Our Beautiful Universe - Photos and Videos

[Drakkith]

I was hoping to get a good image of Thor's Helmet (NGC 2359) tonight to show off, but due to flexure issues I couldn't do any guiding, so I called it a night soon after starting a headed in out of the cold to eat some hot tortilla soup. So here's a picture from the ESO instead:

I don't know when I'll break out the telescope next. It's become such a huge source of frustration that I sometimes think about getting rid of it all.

 

[timmdeeg]

That's a wonderful nebula, I put it on my to do list.

Frustration, yes well known. Within the last month's we had 2 to 3 times clear sky only and at these occasions when I was ready with alignment dew issues started and I had to give up.

I don't know what "flexure issues" means, is this a mechanical problem?

 

[Drakkith]

I don't know what "flexure issues" means, is this a mechanical problem?

Flexure occurs when imaging with one telescope and guiding with another telescope that is typically attached to the first. There is a slight amount of flex in each telescope and its mountings and attachments, which is different between the two telescopes. One flexes more than the other basically. So when guiding this difference in flex makes the imaging telescope shift off target slightly even though the guide scope stays locked on its guide star.

Long story short, things flex and it makes blurry images.

 

[collinsmark]

I don't know when I'll break out the telescope next. It's become such a huge source of frustration that I sometimes think about getting rid of it all.

I can appreciate/empathize with the frustration.

I started with astrophotography about 15 years ago. I kept up with it for about a year, with only limited instances of success. Then I gave it up for awhile, sort of. It was just so much work and effort, just to have things dashed by vibration or tracking or flexure or dew or unexplainable gradients or god-knows-what.

Rare instance of something I would call a "success," from back in 2008.

While I stopped setting up the telescope regularly, it was always in the back of my mind. It was at about the same time I discovered Physics Forums. And at that period of my life I found that working through problems in physics (helping students with homework sort of stuff) was as gratifying, if not more-so, than mucking with the telescope and the dew and all the layers of failure and frustration.

But the telescope was always there in the background (literally). I was always keeping an eye on new developments in backyard astronomy technology (e.g., the rise of lower noise CMOS sensors), as well as ways to improve my techniques. As the years progressed, it became apparent that a good chunk of my gear (e.g., old Meade camera) became obsolete. But I kept studying astrophotography none-the-less, even if I wasn't actually doing it physically.

Then, finally, a couple years ago I bit the bullet and made some equipment upgrades, thrusting me deep down the astrophotography rabbit hole. And a result, like many before me, it's overtaken my life and I can't get out.

The telescope, the camera and I have become bounded by a fortified garrison of dovetail adapters & plates, mounting rings, field flatteners, reducers (a.k.a., telekompressors), dew straps, and T-thread adapters (M42, M48 and more). Mentally speaking, stacking techniques, image processing algorithms and the Central Limit Theorem lord over my existence. There is no escape. Thanks to all the studying, however, at least I know where I am.

The upside is my astrophotos have shown some limited signs of improvement.

A little bit better image from late 2021.

I don't know what "flexure issues" means, is this a mechanical problem?

Yes, flexure is a mechanical problem. It means that due to gravity*, the guide camera system gets pulled slightly differently than the main camera system. This causes unwanted drift in the main image, as both systems track across the sky (to counteract the Earth's rotation).

*(while it's usually due to gravity, it can also be caused by improper cable management.)

It's the reason I try to avoid guidescopes when possible, and use off-axis guiders (OAGs) instead. This is particularly true for me when imaging with long focal length setups.

 

[timmdeeg]

Thanks for explaining flexure. If I understand it correctly it isn't a problem of the mount driving system but of a non rigid connection between the autoguiding camera and the telescope. Its hard to imagine though, my StarAid Revolution is fixed on the tubus by screws such that no relative motion is possible due to gravity.

I started in 1998 deep sky visual observing always thinking that I can watch the nicest astro images any time. It needed a particular experience to convince myself to start astrophotography. That happened just one year ago. And I had a hard time to learn all these details to do it and to process the pictures. But now its almost an obsession.

 

[collinsmark]

Thanks for explaining flexure. If I understand it correctly it isn't a problem of the mount driving system but of a non rigid connection between the autoguiding camera and the telescope. Its hard to imagine though, my StarAid Revolution is fixed on the tubus by screws such that no relative motion is possible due to gravity.

Flexure isn't necessarily just a issue with the guide camera to telescope connection (although that could be one potential cause). It could be any differential flexing. Another example could be slight sag caused by the weight of the comparatively heavy main camera with respect to the optical tube assembly (OTA), without a corresponding sag on the guidescope and guide camera system. Another potential cause is having USB cables (or any cables) dangling down causing torque on whichever cameras they're connected to. With long focal length systems, it doesn't take much.

