Jumps in guiding

[Georg Albrecht]

Hi guys,

Lately I have seen guiding having a bit more difficulties on my widefield rig.
I have an Atlas II EQ-G mount (3 years old) caring a RedCat51 with a Player One Poseidon-C camera, a 2"x5 filter wheel, Falcon v2 electronic rotator, EAF, and a DeepSkyDad flip flat panel. Guiding is done through ZWO mini guide scope30/120mm with an ASI290MM mini.
The rig is quite well balanced (small rigs are a pain) and cable management is tight.
No feel of any backlash and the axes rotate smoothly. I adjusted all that 2 years back.
The rig is in use July to December in SoCal pretty every clear night and stored inside home during the rest of the year. Last year and before, guiding averaged around 0.5-0.7 but this year I can be happy if it guides around 0.9-1.1. With the resolution I have of 3.09"/pixel I am not too concerned about the total RMS but more about the max peak due to the jumps I get mostly on the RA axis.
Last night I ran a new calibration to have that showing in the PHD2 log, then a short Guiding Assistant run for seeing conditions. The rest was guiding while imaging. Short night interrupted by clouds...  :(

My question is how to address these spikes and get the mount back to it's lower RMS guiding it previously showed. Is it time to disassemble, degrease and get some fresh lubricant in there??? Or could there be other things to check first. Belt tension looks fine and I don't "feel" and backlash on both axes.

Thanks for your help and CS - Georg

https://openphdguiding.org/logs/dl/PHD2_logs_5oVq.zip

[Bruce Waddington]

It's difficult to know exactly what's going on, there probably isn't enough data here.  Here's a look at your 37 minute guiding session (RA in red, Dec in green):

You can see the spikes here, but notice how often both RA and Dec have correlated large excursions, in addition to the solitary Dec move at the right.  The Dec motor was idle at these times.  Before you start tearing things apart, I think you should look more carefully at your guiding assembly.  You're working with a very coarse guider image scale, 5 arc-sec/px so these "big" excursions of 3-4 arc-sec can be caused by a shift in the guiding assembly of only 9-12 microns, about 1/5 the thickness of a human hair.  Was there any window during the session?  These "flip-flat" gizmos can act as sails.  I don't think you can assume that seeing conditions were the same as last year especially during hot summer months.  Has anything changed in your near-scope environment that would produce more heat convection?  If you isolate only the short sections of the guiding session where there were no spikes, the total guiding RMS was no better than 0.7 arc-sec, so that puts you at the high end of your perceived guiding performance from a year ago.  Of course, it you keep your PHD2 guide logs, it would be a good idea to compare the current and year-old performance pictures to be sure you are looking at real numbers.

In order to convince yourself that the mount performance has degraded, you might want to spend a test session guiding only through the main scope with no imaging.  That will rule out mechanical error sources from the guiding assembly.  Better still, try to do the testing during a period when you have evidence that the seeing is better than average.  If you decide to look at the mount drive system, there are some fairly clear contributors to periodic error:

The most important one is probably the 3 arc-sec peak-peak error with a period of about 120 seconds.  The other peaks to the right are at 240 and probably 480 seconds.  By using fairly long guide camera exposures of 2.5 seconds, it will be more difficult for PHD2 to stay on top of the error component with the 120 second period.  You should try using the PHD2 PPEC algorithm for RA, giving it a period length of 120 seconds and disabling the 'auto-adjust' checkbox.  Doing these things can't hurt but I'm not convinced it will have much to do with the source of the error spikes.

Good luck,

Bruce

[Brian Valente]

>>> Was there any window during the session?

I think Bruce meant to say "wind" not "window"?

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/1ef2ee13-a937-47e8-8a3e-343fdd8d2e5bn%40googlegroups.com.

--

Brian 

Brian Valente

astronomy portfolio https://www.astrobin.com/users/bvalente/ 

[pollya...@gmail.com]

Hello Georg,

I see you've checked the 'Assume orthogano axes' box in the Calibration setup. I think you might be better with that left at the default, unchecked.

Cheers,

- Jack T

[Georg Albrecht]

Thanks Bruce for taking the time to answer.

If helpful, I can upload a log of a full night of imaging. I created this last shared log because all the others still in my folder, didn't show a calibration. The rig is unchanged from past years setup as is the imaging position in the courtyard. With regards to guide scope montage, I have attached an image of the entire setup. From inside my courtyard, I am well shielded from wind (Brian ;) ) and there was no wind that evening. Although the connection feels solid, I completely understand your hair thickness comparison 😀. But I don't think I can improve there much. Unfortunately, I cleared out all my logs this spring. So I can only base my comparisons on memory. 

