Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

RA spikes during guiding with a new iOptron CEM40EC

  • Please log in to reply
61 replies to this topic

#51 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 26 July 2019 - 03:41 AM

Gday Tim

 

That is exactly what I'm saying. I think that there is logic in the mount firmware that makes decisions about when a guider correction is permitted or rejected

The only thing i see wrong with that theory is why does a reconstituted curve have such a repeatable sinusoidal error at exactly the worm period.

 

In theory, PHD was keeping the star at a fixed position on the sensor and as such its commands to guide had to be proportional to the measured error. For it to be so consistent in shape indicates to me that the error in tracking is probably real. Still all hypothetical guesses tho.

Andrew Johansen Melbourne Australia



#52 spokeshave

spokeshave

    Vanguard

  • *****
  • Posts: 2033
  • Joined: 08 Apr 2015

Posted 26 July 2019 - 03:58 AM

Gday Tim
The only thing i see wrong with that theory is why does a reconstituted curve have such a repeatable sinusoidal error at exactly the worm period.
 
In theory, PHD was keeping the star at a fixed position on the sensor and as such its commands to guide had to be proportional to the measured error. For it to be so consistent in shape indicates to me that the error in tracking is probably real. Still all hypothetical guesses tho.
Andrew Johansen Melbourne Australia


I can only comment on how my mount behaves, which is much more like that in post #32, where the inferred drift appears uniform and in one direction. The reason that the drift appears to suddenly change direction can be explained by which side of zero the RA trace happens to be. If the mount truly does ignore some RA corrections, it would do so whether the trace was above of below the center line, and the direction of the inferred drift would depend on which side of zero the trace happens to be. Additionally, the position of the trace with respect to the center line can change due to random factors such as post-dithering position, wind, etc. Regardless of which side of zero the trace ends up on due to these factors, the encoder keeps it there and the ignored corrections give the false appearance of drift.

I can't explain the data seen in post #26. That appears to be a completely different issue in which the encoder simply isn't working. I think you'll agree that what you see in post #26 and #32 seem to be completely different.

Tim

#53 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 26 July 2019 - 04:28 AM

Gday Tim

I think you'll agree that what you see in post #26 and #32 seem to be completely different

Yes and no :-)

Post 32 shows 4 very distinct regions of drift, but if you break out each section, and analyse it with the drift removed, you still get a sinusoidal error overlaid, and it roughly matches what post 26 shows.

Ref attached plot of the first section of post 32 deconstructed

I cant explain it yet, but the fact the sinusoid matches the worm freq is too neat to be coincidence.

The reason that the drift appears to suddenly change direction can be explained by which side of zero the RA trace happens to be.

Dont understand??? In all these traces, the RA position is plotted relative to the first point, ie whether it is above or below zero on the graph is a function of the first point.

Additionally, the position of the trace with respect to the center line can change due to random factors such as post-dithering position, wind, etc.

Agreed, but he didnt have dithering going and wind effects would only cover a few seconds, not the thousands of seconds he had.

Andrew Johansen Melbourne Australia

First section.jpg


Edited by OzAndrewJ, 26 July 2019 - 04:35 AM.


#54 Sasho_Panov

Sasho_Panov

    Explorer 1

  • -----
  • topic starter
  • Posts: 68
  • Joined: 19 Aug 2017

Posted 31 July 2019 - 07:17 PM

Gday Tim

Yes and no :-)

Post 32 shows 4 very distinct regions of drift, but if you break out each section, and analyse it with the drift removed, you still get a sinusoidal error overlaid, and it roughly matches what post 26 shows.

Ref attached plot of the first section of post 32 deconstructed

I cant explain it yet, but the fact the sinusoid matches the worm freq is too neat to be coincidence.

Dont understand??? In all these traces, the RA position is plotted relative to the first point, ie whether it is above or below zero on the graph is a function of the first point.

Agreed, but he didnt have dithering going and wind effects would only cover a few seconds, not the thousands of seconds he had.

Andrew Johansen Melbourne Australia

attachicon.gif First section.jpg

Hi Andrew,

 

Tonight, I had finally succeeded to perform the guiding test as you described it in your previous tests.

 

The guiding log is attached and the DebugLog is too large to attach, so if you like, I could send you by WeTransfer or similar way.

 

Please consider that I am located in the very center of the city with heavy light pollution and this night the seeing was bad with occasional wind gusts.

 

The PHD2 log contains first calibration and subsequent four short attempts to guide that I terminated sine were not satisfactory.