 

[timmdeeg]

Flexure isn't necessarily just a issue with the guide camera to telescope connection (although that could be one potential cause). It could be any differential flexing. Another example could be slight sag caused by the weight of the comparatively heavy main camera with respect to the optical tube assembly (OTA), without a corresponding sag on the guidescope and guide camera system. Another potential cause is having USB cables (or any cables) dangling down causing torque on whichever cameras they're connected to. With long focal length systems, it doesn't take much.

ah I see, multiple possible reasons. Then it might be not trivial to detect the origin of flexure. Thanks!

 

[Devin-M]

Isn’t flexure not a problem after the images are stacked? I thought the registration stage where each image is adjusted for X, Y and rotation differences during the stacking process would take care of it. I don’t use guiding and consequently during a 2 hour imaging campaign I expect some field drift but that doesn’t seem to significantly affect the final output in my experience.

https://free-astro.org/index.php?title=Image_registration

“Registration is the process of aligning a set of images so that they can be used as if taken from the same point of view, in astronomy this applies to aligning images in order to be able to stack them.”

 

[Drakkith]

Isn’t flexure not a problem after the images are stacked?

Flexure blurs individual exposures (at least for me it does) so stacking does nothing to help.

 

[Drakkith]

It's the reason I try to avoid guidescopes when possible, and use off-axis guiders (OAGs) instead. This is particularly true for me when imaging with long focal length setups.

Indeed. I was using a 2000 mm focal length OTA as my imaging scope last night. My much shorter focal length refractor doesn't seem to suffer from flexure as far as I can tell. I actually made sure to get an OAG when I last upgraded my camera to a new Atik ccd, but the OAG requires a specific Atik guide camera. Which I didn't find out about until I got it in and tried to attach my Orion guide camera.

 

[Andy Resnick]

Well, it's been about 3 straight weeks of 100% overcast skies, day and night.... there were only a few hours here and there Here's what I have for Neptune, in October it was at the bottom, moving upwards until early December when Neptune switched back to prograde motion. I have been able to image on only 4 nights since that point:

Uranus is still in retrograde for another couple of weeks, here's what I have so far- beginning in November at bottom:

I'm able to occasionally capture 4 of Uranus' moons (200% zoom), which I'm very pleased about!

This time of year M45 (Pleiades) is well positioned, but I'm not sure how many more clear nights I'll have (this year).

11.9 hours of integration, 6s subs. I'm starting to capture the Horsehead nebula (IC 434) and neighborhood (NGC 2023, NGC 2024, IC 431, IC 432, IC 423, IC 426), 6.8h integration, 6s subs:

The first 3 images taken @ 800/8, the last taken @ 400/4.

Long story short, I figured out the root cause of my tracking problems- I ended up replacing the gearbox. It seems I have to replace it every few years (I suspect tips of the plastic gearteeth begin to snap off..?), so I should have figured it out sooner. It's only $30, so I consider it a consumable item.... now the problem is, all of the PEC training I did with the bad gearbox has to be deleted :)

 

[Andy Resnick]

I don't know when I'll break out the telescope next. It's become such a huge source of frustration that I sometimes think about getting rid of it all.

I empathize! Even with my simple setup, it seems there are several dozen things that can go wrong on any given night.....

 

[Drakkith]

@Andy Resnick Beautiful images! Do you have a permanent setup like a pier for your equipment? How do you keep you PEC data between sessions?

 

[DennisN]

I'm able to occasionally capture 4 of Uranus' moons (200% zoom), which I'm very pleased about

That's very cool!
And very nice images too!

 

[Devin-M]

 

[DennisN]

Even with my simple setup, it seems there are several dozen things that can go wrong on any given night.....

I have the same experience.

I remember one time I was going to shoot the Moon at a new location in a city park further away from home. I was in a bit of a rush since the Moon was not going to be visible for very long (it was setting), so I quickly packed my backpack with the gear and went to the park.

When I arrived I picked a spot and started to assemble the gear, but then I noticed I had forgot the tripod at home.

I didn't get that sad, I mostly laughed at the situation.
Good luck shooting close-ups of the Moon without a tripod.

Nowadays I have a checklist on my phone with a list of needed gear.

 

[Andy Resnick]

@Andy Resnick Beautiful images! Do you have a permanent setup like a pier for your equipment? How do you keep you PEC data between sessions?

Thanks!

I don't have a pier. My understanding is that PEC (rather, the error itself) arises from mechanical (machining) errors in the worm gear, and so PEC acts on the errors inherent to the mount, as opposed to polar alignment error (which does vary from night to night). I'm not sure about the specifics, but I believe my mount driver (Gemini 2) saves consecutive PEC runs as a moving average, converging on the 'optimal' solution.

For me, correcting polar alignment error is part of the initial setup, a process that aligns the Gemini's map of the sky to where my lens is actually pointing. Periodically during imaging (say, every 30 minutes or so), I will re-run the alignment routine, essentially performing a drift alignment process that also compensates for additional mount mechanical errors (non-perpendicularity, for example) as the night progresses. It's unclear (to me) if the PEC corrections and the alignment corrections are independent or not.

 

[Devin-M]

…its been too cloudy to bring out the telescope for the past few weeks so I made a few renaissance painting inspired AI renderings celebrating the pursuit of astronomy…

 

[Drakkith]

Thanks!