You point out the exposure length I am using as difficult to address the 120sec period? Should I try a shorter exposure time to see what happens? My seeing here is most of the time average 3/5 sometimes 4/5, rarely 5/5 https://www.cleardarksky.com/c/RcMrLbObCAkey.html and I thought a longer exposure time would keep the mount from chasing. The other suggestion of yours to use PPEC: Is that useful for a rig that is rolled out/in every night? I thought that would be more something to be used when a rig is stationary? I am willing to try this out. Do I need to set the 120sec period somewhere in the PHD2 software? I am not at my PC at the moment to check. 

The other suggestion you had about using main scope. I would create a new profile in PHD2 for that with dark frames, then running a calibration and thereafter let the rig guide for a period of time. Correct? And then compare the logs. 

I also saw a comment by Jack T this morning about the setting "Assume orthogano axis" I will have to check in the software.

Much appreciated and thank you - Georg

--

[Georg Albrecht]

Hi Bruce,

Would this be the page I set the "Period Length" to 120 and uncheck "Auto-adjust periode". Once I start guiding, I assume this will take some period of time to record the needed information? Would I need to adjust anything else here?

Thanks for your help. Georg

[Brian Valente]

Georg

Yes that's the page. I would personally increase the predictive gain a bit to 70 to start

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CACVq1zE7x-UjQLhQHN%2BHvZQ-0jGVTGpqmEkj6TGN6ta%3DOp7WAQ%40mail.gmail.com.

[Georg Albrecht]

Thanks Brian,

Now I only need better weather so I can test...

Cheers - Georg

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i6Ki4q9QGk%2Bx%2Br4RPWeOZQH%3Dem1eYx9GLxw-gmpQAKQwQ%40mail.gmail.com.

[Bruce Waddington]

Hi Georg.  I think we should take a step back from this once you've done the easy things like using PPEC for RA and decreasing the exposure time a bit.  I think there's a lot to not like about your guiding setup but we need to ask the more important question: does it matter?  If you're imaging at over 3 arc-sec/px, your guiding requirements are probably pretty minimal.  So you could easily end up spending a bunch of time and potentially screwing things up just because you're "chasing the numbers" with respect to guiding statistics.  If you have to set up every night and your imaging time is limited, I think you should spend that valuable time getting images.  If you reach a point where your results are suffering because of elongated stars or whatever, then it might be worth the time to go after improving your setup.  Just my opinion, of course, but that's what I would do.

Bruce

[Georg Albrecht]

Thanks Bruce,

Well, last night turned out "interesting".
I switched to PPEC but for the first calibration I forgot to decrease exposure time. The second calibration in the file is with 1.5sec exposures.
I ran directly after calibration at a new sky position, a short (120+ sec) Guiding Assistant with backlash measuring. That's when I remembered to reduce exposure time and I repeated it all again.
I also checked on what Jack T had mentioned and disabled "Assume DEC orthogonal to RA" (don't remember why that was enabled - must have somehow clicked on that during a boring winter night)
Anyway, my backlash suddenly is completely bonkers ???? Let me guess > disabling  "Assume DEC orthogonal to RA"???
Also looking at the second calibration, I assume there is too much tension in the DEC movement where it receives pulses but is stuck and then jumps far in? 
The average guiding seemed lower, but I am sure I need to work on a few things. What would be the best approach? I have free time coming up, now we are going towards a full moon.

I also added one of my older logs from an imaging night beginning of the month that has a full night of info. And below I added screen shots of the Guide Assistant runs with backlash graph. The one from 08/26 is from the night I shared the first guide log.
The other 2 are from last night's run with PPEC, just after calibration.

https://openphdguiding.org/logs/dl/PHD2_logs_Ui2T.zip

I am looking forward to your comments and insights.
Thank you and best regards - Georg

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/9c4ca59a-24ec-4dbf-ab51-92bbdc7d5e09n%40googlegroups.com.

[Brian Valente]

Georg

What is the image scale you are imaging at? arcsec/pixel

 Your calibration (even with 'assume axis ortho' disabled) looks fine. I think the first report was bogus. Subsequent runs put your dec backlash around 40-70ms, and it seems fairly well managed by PHD2. Dec is not your constraint to guiding so I wouldn't be sweating this for now. 

