26 replies to this topic

[UP4014Fan]

I'm using NINA's advanced sequencer.  I've added the trigger of running autofocus if there's an HFR change of 5% over a sample of 5 images.  This works really well if the air is calm, and I like it because it accounts for flexture, atmospherics, and any slight mirror movements during the night. 

 

Right now, it's a bit breezy where I live, so the scope is being jostled every once in a while by the wind.  That results in slight "egging" of stars every few subs and NINA then decides to re-run autofocus, generally winding up almost exactly where it started.  I obviously don't want to disable the autofocus trigger, but I don't want it eating into imaging time by running unnecessarily.  Do I want to change the sampling or any other thoughts?

[Jim Waters]

Increase the HFR to 7% and see what happens.  Try to address the mount movement or stability later.

[UP4014Fan]

Increase the HFR to 7% and see what happens.  Try to address the mount movement or stability later.

I'll go with 7%.  The mount itself is pretty solid.  The wind's died down to five knots, so that's no longer a factor. 

 

I have added a second counterweight recently because with only one, it was at the end of the shaft and the mount was barely balanced.  I'm starting to think that I've got the mount too well balanced with the additional counterweight and it's "floating" in RA.

 

EDIT:  wind's down to five knots at Spirit of St. Louis.  Our house sits about 100 feet higher and the wind is still blowing here.

Edited by UP4014Fan, 22 June 2025 - 12:13 AM.

[PIEJr]

I run NINA for my acquisition and have the basic settings working fine.

My Autofocus settings read as: On Start; 30 minutes; 10 exposures; 5° (Temp Change); and 10% HFR amount.

 

If you think you have a wind problem affecting your stability, you could do what I used to do with my AVX mount and hang a 5 gallon bucket of water under the mount to weight it down.

That worked wonders for that mount. But since we have a pool, I would simply dip a pail of water, then hang it from the eyepiece tray.

 

I don't know what effect your Schmit-Cassigrain might be imposing on your AP. But for me, I shied away from reflectors so I'd have less to worry about.

 

I would rather err toward wasting time doing more focusing, that to deal with fuzzy pictures. Been there, done that, it wasn't purdy.

I recently changed from focusing every 5 images, to every 10 images. So far, so good.

 

The wind here is running up to 17.4 MPH, and it is causing extra focusing runs. One right now at 8 exposures. Abnormal condition noted.

Mostly, I monitor through a wireless camera and can hear the wind blowing through the camera app.

I wish you hadn't brought up the wind, doggone it. 

[UP4014Fan]

I run NINA for my acquisition and have the basic settings working fine.

My Autofocus settings read as: On Start; 30 minutes; 10 exposures; 5° (Temp Change); and 10% HFR amount.

 

If you think you have a wind problem affecting your stability, you could do what I used to do with my AVX mount and hang a 5 gallon bucket of water under the mount to weight it down.

That worked wonders for that mount. But since we have a pool, I would simply dip a pail of water, then hang it from the eyepiece tray.

 

I don't know what effect your Schmit-Cassigrain might be imposing on your AP. But for me, I shied away from reflectors so I'd have less to worry about.

 

I would rather err toward wasting time doing more focusing, that to deal with fuzzy pictures. Been there, done that, it wasn't purdy.

I recently changed from focusing every 5 images, to every 10 images. So far, so good.

 

The wind here is running up to 17.4 MPH, and it is causing extra focusing runs. One right now at 8 exposures. Abnormal condition noted.

Mostly, I monitor through a wireless camera and can hear the wind blowing through the camera app.

I wish you hadn't brought up the wind, doggone it. 

I tend to agree on focusing too often.  

 

I think the issue is that the dew shield acting as a sail.  It's got a lot of surface area and a long arm from the pivot points of the RA and DEC.

[rgsalinger]

I would suggest looking a bit more deeply to see just how much the HFR increases when you get a bad sub. Seven percent sounds fine (I use 8). 

 

You might want to increase the sub count to 10 which would reduce the effect on AF timing by 1/2.