I don't have a pier. My understanding is that PEC (rather, the error itself) arises from mechanical (machining) errors in the worm gear, and so PEC acts on the errors inherent to the mount, as opposed to polar alignment error (which does vary from night to night). I'm not sure about the specifics, but I believe my mount driver (Gemini 2) saves consecutive PEC runs as a moving average, converging on the 'optimal' solution.

Do you do a new PEC every time you get your gear out to image?

For me, correcting polar alignment error is part of the initial setup, a process that aligns the Gemini's map of the sky to where my lens is actually pointing. Periodically during imaging (say, every 30 minutes or so), I will re-run the alignment routine, essentially performing a drift alignment process that also compensates for additional mount mechanical errors (non-perpendicularity, for example) as the night progresses. It's unclear (to me) if the PEC corrections and the alignment corrections are independent or not.

As far as I understand, PEC only compensates for errors in the RA worm gear. As the worm gear turns, any errors in its shape cause the connecting gear to turn at different rates, leading to a faster or slower mount speed that reoccurs every time the worm gear makes a full turn. Hence the 'periodic' in 'periodic error correction'. Polar alignment is separate and has nothing to do with PEC.

 

[Devin-M]

With my Starwatcher 2i equatorial mount with no guiding & the 2175mm focal length reflector on board, if I limit my exposures to 90 seconds I can keep about half of them, the other half being wobbled by periodic error. I should note this setup is best guess 4x over the official weight limit of the mount so required modifying with additional counterweights, but including the full frame 36MP Nikon D800 camera used the whole setup is sub $2k USD.

Starwatcher 2i equatorial mount ($429):
https://www.bhphotovideo.com/c/product/1604032-REG/sky_watcher_s20512_star_adventurer_pro_pack.html/?ap=y&gbraid=0AAAAAD7yMh1dZbFi8tc5Lg9RH2dNka5Gn&gclid=CjwKCAiAk--dBhABEiwAchIwkXrvCY1AQQ-eKxDwpKKP-J8uwHqiTSp9hrykq8fSP2c67IPnXp6IhRoCty8QAvD_BwE&lsft=BI:514&smp=y

2175mm focal length reflector ($799):
https://www.walmart.com/ip/Meade-In...assegrain-Telescope-OTA-Optical-Tube/50020290

Nikon D800 ($625 used):
https://www.keh.com/nikon-d800-36-3...nRmOURJ3XWD_jmwiXtMWVd4M3yUioOhxoC7UUQAvD_BwE

Tripod ($100):
https://www.horronline.com/shop/manfrotto-055xprob-tripod

 

[Andy Resnick]

Do you do a new PEC every time you get your gear out to image?

No. I'm lazy- my conditions for a PEC run are (1) imaging @ 800mm, and (2) weather warm enough so I'm not freezing my giblets :) I typically PEC train in Fall and Spring.

As far as I understand, PEC only compensates for errors in the RA worm gear. As the worm gear turns, any errors in its shape cause the connecting gear to turn at different rates, leading to a faster or slower mount speed that reoccurs every time the worm gear makes a full turn. Hence the 'periodic' in 'periodic error correction'. Polar alignment is separate and has nothing to do with PEC.

Yes- I meant that PEC only compensates for the RA worm drive. Polar alignment (meaning the manual pointing of the mount) is separate, but it's not clear when I perform successive GoTo alignment procedures if the tracking algorithm (as opposed to the pointing algorithm) compensates for drift caused by worm gear error (in addition to all the other errors).

I'm also easily confused, so there's that.

 

[Drakkith]

No. I'm lazy- my conditions for a PEC run are (1) imaging @ 800mm, and (2) weather warm enough so I'm not freezing my giblets :) I typically PEC train in Fall and Spring.

Perhaps I've misunderstood PEC. I was under the impression that I needed to do a new PEC every time I set up my gear since the mount gets moved around in the RA axis during polar alignment. But now that I think about it, I guess the gears don't move since I'm disengaging the clutch each time I rotate it.

Yes- I meant that PEC only compensates for the RA worm drive. Polar alignment (meaning the manual pointing of the mount) is separate, but it's not clear when I perform successive GoTo alignment procedures if the tracking algorithm (as opposed to the pointing algorithm) compensates for drift caused by worm gear error (in addition to all the other errors).

Why do you do multiple GoTo alignments? I find that one is usually all I need unless I'm using my 2,000 mm focal length RC and I move all the around to the other side of the sky from where I was just imaging.

As for whether the target finding and movement command algorithms use the PEC, I don't know. Periodic error usually isn't so great that it would cause a GoTo problem. If it does, then you may need a new worm gear.

 

[collinsmark]

Perhaps I've misunderstood PEC. I was under the impression that I needed to do a new PEC every time I set up my gear since the mount gets moved around in the RA axis during polar alignment. But now that I think about it, I guess the gears don't move since I'm disengaging the clutch each time I rotate it.

Allow me to ramble about PEC for a bit.

