Guiding Issues - Large Error in R.A. Axis

[lrt75914]

Hi guys,

I've been having a lot of issues with the R.A. guiding of my Skywatcher AZ-EQ 6 mount and I can't
seem to isolate the source of the problem. This PHD2 (unguided) graph and guiding assistent was
last week while I was experimenting with my mount. I have also uploaded the PHD 2 log file in case
anyone is interested.

 

And here is the equipment I am using:

Equipment:

Skywatcher Esprit 100
ZWO ASI 1600 MM C (imaging)
ZWO EFW Mini
Starizona Microtouch Focuser
Borg 50mm f/5 guide scope (incl borg helical focuser)
ZWO ASI 174 MM (guiding)
Skywatcher AZ-EQ 6

Software:

Sequence Generator Pro
PHD 2

 

I have already tried several measures to improve the guiding performance. The R.A. and Dec axis were striped down and regreased to reduce friction. While doing so I found that
a seized grub screw, securing the R.A. axis in the aluminium block, had been haphazardly sanded down - I'm guessing that the worm gear had been hammered into its socked to
clear the grub screw. Said screw was properly screwed down.

I have also tried adjusting the worm gear play, east heavy balancing, changing PHD2 settings  - both for the hysteresis and ppec algorithm, using the PHD2 guiding assistent and its reccomendations etc.
Since nothing has helped so far, I'm starting to wonder if the R.A. worm gear itself is damaged and needs replacing. Another issue I've found is that the helical focuser on my Borg Guide has
a lot of play. However, I'm not sure if this is the root cause of the problem since the guiding issue is only in the R.A. axis.

 

I was wondering if anybody here can give me any pointers and tell me why my R.A. guiding is this bad.

 

Thank you.

Patrick

PHD2_GuideLog_2020-01-01_222712.txt

Edited January 8, 2020 by lrt75914

[wimvb]

Just to rule this out: was making corrections turned on? The plot should show phd2 trying to make corrections, and the checkbox to show corrections is ticked.

[wimvb]

Just looked at the log: you turned off guiding at 22.52 to run guiding assistant, before that guiding was actually very good. What you see during GA is the mount's periodic error in RA and some DEC drift due to polar misalignment (only 1 '). This seems to be guided out very well.

Excluding the GA period, your RMS error is 1.38" or 0.29 px, about 1.1" in RA and 0.76" in DEC.  MinMo is 1.1" in RA, and 0.86" in DEC (0.24 px and 0.18 px). As long as tracking deviation is smaller than MinMo, PHD won't apply any corrections. I don't think you can expect a guide RMS smaller than the MinMo settings.

To improve the periodic error, you would have to hypertune the mount. You could try PPEC, but since most PE is guided out, I'm not sure you'll see much improvement from that.

With your camera and scope (3.8 um pixels and 550 mm fl) you get an imaging pixel scale of 1.2"/px, assuming the 0.85x FR. Your guiding pixel scale is 4.8 or four times as much. You could try to reduce that by using a longer fl guide scope or a different camera (ASI120?) With the ASI174 as guide camera you could also consider using an oag.

Edited January 8, 2020 by wimvb

[lrt75914]

2 hours ago, wimvb said:

Just to rule this out: was making corrections turned on? The plot should show phd2 trying to make corrections, and the checkbox to show corrections is ticked.

I only saved an image of the graph once the corrections were turned off. Having unearthed the production mishap
in the R.A. axis I was starting to get worried that the gears themselves were bend. I figured that the best way to
visualize any worm gear problems would be to record an unguided graph. I have to be honest though, half the
time I don't know what I'm doing^^.

2 hours ago, wimvb said:

Just looked at the log: you turned off guiding at 22.52 to run guiding assistant, before that guiding was actually very good. What you see during GA is the mount's periodic error in RA and some DEC drift due to polar misalignment (only 1 '). This seems to be guided out very well.

Excluding the GA period, your RMS error is 1.38" or 0.29 px, about 1.1" in RA and 0.76" in DEC.  MinMo is 1.1" in RA, and 0.86" in DEC (0.24 px and 0.18 px). As long as tracking deviation is smaller than MinMo, PHD won't apply any corrections. I don't think you can expect a guide RMS smaller than the MinMo settings.