 

After the second calibration and short guiding to verify that it is OK, I have selected a target and followed your instructions.

 

The target (M26) was near the celestial Equator (Dec -9 Degrees).

 

After app. 30 minutes of guiding, I have disabled it in PHD2 advanced settings for 30 minutes and after that, enabled it again for 30 minutes.

 

Thank you in advance for your invaluable help and support.

 

Best regards,

 

Sasho

Attached Files


Edited by Sasho_Panov, 31 July 2019 - 07:29 PM.


#55 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 31 July 2019 - 08:38 PM

Gday Sasho

the DebugLog is too large to attach

No bother for me, i dont understand it :-)

After the second calibration and short guiding to verify that it is OK, I have selected a target and followed your instructions.

Yep, looks good with a few questions

Using the ripples in the DEC trace as a rough guide of seeing, the RA doesnt show any remarkable spikes or oscillations. That said, the guider was working overtime in RA, but it wasnt heavily unidirectional this time???

Once more i plotted your raw data and overlaid the RA pulses ( at an expanded scale )

You can clearly see from the guide pulses that the direction of drift it was fighting changed at the worm period, so once more i loaded it into PEMPro viewer and reengineered what the unguided trace would have looked like ( assuming the pulses were real )

This was very illuminating

From the first plot we can see when you stopped guiding, the RA started to drift down at a slow rate.

Something weird happened to DEC part way though that drifting time, no idea there.

From the reconstituted plot, you can see it begins with what "appears" to be the trace of a raw worm drive with drift

The unguided section then reverts to a flat line of the same drift rate, then back to a sinusoid with drift????

ie once more, it looks like when you were guiding, the encoder wasnt keeping anything under control???

but the guiding itself was???

Unguided, you had drift but the encoder was removing the PE

( The FFT of the guided sections once more shows a 42x fundamental, but it is small compared to the rest of the data. )

Edit

Just added a third plot where i cut yr log up ( a bit )

The third plot shows the raw RA from the first section

then the unguided section, but with drift removed.

I cant see anything unusual as the guiding stops, as if there is something there, its hidden in the seeing

or occurs between frames.

 

Andrew Johansen Melbourne Australia

GuideUnguide.jpg

GuideUnguide decon.jpg

RA compared.jpg


Edited by OzAndrewJ, 31 July 2019 - 08:59 PM.


#56 Sasho_Panov

Sasho_Panov

    Explorer 1

  • -----
  • topic starter
  • Posts: 68
  • Joined: 19 Aug 2017

Posted 01 August 2019 - 08:58 AM

Gday Sasho

No bother for me, i dont understand it :-)

Yep, looks good with a few questions

Using the ripples in the DEC trace as a rough guide of seeing, the RA doesnt show any remarkable spikes or oscillations. That said, the guider was working overtime in RA, but it wasnt heavily unidirectional this time???

Once more i plotted your raw data and overlaid the RA pulses ( at an expanded scale )

You can clearly see from the guide pulses that the direction of drift it was fighting changed at the worm period, so once more i loaded it into PEMPro viewer and reengineered what the unguided trace would have looked like ( assuming the pulses were real )

This was very illuminating

From the first plot we can see when you stopped guiding, the RA started to drift down at a slow rate.

Something weird happened to DEC part way though that drifting time, no idea there.

From the reconstituted plot, you can see it begins with what "appears" to be the trace of a raw worm drive with drift

The unguided section then reverts to a flat line of the same drift rate, then back to a sinusoid with drift????

ie once more, it looks like when you were guiding, the encoder wasnt keeping anything under control???

but the guiding itself was???

Unguided, you had drift but the encoder was removing the PE

( The FFT of the guided sections once more shows a 42x fundamental, but it is small compared to the rest of the data. )

Edit

Just added a third plot where i cut yr log up ( a bit )

The third plot shows the raw RA from the first section

then the unguided section, but with drift removed.

I cant see anything unusual as the guiding stops, as if there is something there, its hidden in the seeing

or occurs between frames.

 

Andrew Johansen Melbourne Australia

attachicon.gif GuideUnguide.jpg

attachicon.gif GuideUnguide decon.jpg

attachicon.gif RA compared.jpg

Hi Andrew,

 

Thank you very much for your efforts and time. 

 

Thus this means that future firmware updates from iOptron could possibly led to a better guiding curves?

 

 

Meanwhile, is there some adjustments I could set to improve the guiding?