Whether or not periodic error correction (PEC) is stored permanently between sessions is dependent on the make and model of your mount. The terminology of "permanent" PEC also varies by manufacturer (just to make things more confusing).

As for some examples of mounts I'm familiar with, both my Meade LX200 mount and my Sky-Watcher EQ6-R Pro both support "permanent" PEC.
The Meade refers to it as "Smart Drive."
The Sky-Watcher refers to it as "PPEC."

Some mounts support PEC permanently and don't require you to do anything between sessions, Some mounts will remember the PEC settings, but you'll have to be sure to "Park" the scope between sessions, and some don't support it at all.

--

In order to support permanent PEC, the mount will need the following capabilities:

• Nonvolatile memory to store the PEC data.
• Some sort of internal mechanism to inform the algorithm of what part of the period the gears happen to be on (such as when the scope powers up). This could be an electromechanical mechanism, but it might be nothing more than re-loading a value stored in nonvolatile memory of the position that was recorded at the most recent time the scope was parked.

That second bullet is important. In many mounts, the current position of the gears within their period is not stored until the command is given to the scope to "Park." Then, the next time the scope is powered up, it just uses that value.

Meade's "Smart Drive" system for the LX200 has an optical encoder and a hole in a gear that lets it know where in the period it happens to be. When the mount is powered up (with "Smart Drive" enabled), it will automatically slew along the RA axis to determine the gear period position. Alternatively, if the scope is "Parked" between sessions, it will skip the "Smart Drive Initialization" sequence, since the PEC index was stored in Nonvolatile memory.

The Sky-Watcher EQ6-R Pro apparently has some mechanism to determine the gear position of the gears.

But a lot of other mounts can store the PEC data in nonvolalitile memory, but don't have an electromechanical mechanism for determining the gear positions. The implication there is that if you ever forget to "Park" your scope before packing up, or if the scope unexpectedly loses power while tracking, you'll need to redo the PEC at the beginning of your next session, for those mounts. (Sky-Watcher refers to this type of PEC as "SPEC.")

'Best check the mount's user's manual.

--

The interaction between PEC and autoguiding also vary between mount models. On the mounts that I've used, they both can be used together just fine. But I've heard that on some mounts, it's either one or the other.

If you want to enable PEC though, you'll also want autoguiding available to train your PEC. Sure, you could train the PEC by peering through a reticle eyepiece constantly for a half hour or so, but it's so much easier with an autoguider (and less backbreaking). With an autoguider, you can set things in motion, then go back inside and eat a sandwich or something.

--

And yes, @Drakkith, ideally, if you loosen the clutches and move the telescope around manually (on the RA and Dec axes), that shouldn't affect the PEC, since that motion doesn't affect the gear positions. Of course that will destroy any existing star alignment, but it shouldn't affect PEC.

 

[DennisN]

Comet ahoy!

I read in the news about a green comet becoming visible from Earth in the Northern hemisphere.
The comet is C/2022 E3 (ZTF);

C/2022 E3 (ZTF) is a long period comet that was discovered by the Zwicky Transient Facility on 2 March 2022. The comet will reach its perihelion on January 12, 2023, at a distance of 1.11 AU (166 million km) and the closest approach to Earth will be on February 1, 2023, at a distance of 0.28 AU (42 million km). The comet is expected to get brighter than magnitude 6 and thus become visible with the naked eye.

I looked on the net for some data about angular size, and according to this page:

The Comet C/2022 E3 ZTF or the iceball tail is expected to be stretching 2.5 degrees wide FOV. It is reported to be the closest to the Sun on 12h Jan 2023.

...and here is a page with a couple of very nice photos of the comet (by Aleix Roig).

Regretfully we've got very bad weather where I am at, but maybe some of you will have an opportunity?

E.g. :

@collinsmark could do a close-up of the comet's surface, so we can look at the geology
@Drakkith could do spectroscopy so we get to know what it's made of
@Andy Resnick could plot the orbit/trajectory in more detail
(it's your only chance with this one, next time will be in ≈50 000 years )

But don't feel any pressure, anyone!
Clear skies and good luck!

@timmdeeg , @bruha , @Devin-M , @Chemistree, @russ_watters may also be interested, perhaps?

 

[Drakkith]

@Drakkith could do spectroscopy so we get to know what it's made of

Sure. Got a spectrometer I can borrow?

 

[DennisN]

Got a spectrometer I can borrow?

Luxury.
When I grew up I had to determine compositions using my eye only.
(I'm of course referring to this sketch)

 

[Andy Resnick]

Why do you do multiple GoTo alignments? I find that one is usually all I need unless I'm using my 2,000 mm focal length RC and I move all the around to the other side of the sky from where I was just imaging.

As for whether the target finding and movement command algorithms use the PEC, I don't know. Periodic error usually isn't so great that it would cause a GoTo problem. If it does, then you may need a new worm gear.

Long post follows- sorry!

I have a really hard time visualizing what is happening when my mount is not perfectly polar aligned (know any good URLs that illustrate this?). In my mind, the idealized heavenly sphere (apparently) rotates only in RA and polar alignment error introduces slow RA (oscillations?) and DEC (linear drift?) movement into my field of view. For now, forget about PEC.

Here's my thought process (Losmandy GM8 w/ Gemini 2)- I start off with some small pointing error (meaning my mount is not perfectly polar aligned). This initial pointing error leads to a mismatch between, for example, the mount thinking I am centered at the bright star Capella and where the center of my field of view actually is. The initial GoTo alignment corrects for this initial pointing error. I do this initial alignment process for 3 or 4 bright stars, all well-separated in RA and DEC- then I can confidently point my mount at whatever dim object I want to image, and it is (initially) centered in my field of view.

Then, as time goes on, although the tripod itself is not moving, my field of view changes because the (apparent) movement of the stars that my mount is correcting for is not true RA-only movement. The drifting leads to worse- and worse-quality subs as the night progresses. If I trace the position of a star in successive frames, it traces out a sort-of sawtooth pattern (see what I've posted earlier) that (slowly) grows in RA amplitude and DEC 'wavelength'. There are two approaches I can use here- periodically re-polar align and periodically re-perform the GoTo alignment procedure. I don't need to turn off the mount for this; typically I do both and also throw in a focus check for good measure. My hypothesis is that either one 'resets the clock', and I see this in the stacking process- star roundness slowly degrades and is restored after each resetting.

In this hypothesis, periodic error simply increases the rate of total misalignment.

This is all phenomenological- I have no idea what the motor driver software is actually doing, I only have some semi-quantitative readouts of the motors' duty cycles and mount 'alignment parameters'. I have yet to make these respond in any repeatable or predictable way (for example, intentionally doing a poor initial alignment and watching the parameters converge to something).

Prior to replacing the gearbox, I took a close look at the RA worm gear and there were no obvious defects (scratches, worn areas or flat spots, etc). There are still some mount anomalies I haven't been able to identify:

1) The RA motor sound while slewing is a steady tone in one direction and a different tone (with slight warbles) in the opposite direction. It's greatly improved after the gearbox replacement, but the asymmetry is odd. The DEC motor gives steady tones in either direction.

2) perhaps related, moving/guiding the mount in 3 directions behaves as expected- the stars move when I push the button and stop when I release. In the 4th direction, tho- one of the RA directions- the stars 'coast to a stop' when I release the button. Again, this is well improved post-gearbox replacement, but it's not entirely gone.

Maybe these do indicate the worm gear needs replacing? A new one is about $120 and the procedure is moderately delicate (feeler gauges for the guideblocks, for example) so I prefer not to deal with it.

Does any of that make sense?

 

[Andy Resnick]

Sure. Got a spectrometer I can borrow?

FWIW, I've played around with putting a transmission grating at the aperture stop, that gives nice spectra at the image plane: Here is Jupiter and Spica, an image I took a few years ago-

 

[Drakkith]

1) The RA motor sound while slewing is a steady tone in one direction and a different tone (with slight warbles) in the opposite direction. It's greatly improved after the gearbox replacement, but the asymmetry is odd. The DEC motor gives steady tones in either direction.

Not sure. It might be the asymmetry in the weight of each end of your mount's RA axis bar. That is, the telescope weighs X, the counterweights weigh Y, and unless they are perfectly balanced you get a different torque in each direction. In one direction the motor pulls against the weight, while in the other the RA axis wants to continuously 'fall' into the slack between the worm gear and worm wheel, perhaps leading to the warbling you're talking about. But that's mostly a random guess.

2) perhaps related, moving/guiding the mount in 3 directions behaves as expected- the stars move when I push the button and stop when I release. In the 4th direction, tho- one of the RA directions- the stars 'coast to a stop' when I release the button. Again, this is well improved post-gearbox replacement, but it's not entirely gone.

I have an identical problem. I'm almost certain it's because of slack between the RA worm gear and worm wheel. This is also why backlash happens. The slack causes the worm gear to have to turn some amount before before it touches the worm wheel when you reverse directions. Your mount should have a backlash compensation setting somewhere that can help, as should your guiding software.

As for your comments on polar misalignment, my understanding is as follows.

I agree that small polar misalignment (less than a degree) will cause a slight DEC and RA drift as your mount tracks the sky. I was under the assumption that the DEC drift was much more substantial than the RA drift, but I'm not certain. I think of it as the target and the mount tracing two different circles across the sky. The larger the polar misalignment, the greater the difference in these two circles.

Given that these two circles probably cross, the DEC and RA errors should oscillate back and forth over the course of 24 hours as the tracks diverge and then come back together to cross. Depending on how your misaligned you may get more or less of each error when you first go to the target. However, I think the sum of the two errors is more or less the same throughout the night.

Let's say you are misaligned by half a degree towards zenith from the pole. Now trace a line from the point in the sky where your mount's RA axis is pointing, through the actual celestial pole, and on down through the sky to the opposite pole before coming back around to make a circle. If you point your scope at any target along this great circle, you should have zero DEC error, but maximum RA error. The arc your scope makes as it moves along its RA track should be tangent to any arc the stars make when both are touching that circle. In effect, your scope/mount is moving along a different sized circle than the target is, so even though they are tangent for an instant, they are moving at different speeds.