To improve the periodic error, you would have to hypertune the mount. You could try PPEC, but since most PE is guided out, I'm not sure you'll see much improvement from that.

With your camera and scope (3.8 um pixels and 550 mm fl) you get an imaging pixel scale of 1.2"/px, assuming the 0.85x FR. Your guiding pixel scale is 4.8 or four times as much. You could try to reduce that by using a longer fl guide scope or a different camera (ASI120?) With the ASI174 as guide camera you could also consider using an oag.

I had lowered the MinMo values without much difference to the guiding performance. I did not, however, consider the different pixel scales between the imaging and
guiding scope. I guess that's what you get for buying a guiding camera that could possibly double as a planetary camera in the future.

The stars in my pictures are a bit elongated due to the R.A. guiding error (see horse head picture), so I have to reduce this error somehow. Would switching to another guide
camera (e.g. ZWO ASI 178 MM) be sufficient to achieve round stars or is it impossible to tell in advance?

 

Thank you very much for your help so far!

Cheers,

Patrick

Edited January 8, 2020 by lrt75914

[wimvb]

1 minute ago, lrt75914 said:

I only saved an image of the graph once the corrections were turned off. Having unearthed the production mishap
in the R.A. axis I was starting to get worried that the gears themselves were bend. I figured that the best way to
visualize any worm gear problems would be to record an unguided graph. I have to be honest though, half the
time I don't know what I'm doing^^.

I had lowered the MinMo values without much difference to the guiding performance. I did not, however, consider the different pixel scales between the imaging and
guiding scope. I guess that's what you get for buying a guiding camera that could possibly double as a planetary camera in the future.

Combined with a longer fl, your guide camera is ok. I use the cooled version as my main imaging camera, but also a 1000 mm scope. To get a better match between guide camera and imaging camera, the pixel scales shouldn't differ much more than a factor of 2. With 1.2"/pixel for your imaging camera, a pixel scale of about 2.4"/px would be better for your guide camera.

If you use an oag, the match between guide camera and imaging camera will be much better. Oags are not that difficult to set up, and since you already have ZWO imaging and guide cameras, there won't be a distance problem. The ZWO oag is a match to your cameras.

In general, you can't expect better than a few tenths of a pixel guide RMS. At 4.8"/px, 0.2 pixels is almost 1". That's why I think your guiding is good. Also consider that guiding quality is determined by other factors than the mount, eg seeing, balance, etc. With my AZ-EQ6, I get guiding RMS from 0.6 - 1", mainly depending on sky quality (seeing, transparency).

[lrt75914]

On 08/01/2020 at 23:30, wimvb said:

Combined with a longer fl, your guide camera is ok. I use the cooled version as my main imaging camera, but also a 1000 mm scope. To get a better match between guide camera and imaging camera, the pixel scales shouldn't differ much more than a factor of 2. With 1.2"/pixel for your imaging camera, a pixel scale of about 2.4"/px would be better for your guide camera.

If you use an oag, the match between guide camera and imaging camera will be much better. Oags are not that difficult to set up, and since you already have ZWO imaging and guide cameras, there won't be a distance problem. The ZWO oag is a match to your cameras.

In general, you can't expect better than a few tenths of a pixel guide RMS. At 4.8"/px, 0.2 pixels is almost 1". That's why I think your guiding is good. Also consider that guiding quality is determined by other factors than the mount, eg seeing, balance, etc. With my AZ-EQ6, I get guiding RMS from 0.6 - 1", mainly depending on sky quality (seeing, transparency).

I wasn't sure if I could set up a OAG in a timely manner, so I got my hands on a ZWO ASI 178 MM instead - just to check if a better match between my guiding/imaging setup would do the trick. Pixel scale was 1.42" and 1.98" for the imaging and guiding rig respectively. I guess this is as close as I can possibly get the two without buying a different guide scope.

Unfortunately the results aren't all that encouraging.