 

Each sub then counts for less when calculating whether the trigger fires. 

[Jim Waters]

 I'm starting to think that I've got the mount too well balanced with the additional counterweight and it's "floating" in RA.

 

IMO floating in RA or DEC isn't good.  Being just a little off in polar alignment is best.  

[UP4014Fan]

I would suggest looking a bit more deeply to see just how much the HFR increases when you get a bad sub. Seven percent sounds fine (I use 8). 

 

You might want to increase the sub count to 10 which would reduce the effect on AF timing by 1/2.

 

Each sub then counts for less when calculating whether the trigger fires. 

That's what my instinct was telling me, but I initially went with the NINA defaults.  Next run, I'll do that.  Thanks!

[UP4014Fan]

IMO floating in RA or DEC isn't good.  Being just a little off in polar alignment is best.  

I've read that the slight unbalance should be opposite of where the scope is pointed.  How does one handle a meridian flip in that case?

[PIEJr]

I've read that the slight unbalance should be opposite of where the scope is pointed.  How does one handle a meridian flip in that case?

The imbalance (strictly from a mechanical viewpoint) causes a preload on the drives so while logic wants perfect balance, running "East Heavy" allows a preloaded condition on your mounts drive gearing.

A mechanical advantage to avoid a floating unbalance. Make sense? (Links for clarification.)

 

Incidentally, some of this terminology is confusing, to me, because in outer space there are no compass marks. (If you "Turn Left at Orion", is that E? W? N? or S?) Brain teaser....

 

How is it possible to have East Heavy when the telescope's weight offset flops the other way? (Meridian Flip)

Plate solving (I still like ASTAP myself) will adjust for any "flop" in the mount and telescopes error tendencies.

While I fancy myself Galilean with my refractor preference, I am glad I'm not in his place in history. I like my A.I. if you will. 

 

Anyway, a preload is desirable from a mechanical point of view of your mount.

And then, the clouds come, the wind blows, the Earth quakes, and we are left scratching our heads at what went wrong.   

[UP4014Fan]

The imbalance (strictly from a mechanical viewpoint) causes a preload on the drives so while logic wants perfect balance, running "East Heavy" allows a preloaded condition on your mounts drive gearing.

A mechanical advantage to avoid a floating unbalance. Make sense? (Links for clarification.)

 

Incidentally, some of this terminology is confusing, to me, because in outer space there are no compass marks. (If you "Turn Left at Orion", is that E? W? N? or S?) Brain teaser....

 

How is it possible to have East Heavy when the telescope's weight offset flops the other way? (Meridian Flip)

Plate solving (I still like ASTAP myself) will adjust for any "flop" in the mount and telescopes error tendencies.

While I fancy myself Galilean with my refractor preference, I am glad I'm not in his place in history. I like my A.I. if you will. 

 

Anyway, a preload is desirable from a mechanical point of view of your mount.

And then, the clouds come, the wind blows, the Earth quakes, and we are left scratching our heads at what went wrong.   

Your explanation made perfect sense - it's like taking slack out of train.  I just wasn't sure which side the bias should be on and you and others helped.  Thinking about the rotation of the scope during the night, the east side makes sense, as the drive train is "lifting" the load as it heads towards the meridian. 

 

I biased the weights this evening, changed the trigger to 7% over 10 exposures and things seem to have stabilized.  The HFR trends have all been hanging around 2% - 3%.

[dx_ron]

A lot depends on which mount you are talking about. Mounts like the Losmandys and EQ6 apparently can benefit from east-heavy. The reason for 'east' is to keep the pre-loading in the same direction that the gear is constantly turning. So yeah, which way to make heavy changes after a meridian flip. If you search hard enough, there are several old threads about creative ways to do it, though many of them require manual intervention.

 

But some mounts do not need it anyway. iOptron mounts have spring-loaded worms and do fine with neutral balance. Strainwave mounts don't care what you do.

[fewayne]

Strainwave mounts don't care what you do.