 

All the best,

 

Sasho



#57 gotak

gotak

    Surveyor 1

  • *****
  • Posts: 1996
  • Joined: 18 Sep 2016
  • Loc: Toronto, CA

Posted 01 August 2019 - 09:21 AM

When is Andrew going to get one of these to get really dug in?

#58 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 01 August 2019 - 05:08 PM

Gday Sasho

Thus this means that future firmware updates from iOptron could possibly led to a better guiding curves?

I would certainly hope so

The main thing your plot shows ( as you have guided and unguided data in one set )

is that the theory that the mount may be blocking some guides is probably wrong.

In all cases when you are guiding ( at different guide frequencies ) you get the same reconstituted shape.

ie ( to me ) it indicates that even at a 3sec rate like your last data, whilst guiding, the encoder doesnt appear to be getting used???

 

 

Meanwhile, is there some adjustments I could set to improve the guiding?

No idea. You could try a much longer guide rate, but what is really needed is a way to determine whats really going on and what triggers it. That probably requires bench testing ( or access to internal microstep/encoder data so it can be plotted accurately over time . That can easily be done with the Syntas or Meades, but no idea what data IOptron exposes to end users )

 

Andrew Johansen Melbourne Australia



#59 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 01 August 2019 - 05:15 PM

Gday Gotak

 

When is Andrew going to get one of these to get really dug in?

Need to win the lotto first. ( prices for stuff down here is too high to just buy stuff "for fun", esp based on the no of clear nights we get now )

Also, I already have enough to do playing with the Meade and Synta firmwares for mounts i currently own.

This is still just a learning exercise in how encoders work, and why just mentioning something has an encoder doesnt mean it will work properly and consistently.

As i mentioned in many threads very early on, you can't beat getting lots of good data from lots of different sources.

 

Andrew Johansen Melbourne Australia



#60 Sasho_Panov

Sasho_Panov

    Explorer 1

  • -----
  • topic starter
  • Posts: 68
  • Joined: 19 Aug 2017

Posted 01 August 2019 - 05:39 PM

Gday Sasho

I would certainly hope so

The main thing your plot shows ( as you have guided and unguided data in one set )

is that the theory that the mount may be blocking some guides is probably wrong.

In all cases when you are guiding ( at different guide frequencies ) you get the same reconstituted shape.

ie ( to me ) it indicates that even at a 3sec rate like your last data, whilst guiding, the encoder doesnt appear to be getting used???

 

No idea. You could try a much longer guide rate, but what is really needed is a way to determine whats really going on and what triggers it. That probably requires bench testing ( or access to internal microstep/encoder data so it can be plotted accurately over time . That can easily be done with the Syntas or Meades, but no idea what data IOptron exposes to end users )

 

Andrew Johansen Melbourne Australia

Hi Andrew,

 

Once again, thank you very much. I hope the thinks will clarify soon.

 

Regards,

 

Sasho



#61 OzAndrewJ

OzAndrewJ

    Mercury-Atlas

  • *****
  • Posts: 2819
  • Joined: 30 Nov 2010

Posted 01 August 2019 - 08:38 PM

Gday Sasho

 

I hope the thinks will clarify soon

Based on the other threads going re the 60 and 120, it sounds like the manufacturers know about it :-)

At least with your mount, it does appear to guide without the strange spikes/oscillations etc seen in some of the other mounts. Based on yr comment re seeing an pollution, the guided scatter plot isnt too bad.

 

Andrew Johansen Melbourne Australia

First 30 mins raw x/y ( at 1.8 arcsec/pix )

Guided Scatter.jpg



#62 Sasho_Panov

Sasho_Panov

    Explorer 1

  • -----
  • topic starter
  • Posts: 68
  • Joined: 19 Aug 2017

Posted 02 August 2019 - 07:02 AM

Gday Sasho

Based on the other threads going re the 60 and 120, it sounds like the manufacturers know about it :-)

At least with your mount, it does appear to guide without the strange spikes/oscillations etc seen in some of the other mounts. Based on yr comment re seeing an pollution, the guided scatter plot isnt too bad.

 

Andrew Johansen Melbourne Australia

First 30 mins raw x/y ( at 1.8 arcsec/pix )

attachicon.gif Guided Scatter.jpg

Yes, indeed. And the stars are pretty round. I haven't discard any sub yet during the processing of the astrophotos.

 

Sasho

Attached Thumbnails

  • M101_s.jpg

  • Mondeclerk likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics