dark sky site about 10 km from a ₅,000 population "village" and they are reckoning on <Bortle 2. I'm sorry, this must be making you green!!!

At the moment I have ₄₀% scattered cloud, so I'm going to give it a shot!

Paul.

cpl42

Chen, your revised SCP/NCP PA method worked. My eyeball alignment gave me ₁.5 deg RA and ₀.5 deg Dec error, and I was able to reduce this down to a total error of ₃". I'm at present trying to Autorun 600sec exposures on M83, but partly cloudy conditions are becoming a nuisance.

For those who are reading this/might read this in the future, my method is as follows:

Preselect 2 sky locations, 4 hrs apart in RA, and up to 30deg Dec from the normal 0deg starting position. For me, these were 1.5hrs West RA from the polar position (slewing to 2.5hrs East in the PA routine), and 25deg West of the SCP. These locations must be free of obstructions so that plate solving is not compromised.
Using the direction controls in the Preview screen, slew the OTA to the starting location. For me, this is the 1.5hrs RA West and 25deg Dec West as my PA routine rotates the RA axis in the easterly direction. Use the Preview and Solve controls on the starting sky field to check that ASiair is capable of doing its "thing" in the PA routine.
Begin the PA routine. It's start-up screen has the usual Celestial Pole instructions with the dotted line from the telescope image to the Pole. Ignore these instructions, you are already at the starting point. Proceed with the PA process. After it slews the 'scope to the 2nd position and plate solves, the telescope's axis pointing error should be displayed. Make what adjustments you deem necessary.
Note: I have opted to use the Preview telescope slew controls to move the OTA to the starting position. This is because at the end of the PA process, ASiair knows where the OTA is pointed, and can now move to any target. The alternative of moving the OTA by hand after loosening the axes clutches will mean that ASiair doesn't know where the Park/Home position is and this will need to be rectified before proceeding with the night's imaging work.

I hope all this helps. Paul.