I used the standard parameters as well as the suggestions from the guiding assistant. I also gave the PPEC algorithm a go but I still have trailing stars in my images. I tried aggressive east bias balancing, fiddled about with all sorts of parameters available to me, but the results remain unsatisfying. I'm starting to wonder if I'm to stupid for this hobby.

I*m seriously considering selling my gear and turning my back on astronomy. I only have a handful of nights each year that are good enough
for imaging and all I'm doing is dicking around trying to fix mechanical and electrical issues that keep popping up.

P.S.: Seeing and transparency were good tonight.

Edited January 17, 2020 by lrt75914

[wimvb]

The best way to analyze guide logs is with phd guide log viewer.

https://openphdguiding.org/phd2-log-viewer/

It can show you if you have an oscillation, such as pec, or a gear acting up. You should analyze th ra behaviour in the second graph. This graph seems to show some sort of mechanical issue, but the third graph may very well just show poor seeing.

The phd forum is also a good place to get help.

Also, be aware that phd won't apply corrections if the deviation of the guide star is less than MinMo. That's why the second graph shows so few corrections in dec. The random movements in the third graph may, as I already said, be caused by seeing. I probably would keep using those settings. The final rms is about 1 pixel on the imaging camera, not too bad at all. If you use the settings next time, you'll develop a sense for how atmospheric conditions affect guiding. 

One phd tool that I find helpful, is the bulls eye. If ra and dec movements are truly random, the errors will show a round "cloud" in the center. And the better your guiding is, the tighter that cloud will be. But if ra and dec movements follow each other, that cloud will become a streak. The stars in each sub will show the same pattern as the bulls eye. If they don't, you probably have flexure somewhere. Especially if the bulls eye shows a round cloud, but the stars are elongated. In that case, there should also be a constant drift between subs. You'll notice that during stacking. 

Edited January 17, 2020 by wimvb

[michael8554]

You seem to have good PA and not much Dec Backlash, so Dec should hardly need guiding at all - try balancing slightly heavy in one Dec direction, upping the MinMo to 0.3 or 0.4, and see if you get a straight Dec line !

The PPEC seems to work well with your mount, but be aware it has to go through a couple of worm cycles each night to learn the Periodic Error, so don't start imaging until it been chugging along for say 10 minutes.

Finally I would say 3 secs exposure is too long, come back to 2 secs to allow PHD2 to catch the RA drift more quickly.

Michael

[wimvb]

7 hours ago, lrt75914 said:

I*m seriously considering selling my gear and turning my back on astronomy. I only have a handful of nights each year that are good enough
for imaging and all I'm doing is dicking around trying to fix mechanical and electrical issues that keep popping up.

Unfortunately, ap has a learning curve, and the European climate isn't supportive. But you have a few options.

1. High gain short exposure imaging. Not quite lucky imaging, where you try to beat seeing, but rather try to beat guiding. You'll end up with lots more subs, and probably need to throw many away, butwith cooled cmos, this is an option. I've done it on occasion when the guiding gods were against me.

2. Go for a shorter focal length with a light weight refractor or tele lens where guiding is less critical. I've done some of my better images with a pentax 135 mm lens. When you've worked out the wrinkles in your guiding, you can go back to a longer focal length reflector.

3. Live with guiding rms slightly above 1", and enjoy those rare nights when it all just works, and rms suddenly touches 0.6".

A friend of mine with a high end mount gets guiding rms of 1" more often than he would like, but he still manages to produce excellent images. Just not of super detailed galaxies. On those nights of good guiding, you make the luminance filter your best friend. On other nights you go for colour with rgb filters.

[knobby]

Just relax and enjoy your images, don't be hyper self critical ... Stars look fine to me !

Keep at it and enjoy the end result 👍

[newbie alert]

Granted that the RA looks not very good but can't see where the stars are elongated in the image, it's a great image..

But seems like your problem could stem from the worm, it could be  serious backlash from the gears, it could be a over tightened locking nut on the RA shaft.. give the counter weight bar a good wiggle.. if you see or feel movement then that's what it is..