Conversely, rotate around your RA axis by 90 degrees either way and you should have maximum DEC error, as these points should be where the two arcs by the telescope and the sky are misaligned the most. That is, the two arcs should have the greatest angle between them, and any motion perpendicular to the circles the stars make is declination by definition.

If I trace the position of a star in successive frames, it traces out a sort-of sawtooth pattern (see what I've posted earlier) that (slowly) grows in RA amplitude and DEC 'wavelength'. There are two approaches I can use here- periodically re-polar align and periodically re-perform the GoTo alignment procedure. I don't need to turn off the mount for this; typically I do both and also throw in a focus check for good measure. My hypothesis is that either one 'resets the clock', and I see this in the stacking process- star roundness slowly degrades and is restored after each resetting.

That's strange. I can't really understand why this would happen. I'm pretty sure your GoTo alignment does nothing. The mount, once on target, should only use its RA axis motor (or whatever axis if you're using an alt-az with a wedge and not a GEM mount) to track the target. Even if you don't do a GoTo alignment at all, the telescope should track just fine as long as you can find the target.

As for redoing your polar alignment, I can't say. If you do it initially and it's good, then I can't see how redoing it would help unless you became misaligned somehow.

Maybe these do indicate the worm gear needs replacing? A new one is about $120 and the procedure is moderately delicate (feeler gauges for the guideblocks, for example) so I prefer not to deal with it.

If you're not having large or uncorrectable errors in your RA axis, then you should be fine. All worm gears are imperfect to some degree.

 

[Andy Resnick]

Sure, you could train the PEC by peering through a reticle eyepiece constantly for a half hour or so, but it's so much easier with an autoguider (and less backbreaking). With an autoguider, you can set things in motion, then

The GM-8's worm rotates once every 8 minutes, so I only have to stay hunched-over-motionless-staring_at_a_bright_dot for 8 minutes, not pleasant but not terrible.

go back inside and eat a sandwich or something.

:P Go outside and enjoy nature! All the critters come out to play :)

 

[Andy Resnick]

Let's say you are misaligned by half a degree towards zenith from the pole. Now trace a line from the point in the sky where your mount's RA axis is pointing, through the actual celestial pole, and on down through the sky to the opposite pole before coming back around to make a circle. If you point your scope at any target along this great circle, you should have zero DEC error, but maximum RA error. The arc your scope makes as it moves along its RA track should be tangent to any arc the stars make when both are touching that circle. In effect, your scope/mount is moving along a different sized circle than the target is, so even though they are tangent for an instant, they are moving at different speeds.

Conversely, rotate around your RA axis by 90 degrees either way and you should have maximum DEC error, as these points should be where the two arcs by the telescope and the sky are misaligned the most. That is, the two arcs should have the greatest angle between them, and any motion perpendicular to the circles the stars make is declination by definition.

Many thanks for the suggestions- for example, I do have a backlash compensation setting, I'll give it a try and see what happens.

The paragraphs above are well-written but I can't follow along. In my mind, it's a conceptually simple problem (for a German equatorial mount [GEM]):

Imagine you are at the common center of 2 rotating spheres, one is rotating coincident with the Earth's rotation axis and the other rotating on an axis at some small angle δ relative to the Earth's rotation axis- the polar alignment error.

Initially, a particular point on the first sphere (the reference sphere, corresponding to perfect polar alignment) is located at coordinates (RA, DEC) and that same point on the second sphere is given by coordinates (RA', DEC'). The coordinate transformation is fairly straightforward: RA' = RA-δ and DEC' = DEC, for the example of pointing error entirely in RA.

However, as time progresses, the reference point remains (RA, DEC) but in the second frame of reference the coordinates are (RA'(t), DEC'(t))- I know this because of drift.

I just don't know how to transform between these coordinate frames- it's got to be a solved problem (it's just geometry!), and I can imagine coding the solution into GEM motor drive software. RA'(t) and DEC'(t) are functions of δ for sure and likely RA and DEC as well, and so it's possible the drive software can be written to compensate for polar alignment error. I just don't know what's going on the Gemini 2's brain.....

 

[DennisN]

FWIW, I've played around with putting a transmission grating at the aperture stop, that gives nice spectra at the image plane

Darn, they were quite cheap on Amazon. I'll buy me a pack of those, they could be fun to play with . Thanks for the idea!
And it was a very cool photo with overlayed spectra!

@Andy Resnick , by the way, I remember you said you bought a new hard drive (or drives?) on Amazon recently. Would you mind saying which one? I'm thinking of buying two new ones, I'm starting to run a bit low on disk space.

 

[collinsmark]

Allow me to ramble about polar alignment for a bit. This is a subject that I thought about quite a bit back in 2008, soon before I joined Physics Forums.

Polar alignment mathematics is a little weird, because it not only involves true angles, it also involves distances that are themselves measured in angular units.