The second calibration run (the one you stuck with) gave you two recommendations: improve your guidescope focus, and improve your polar alignment (not terrible but could be better). Not sure what you did with these. 

The guiding assistant showed some pretty significant jumps in Dec:  this axis is not moving, but you have jumps of over 2"in Dec, and some substantial ones in RA >1.5". I think your mount remains having mechanical problems, but read on...

You were able to achieve an overall RMS of around 0.8" - 0.9" on a fairly consistent basis and on both sides of pier. RA was a bit higher, but you could probably tame this a little more by increasing your RA PPEC algorithm's control gain to around 0.80

Aside from that, you probably have some mechanical improvements you should be looking at for RA. you didn't really run the guiding assistant for long enough to see what is going on, but from what I can see in the residual error, that 119 second is not a constraining factor. (btw the peak residual now is much lower but more like a 96 second period)

So you have what I would call a noisy mount. A lot of various smaller errors that when combined make RA a bit higher than Dec. I can't point to any one factor in this. The guidescope setup you have isn't great. it's possible the mediocre mounting setup and configuration may be causing some of this, you can investigate as you see fit. 

However, does it really matter to your images? I am guessing your image scale is much higher than 1". so even if you improve it a bit, you likely won't see much difference in the results.

you can chase this around some more, or you can get on with imaging and see how it goes. 

Brian

 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CACVq1zGojaCFFMTHeLG3H1x3hmoZNAUooOpz5F1xXjHa7a6%2BNg%40mail.gmail.com.

[Georg Albrecht]

Hi Brian,

Thank you for your time. 
The imaging scale on this rig is 3.1"/pixel / RedCat51 > Poseidon-C pro
The overall RMS I get is more than sufficient for that scale. My concern is all these peaks that jump above the 3" line. That is really what I want to tackle.

Now, let's address the elephant in the room > Guide scope
Sorry but I get a little chuckle when I read big boys comments such as "not great" set-up or "mediocre" etc. (I am not taking it personal nor should you 😉 )
I mean, there are probably hundreds or enve more out there with similar guide scopes and set-ups and some do read these posts.
So, what is bad about this set-up? The guide scope itself only having one support ring??? Should I look for a guide scope with minimum 2 rings?
The position of where the guide scope is mounted > too far from the main scope? At the end of the (freakingly) ridgid Losmandy plate?
Constructive insights would be more helpful for the entire community 😁.

My backlash that creeped up after removing "assume axis ortho" has been addressed. I made some mechanical adjustments in the DEC and it looks good now.
I will work on the RA this week while I have cloudy skies 😞 

I also had already increased (last night) the  RA PPEC algorithm's control gain to 0.80
Not sure about the seeing conditions last night. It was probabil only OK seeing - Overall RMS was fine but again, these repeated jumps producing several rejected images...

A few other things you mentioned:
I am not sure why at the 2nd calibration it recommended to check guide scope focus. The focus looked just fine.
Polar Alignment - Unfortunately Polaris is hiding away behind the roof line. I use NINA's 3PPA > home to 1st 3PPA (ALT 47°/AZ 58°> adjusting below 1' > home > home to 2nd 3PPA > adjusting below 0.5' > home >  home to 3rd 3PPA >  adjusting ends up below 0.2' > home. Thereafter I do a Guide Assistant (see below) run for 120-180sec at ALT 49° AZ 100°


I have attached logs for 08/30 and from last night.

I have 2 burning questions:
- Guiding Assistant runs - Where in the sky should they be done? Reading the Baseline Measurements, it says to run the GA directly after the calibration. So, does that mean the GA should be run at the same intersection calibrations are done?
- I read once more through the PHD2 manual and came across this -  If you’re following best practices, you will have programmed periodic error correction in your mount (assuming that feature is available to you)"
Somehow, I was under the impression that if a PEC training was done in the mount itself, PPEC with PHD2 should not be used at the same time??? Reading this, it looks more like it is recommended to do just that. 
As long as my rig is not broken down, a trained PEC in the mount should keep its value, right? My rig is on a wheely cart. This could be an additional step... Please advise. 

Thank you again for your time - CS - Georg

https://openphdguiding.org/logs/dl/PHD2_logs_Uhmy.zip

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i5xAUgtY85o%2Bhwbn_SNr4sYx4x%3DqnmJzFALb8RHn9OFeg%40mail.gmail.com.