Things to check is the worm adjustment, not too sure on a AZ-EQ6 but where the bearings are on the worm spindle is the locking nut tight, but not too tight to lock the worm up, has the threads been loctight'd, has the locking ring been over tightened on the RA shaft..are you over tightening the clutch lever..all of this could have a impact on the performance of the mount 

[MarcusH]

On ‎08‎/‎01‎/‎2020 at 15:49, lrt75914 said:

Another issue I've found is that the helical focuser on my Borg Guide has
a lot of play.

Try to eliminate this if possible. Any slop, tilt or play especially in your guiding gear is going to cause you a lot of headaches. Your rig looks solid and well built with a payload that shouldn't cause any problems to the AZ-EQ6 whatsoever. It's nothing like the behemoth C11 I'm punishing my mount with (I have the same AZ-EQ6).

Those guiding graphs look much like mine did last season. I kept reading that this mount was capable of sub-arcsec tracking with guiding and yet I was struggling to get it under 2" RMS.... Nothing I did seemed to improve things, and more often than not I didn't even know what I was doing. I too was ready to call it a day after the first season, sell my gear and do something else as I obviously was too stupid for this hobby....

Luckily we have this mandatory break up here in the north when any form of imaging is out of the question (apart from solar, obviously), no astronomical darkness for 4 months. That gave me time to put things in perspective and as you have done, I also stripped the mount down and did a regrease of the bearings. While I had the mount in pieces I decided to swap out the Chinese worm bearings for a good set of SKF bearings (nothing fancy as some super duper hybrid ceramic bearings, as I cannot see why those would be superior in a ultra-slow rpm application like this).

I  spent quite some time on trying to reduce the backlash between the worm and the worm gear. I really don't know how much backlash is too much, but when I run the guiding assistant I rarely get anything >200ms, more often than not I've been getting the backlash-so-small-that-nothing-needs-to-be-done message. So at 915ms I feel that there still would be room for at least a little improvement.  I also spent numerous nights when it was not dark enough to image but plenty dark to test guiding in doing PEC training. And when I finally was satisfied I made it permanent (PPEC).

The result? I went from the-best-I-can-hope-for-tonight-is >2" to a somewhat solid 0.6" - 1.0" RMS error. But I also have to stress that this was not due to the diy "hypertune" alone. I also perfected my polar alignment, mount leveling, balancing and my whole workflow during those semi-dark nights. Anything and everything you can improve on is going to help you in the long run.

We also have to remind ourselves that even tough this is a capable little mount (AZ-EQ6), it still is (at best) a mid-range mass-produced mount with all the peculiarities and quirks that go into that territory. The RMS error range of 0.6"-1.0" that also @wimvb  quoted is realistically the best we can hope for with repeatability, with a few outliers in the <0.5" and >1.0" range. My own personal best was a 5 hour session where the RMS error hovered at 0.38"-0.56" but I'm going to put that down as a once-in-a-lifetime occurrence. And what might work out for me, doesn't necessarily work for you.

 

On ‎17‎/‎01‎/‎2020 at 01:48, lrt75914 said:

I only have a handful of nights each year that are good enough
for imaging and all I'm doing is dicking around trying to fix mechanical and electrical issues that keep popping up.

It's hard to tell from the picture of your rig if it is a permanent setup or if you set it up for every session. But if you set it up every session, try to make it a habit to set it up even when the forecast is so poor that you are barely able to get it polar aligned. Set it up, get everything going, tear it back down. Rinse and repeat. Try to make it second nature to set it up so you don't even have to think what you are doing.

I did this as a part of perfecting my workflow. And I did it so many times that I now can set up my rig in pitch black darkness with my eyes blindfolded and one hand tied behind my back...😁 I went from a totally ridiculous and mind boggling 4-5 hour setup time to 15-20 minutes. And even though my normal setup consists of 3 PC:s, an imaging camera and focuser all connected wirelessly + a myriad of USB connected equipment, I have yet to loose a single sub this season due to equipment acting up (fingers crossed...). Those few precious nights that are perfect for imaging is now spent imaging, not troubleshooting.