For example (Figs 1 and 2), in normal geometry, the distance of the pink arc length in the image below (Fig. 1) is measured in units of distance.

Figure 1. Traditional geometry. Distances are in units of distance.

But if the radius is itself an angle, the distance is also an angle (Fig. 2).

Figure 2. In polar alignment math, everything is an angle. Even lengths are really angles.

To complicate matters more, we work in 3 dimensions. All the angles involved are bounded to the celestial sphere (Fig. 3).

Figure 3. Even more angles. This time on a sphere.

@Drakkith conceptualized the matters at hand by introducing the idea of two circles, slightly offset from each other. Once circle is a star's path around the celestial pole, and the other is the telescope tracking at the same declination (Dec) as the star. That's a good way of thinking about it. But it gets tricky when considering the resulting "drift."

But the telescope won't notice drift in the right ascension (RA) axis unless it's tracking a star so close to the pole that it's within the ballpark of the polar alignment error. In other words, if your polar alignment is a whole degree off, you won't notice much drift in the RA axis unless you're tracking a star that is a degree or two from the actual pole.

But for the rest of the sky (which is most of it), the drift is almost completely relegated to the Dec axis.

Example 1: Azimuth error. Azimuth error is most pronounced for stars near the meridian. Altitude error has little to no effect on stars near the meridian.

Here's an approximation example where of a star on the celestial equator near the meridian (Fig 4), where the drift is caused specifically by azimuth error:

Figure 4. Star on celestial equator, near the meridian, moving from left to right.

Variable definitions for Fig. 4:

• $t$: Time, in units of minutes
• $R$: Rate of drift across the sky for a star on the celestial equator, in units of radians per minute. Approximately $\frac{\pi}{720}$ rad/min.
• $\theta_{az}$: Azimuth error, in units of radians.
• $L$: Approximate, absolute value of the latitude of the observing location (can be degrees or radians; just make sure the units match that of the corresponding trig functions that act upon it.)
• $d$: Magnitude of the star's declination drift (i.e., what we're trying to minimize).

We can fix this by adjusting the azimuth axis by $a_{az}$, the resulting angular distance that would be seen in the eyepiece or sensor, where

$a_{az} = \theta_{az} \sin(L)$.

This azimuth adjustment would be seen through the eyepeice/sensor as taking place almost completely along the RA axis, even though it's the azimuth axis that's being adjusted.

Figure 5. Azimuth adjustment.

And after some algebraic substitutions and some small angle approximations, we can use our measured values to find that

$\frac{a_{az}}{d} = \frac{\sqrt{1 - \cos^2(Rt)\cos^2(L)}}{\sin(Rt) \cos(L)} \approx \frac{229 \tan(L)}{t}$, for small $t$ (less than a half hour or so).

Of course that $229 \tan(L)$ is a number you can calculate before-hand. You wouldn't ever need to do that in the field. So let's call $C_{az} = 229 \tan(L)$.

Then the equation simplifies to $a_{az} \approx C_{az} \frac{d}{t}$

Figure 6. Azimuth adjustment as would be seen through eyepiece or sensor. Note: adjustment might extend beyond field of view of eyepiece/sensor.

Figure 7. Azimuth adjustment direction based on hemisphere and telescope configuration.

Figure 8. Azimuth adjustment direction based on hemisphere and telescope configuration.

And just to re-iterate, the drift will occur almost completely along the Dec axis. While it appears the adjustment is along the RA axis, it is accomplished by adjusting the azimuth knobs on the mount (not the RA controls).

To be continued with an altitude error example...

 

[collinsmark]

...Continued from last post.

Btw, the reason I'm posting this stuff on this thread is because people were interested in polar alignment concepts, and how polar alignment error affects the telescope's drift. I'm hoping this stuff is on-topic.

Example 2: Altitude error. Altitude error is most pronounced for stars near the celestial equator and the horizons.

Here's an approximation example where a star is on the celestial equator, but kind-of-sort-of near the East or West horizon.

Due to the atmosphere though, we can't choose a star too close to the horizon, or atmospheric refraction would interfere with our measurements. So you'll need to choose a star near the celestial equator with an altitude at least around 20 deg above the horizon -- but not too far above the horizon, no more than maybe 45 deg if obstructions allow.

Once you choose a star, measure it's angle along the celestial equator, starting at the intersection of the equator and horizon. Call this angle, measured in degrees, $\phi_{ra}$.

You can use your outstretched hand to measure $\phi_{ra}$.

Figure 9. Measuring angles with your outstretched hand.
(Image source: https://www.abc.net.au/science/articles/2009/07/27/3169109.htm)

Anyway, Fig. 10 shows how Altitude (Alt) error affects drift.

Figure 10. Example star near the horizon (East here) on the celestial equator, moving from lower left to upper right.