The Honey Badgers of Astronomy.

[rgsalinger]

The idea that you can achieve perfect balance ignores that fact that you have friction/stiction/etc. even with the clutches open.  That means that, just because the mount doesn't move, it's not the case that it's in perfect balance. A motionless system is not necessarily a balanced system.

 

If your RA axis is floating (which I've never experienced) then I suggest adjusting the mesh rather than deliberately misbalancing the mount.

 

Note that this is different from offsetting the mount away from the pole so that all of the DEC (drift) corrections go in the same direction. When you do that, then DEC backlash will never cause poor guiding RMA. 

[UP4014Fan]

I’m talking backlash here. I think I’ve been able to associate the issue with dithering events.

[rgsalinger]

There can never be RA backlash as the RA axis is always moving. If I was getting poor RA guiding then I would use PEC and/or adjust the mesh or belt tension. Anything else is purely hit or miss and some of what gets posted here is dubious when it comes to guiding. 

[dx_ron]

I’m talking backlash here. I think I’ve been able to associate the issue with dithering events.

Do you have sufficient settling time built in to your dither settings? 

As for the original question of triggering autofocus - does NINA implement temperature-based adjustments? Not just "full autofocus run if temp has changed x degrees". For many (most?) scopes, temperature is the main driver of focus change and the relationship between temperature and focuser position is quite linear.

 

My AT130EDT, for example, shifts focus by 42 steps per degree C, as plotted over several autofocus runs through a couple of nights with re-focus triggered on 1° changes. I then told Ekos to move the focuser 21 steps for any 0.5° change, which keeps the need for full autofocus runs to a minimum. I run autofocus on any change of target, but only rarely during capture of any single target on a night. It's also handy for start-of-night focusing. If the focuser is at the position it ended at in the early-morning cool but needing to focus in warmer air the next evening, Ekos calculates a rational starting point.

[dx_ron]

There can never be RA backlash as the RA axis is always moving. If I was getting poor RA guiding then I would use PEC and/or adjust the mesh or belt tension. Anything else is purely hit or miss and some of what gets posted here is dubious when it comes to guiding. 

He mentioned dithering. I've not thought about it much, but a sufficiently large RA dither move could ask the mount to actually back up (as opposed to just slowing down, like a normal guide pulse). RA backlash could come into play there, maybe? Backlash would cause the RA move to be much smaller than requested, causing PHD to issue more RA pulses to try to get the guide star to its new position.

 

Hmm - more thinking. If the dithers are normal guide pulses, then they'd be at most at 1x sidereal - so maybe the above paragraph is a load of road apples.

Edited by dx_ron, 23 June 2025 - 01:46 PM.

[AstroFromHome]

5% is much too sensitive. Let is at its default 10%.

If you have an HFR of 1.7 or 1.8 you will not see visually in the final image. 10% is also better if there is some wind not to trigger every few images a new autofocus run for no reason.

 

The temp trigger I have set to two degrees.

 

Currently this is giving me two or three autofocus runs a night at the moment. That's more than enough.

[rgsalinger]

He mentioned dithering. I've not thought about it much, but a sufficiently large RA dither move could ask the mount to actually back up (as opposed to just slowing down, like a normal guide pulse). RA backlash could come into play there, maybe? Backlash would cause the RA move to be much smaller than requested, causing PHD to issue more RA pulses to try to get the guide star to its new position.

 

Hmm - more thinking. If the dithers are normal guide pulses, then they'd be at most at 1x sidereal - so maybe the above paragraph is a load of road apples

If you are not using an Astro-Physics mount then I know of no manufacturer that recommends anything other than .5x sidereal as the guide speed. While there are differences in how this is implemented, there is just no way for the RA axis to "back up". It just slows down or speeds up by the guide rate.

 

I am doubtful that dithering would be anything other than a random pulse guide command but I've never looked at a log detailed enough to tell me what's happening. That would be an excellent question to pose on the PHD forum. It would be "dumb" of them to do anything else but you nevr know................. 