 

And as far as your images go, sure, you can beat yourself up by pixel-peeping them and trying to find all sorts of imperfections. Or you could take step back, zoom out and look at them for what they are - stunning examples of something incredibly beautiful. I'd be inclined to say that you give people with twice your budget and 100x more clear nights a run for their money.👍

Sorry for this long winded, mindless and pointless rant, but hopefully you can find something hidden in here that keeps you hooked to this brutally unforgiving yet fascinating hobby.

[geeklee]

Great post @MarcusH

Some really good info in there. 

As a beginner, I can relate to the preparation steps and workflow. I've worked toward reducing setup time and thinking of ways to make it easier/more efficient. 

[wimvb]

1 hour ago, MarcusH said:

 I also spent numerous nights when it was not dark enough to image but plenty dark to test guiding in doing PEC training. And when I finally was satisfied I made it permanent (PPEC).

Can you elaborate on that (or pm me). I've tried doing ppec on this mount, but never succeeded.

[dannybgoode]

@lrt75914  I’m new to AP and have spent the last year just getting cross and missing while nights of clear skies, like you, just dicking about with kit that just doesn’t bloody work. 
 

But slowly and surely it’s coming together and whilst I still have a long way to go I’m starting to get some results. I have spent a long time working things out in the warm and dry but that isn’t feasible with things like guiding so it’s still a bit hit and miss. 
 

Similarly I often think of money I have tied up in my gear and wonder if I’d be better freeing it up but then I know I’d miss it. 

[MarcusH]

7 hours ago, wimvb said:

Can you elaborate on that (or pm me). I've tried doing ppec on this mount, but never succeeded.

As the whole concept of periodic error correction was new for me and still is something of a dark art I relied heavily on info found on the net. I went the free software route (i.e. PECPrep).

Essentially what I did was:

1. Used the AutoPEC in EQASCOM to record a PE curve (I used 9 worm cycles)

2. Imported the PE curve into PECPrep and used it to analyze, tweak and create a PEC curve.

3. Wrote the created PEC curve back to the mount

After that I used the method described by user AhBok over at Cloudy Nights to store the PEC curve permanently to the mount (the summary is in post #14):

https://www.cloudynights.com/topic/631100-a-well-averaged-smoothed-ppec-curve-with-eq6-r-pro/

I had to reread that post a few times before it sunk in, but it was pretty simple in the end. The good part is that you can store the PPEC into the mounts motor controller anytime, you don't have to do it at night.

The importing to PECPrep is optional and can be used if you want to look at or tweak the PE curve. AhBok got some pretty impressive results by only using the AutoPEC feature of EQASCOM.

[Thommy]

I always use exposure times between 3 and 5 s with my EQ8. If I use 2 s, guide stars seem to move randomly indicating that I try to chase the seeing. This is always resulting in inferior guiding. In your plots above, it seems that the corrections for the plot with 3 s exposure are very small. As far as I can tell, the problem with using 3 s or more is not the mount, but that corrections are insufficient. It is a bit strange since your agressiveness setting are quite insisting. What is the guide speed setting for your mount?

A few things to check: Is the orientation of your guide camera correct? Is there a slight imbalance of your mount to the east? Do you have a good guide calibration?

 

/Thommy   

[ollypenrice]

Your rig looks nicely put together. I'd have thought you could contrive a device to put a little upward pressure on your guide cam so as to eliminate the movement in the guide focuser. It certainly does need to be eliminated.

I'd also try much shorter guide subs. I sometimes used to use 0.5 sec on our EQ sixes.

Olly

Edited January 27, 2020 by ollypenrice

[Mgarsal]

Pardon my late comment, just stumbled upon this about 2 months old thread wondering how you were doing, hope you are still hooked to the hobby!. Wanted to share I  just had a similar issue of irregular tracking in my EQ-G ( I understand basically the same mount save for color and AZ variation?), irregular tracking ultimately due to the CAP screws and set screws securing the RA/DEC axis and worms becoming loose..... took me 10 minutes to fix it!.  My setup is not the same as yours ( I use an OAG) but now I'm back to sub arc second guided tracking.  I love my EQ-G ( my 3rd mount) and as long as I can still hook it up to a computer I won't ever sell it.

Details here:

https://www.cloudynights.com/topic/698093-beloved-atlas-has-the-flu/

Hope this helps anyone!.