Variable definitions for Fig. 10:

• $t$: Time in units of minutes.
• $R$: Rate of drift across the sky for a star on the celestial equator. Approximately $\frac{\pi}{720}$ rad/min.
• $\theta_{alt}$: Altitude error, in units of radians (i.e., what we're trying to minimize).
• $\phi_{ra}$: Absolute value of the angular distance starting from the intersection of the celestial equator and horizon, ending at the star of interest, measured at the time the drift measurements begin. Units are degrees.
• $\theta_{ra}$: Same as $\phi_{ra}$ except converted to radians.
• $d$: magnitude of star's declination drift, in units of radians.

Once again, as seen in the figure, the drift will occur almost exclusively along the declination axis. The declination drift can be calculated as:

$d = \theta_{alt} \left[\sin(\theta_{ra} + Rt) - \sin(\theta_{ra}) \right]$

We can correct the error by making an adjustment of $\theta_{alt}$ on the mount's altitude knobs, which would have a corresponding change of

$a_{alt} = \theta_{alt} \sin(\theta_{ra} + Rt)$

on the position of the star, as seen from the eyepiece or sensor. This adjustment is shown in Fig 11.

Figure 11: Altitude adjustment. (Star in the East shown here.)

(The equations are same for a star in the West, except that the $Rt$ term gets a negative sign.)

After some algebra and some small angle approximations we find

$$\frac{a_{alt}}{d} = \frac{\sin(\theta_{ra} + Rt)}{\sin(\theta_{ra} + Rt) - \sin(\theta_{ra})} = \frac{\sin(\theta_{ra}) \cos(Rt) + \cos(\theta_{ra}) \sin(Rt)}{\sin(\theta_{ra}) \cos(Rt) + \cos(\theta_{ra}) \sin(Rt) - \sin(\theta_{ra})} \approx \frac{229 \tan(\phi_{ra})}{t} + 1$$

for small $t$ (less than around a half hour or so).

For a star in the West, the approximation is the same except for a negative sign.

$\frac{a_{alt}}{d} \approx \frac{229 \tan(\phi_{ra})}{t} - 1$.

In order to save some effort in the field, you could decide to patch of sky where you will choose your star ahead of time. That way you can calculate the $C_{alt} = 229 \tan \phi_{alt}$ ahead of time. Then in the field, you'll need to calculate

$a_{alt} \approx \left( \frac{C_{alt}}{t} + 1 \right)d,$ if the star is in the East, or

$a_{alt} \approx \left( \frac{C_{alt}}{t} - 1 \right)d,$ if the star is in the West.

Unlike the azimuth adjustment, in the case of altitude error and adjustment, both the drift and the apparent movement of the star during adjustment occur on the Dec axis. Note however, the adjustment is accomplished with the altitude knobs on the mount, not the declination controls.

In the case of a star in the East, the adjustment should be made such that the star's apparent motion in the eyepiece or sensor is in the opposite direction as the drift. See Fig. 12.

Figure 12. Altitude adjustment as would be seen through the eyepiece or sensor for a star in the East. Note: adjustment might extend beyond the field of view of the eyepiece/sensor.

For a star in the West, the adjustment should be made in the same direction as the drift (Fig. 13).

Figure 13. Altitude adjustment as would be seen through the eyepiece or sensor for a star in the West. Note: adjustment might extend beyond the field of view of the eyepiece/sensor.

More thoughts on polar alignment:

If you have the capability to hook up your scope to a digital camera and computer, don't do polar alignment as discussed here. Use plate solving instead (such as what can be done with NINA or SharpCap). I haven't performed polar alignment like what's discussed here for years. Plate solving is the way to go. Once the software is installed and configured, it only takes a few minutes to do polar alignment with plate solving. It's a game changer.

---------------------

@Andy Resnick, you mentioned you were having problems with drift along the RA axis. I can all but guarantee that the culprit is something other than polar alignment error. Unless you're trying to track a star really close to the pole, like Polaris, polar alignment error shows up almost exclusively as Dec drift.

So if you're experiencing RA drift, I'd look elsewhere, like the periodic error of your mount's tracking motors, flexure, balance, or something I haven't though of.

 

[Tom.G]

perhaps related, moving/guiding the mount in 3 directions behaves as expected- the stars move when I push the button and stop when I release. In the 4th direction, tho- one of the RA directions- the stars 'coast to a stop' when I release the button. Again, this is well improved post-gearbox replacement, but it's not entirely gone.

I just dropped in and noticed your postioning problem. The coasting to a stop in one direction only is typical of a counter-balance mis-adjustment.

When the counterbalances were first adjusted, one of them was likely a bit 'off', then later in the procedure a different weight (or 2) was/were adjusted to compensate. The end result is the COG of the scope + weights does not line up with the scope movement axes. This could also be a case of the weights COG not being in the same plane as the scope COG.

Not knowing the adjustment procedure for your particle scope/mount, I can't reasonably suggest any specific procedure other than Redo-From-Start.

You may get a clue by disengaging the drive and manually positioning the scope in different 3-dimensional attitudes.

Hope this helps,
Tom

p.s. I see @Drakkith beat me to this conclusion, also though I've only skimmed it, the post by @collinsmark looks close to exhaustive!