Edited by rgsalinger, 23 June 2025 - 10:19 PM.

[PIEJr]

A lot depends on which mount you are talking about. Mounts like the Losmandys and EQ6 apparently can benefit from east-heavy. The reason for 'east' is to keep the pre-loading in the same direction that the gear is constantly turning. So yeah, which way to make heavy changes after a meridian flip. If you search hard enough, there are several old threads about creative ways to do it, though many of them require manual intervention.

 

But some mounts do not need it anyway. iOptron mounts have spring-loaded worms and do fine with neutral balance. Strainwave mounts don't care what you do.

Just to update your thinking, Ron, Losmandy has spring loaded worms in the drives.

I don't recall how many years I've had my GM811G, (2017? 2018?) but they were fairly new still when I got mine.

GM8 DEC mounted to a G11G RA.

To my way of thinking it is the best of both mounts put into one.

And my PA/guiding is consistently below 0.5. Typically in the low 0.30's lately. Tiny pinprick stars in my images.

Recently I weighed my AT130EDT as it sets in the clamp and it's 36.x pounds, on a mount rated for up to a 50 pound photographic instrument load. (I was kind of worried because I err towards the heavy side. For example, I have two 14" Losmandy D-bars, one under, the other over, to make sure my telescope is rigidly supported (no unnecessary flexture) why I say I err towards heavy. Heavy is stable.)

But the spring loaded worms definitely improve Periodic Error.

[rgsalinger]

Is there any actual "official" documentation about having a mount be "east heavy" to improve guiding performance? I get that there are people who report this. I've just never seen any document from SkyWatcher or Losmandy recommending it. 

 

What bothers me is that since the RA axis is always moving westward, once the mesh is in contact, it can never reverse itself. So, you'd have to overcome the inertia of the system and the power of the motor to move the system eastwards. I even asked Charles Bracken about it a while back. I got back that same meme - some people report that this improves their guiding. 

 

So, I don't recommend east heavy as a technique. Even worse, you have to reverse the mount to "west heavy" after a meridian flip. Just seems odd that this lives on. Maybe I'll borrow a G11 and see if it works some night. 

[UP4014Fan]

Well, for the moment, it looks like not enough settle time after dithering for my issue.  I've upped it and I've only had one autofocus run in over an hour.  BTW, it's direct guiding with a StarSense Autoguider, FWIW.

[PIEJr]

Is there any actual "official" documentation about having a mount be "east heavy" to improve guiding performance? I get that there are people who report this. I've just never seen any document from SkyWatcher or Losmandy recommending it. 

 

What bothers me is that since the RA axis is always moving westward, once the mesh is in contact, it can never reverse itself. So, you'd have to overcome the inertia of the system and the power of the motor to move the system eastwards. I even asked Charles Bracken about it a while back. I got back that same meme - some people report that this improves their guiding. 

 

So, I don't recommend east heavy as a technique. Even worse, you have to reverse the mount to "west heavy" after a meridian flip. Just seems odd that this lives on. Maybe I'll borrow a G11 and see if it works some night. 

I suppose that depends on what your definition is is, is.

Plenty of different discussions about it. Here's what I find in a few milliseconds.

[rgsalinger]

Nothing I would call authoritative but good try. You basically posted the results an AI repeating what others have asserted but not really proven and nothing more. 

 

I think that the argument is that when you decelerate the mount the mesh somehow comes undone. Then to prevent that (which only happens for a few milliseconds) you try to make gravitational force keep the mount meshed. I think that the confusion comes from the inevitable set of false positive results coupled to the fact that people confuse the guiding cadence with the duration (almost instantaneously) of the deceleration phase. 

 

I'm looking for something from someone like Roland Christen or Tom Bisque or even Doug George who might be able to explain this. I tried using Co-Pilot on this and it pointed me back to CN which I thought was interesting. It also said that east heavy would be useful if the mount "changed direction" which, of course, the RA axis never does when guiding, only when slewing.