Guiding Issue with large spikes in Dec and RA
Arthur Williams
Lenny Shaffer
--
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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/be488f3b-06f0-402c-a7ba-a87cfeac8b7an%40googlegroups.com.
Murphy's Law of the Week: "If authority was mass, stupidity would be gravity"
Bruce Waddington
Arthur Williams
Bruce Waddington
Arthur Williams
Bruce Waddington
Brian Valente
OK, so I just finished writing a Python script to analyze the PHD2 Guide Log file from last weekend (uploaded here). I was specifically interested in how large RA spikes (>1.25") correlated with prior Dec guider adjustments. The results:
- 89 of 117 large RA spikes were immediately preceded by large DEC moves (76%)
- 65 of the 92 large DEC moves were reversals (71%)
So roughly 3/4 of the RA spikes can probably be attributed to Dec guiding corrections. And of those, over 2/3 were reversals.
If I increase the definition of an "RA spike" from 1.25" to 2.0", the numbers are even more dramatic:
- 41 of 46 large RA spikes were immediately preceded by large DEC moves (89%)
- 32 of the 46 large DEC moves were reversals (70%)
I think this leaves little doubt that suddent, large Dec movements are not only affecting my Dec guiding, but my RA guiding as well. Somehow, I need to find a way to tame PHD2 so it isn't jolting the Dec axis around so much. Two suggestions given to me so far:
- Disable FastSwitch - I thought I had done that but in looking at the log file I see it was enabled. My bad.
- Back off the aggressiveness which was set at 100%. I actually did reduce it to 80% last weekend and that did seem to improve things in the last guiding run of the night.
Should I also reduce the Minimum Movement from 0.25 to something lower to the Dec adjustments are so drastic? Or would that cause it to overreact?
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/bac6a7d5-029a-4ffc-9bb0-5f4ad64de82cn%40googlegroups.com.
bw_m...@earthlink.net
If you’d like us to give you advice, we need to see your guide and debug log files so we can analyze the raw data. That was the link I gave you in the previous message: https://openphdguiding.org/getting-help/
Bruce
From: open-phd...@googlegroups.com <open-phd...@googlegroups.com> On Behalf Of Bill Richards
Sent: Thursday, July 28, 2022 6:34 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] Guiding Issue with large spikes in Dec and RA
OK, so I just finished writing a Python script to analyze the PHD2 Guide Log file from last weekend (uploaded here). I was specifically interested in how large RA spikes (>1.25") correlated with prior Dec guider adjustments. The results:
- 89 of 117 large RA spikes were immediately preceded by large DEC moves (76%)
- 65 of the 92 large DEC moves were reversals (71%)
So roughly 3/4 of the RA spikes can probably be attributed to Dec guiding corrections. And of those, over 2/3 were reversals.
If I increase the definition of an "RA spike" from 1.25" to 2.0", the numbers are even more dramatic:
- 41 of 46 large RA spikes were immediately preceded by large DEC moves (89%)
- 32 of the 46 large DEC moves were reversals (70%)
I think this leaves little doubt that suddent, large Dec movements are not only affecting my Dec guiding, but my RA guiding as well. Somehow, I need to find a way to tame PHD2 so it isn't jolting the Dec axis around so much. Two suggestions given to me so far:
- Disable FastSwitch - I thought I had done that but in looking at the log file I see it was enabled. My bad.
- Back off the aggressiveness which was set at 100%. I actually did reduce it to 80% last weekend and that did seem to improve things in the last guiding run of the night.
Should I also reduce the Minimum Movement from 0.25 to something lower to the Dec adjustments are so drastic? Or would that cause it to overreact?
On Wednesday, July 27, 2022 at 8:10:44 AM UTC-7 bw_m...@earthlink.net wrote:
It's good to hear you're making progress, hope you can continue to get improvements.
Bruce
On Tuesday, July 26, 2022 at 2:31:07 PM UTC-7 arthur_...@btinternet.com wrote:
Finally got a clear night and have tracked part of the problem down to an issue with the worm gear mesh adjustment on the Dec drive. There was no slack between the motor and the worm gear. Slackened of the limit adjustment a little and it sorted most of the problems with spiking in DEC and RA - there was no movement at all before hand and there should be a little. Also did some work on the cabling - could still be better but my guiding is vastly improved. Still getting some erratic behavior on DEC so still more work to do.
On Thursday, 21 July 2022 at 20:37:51 UTC+1 bw_m...@earthlink.net wrote:
To come back to the main point, you are almost surely not going to be successful guiding a 2800mm scope with a separate guide scope - the differential flexure will kill you unless you're willing to live with only 1-2 minute main camera exposures. You should be using an off-axis-guider of some kind, that's a well-accepted fact.
Setting that aside, I would say you don't really have a cable-routing solution - it's just a mess. All of those hanging cables have the potential to pull on the guide camera, even more so because you've apparently fastened some of them together. Here's a recent post from another skeptic on this topic:
The details of the cable mess will be completely non-deterministic, depending on the recent history of where and how the scope has been slewed. So it's not surprising that your results vary from night to night. Beyond that problem, there are other weaknesses:
1. While there are two mount points on the finder-scope, they are close together, leaving a long, unsupported overhang of the guide camera.
2. The finder-scope is fastened in the rings with Delrin-tipped thumb-screws. While you might be able to get these tight enough at the start of the evening, they usually won't stay tight through slewing operations and temperature changes. These are a well-known source of problems
3. You've got a cable-fed dew shield on the front of the finder, adding yet another opportunity for a cable to tug on the optical assembly, this time in front of the mounting rings.
On another front, we need to ask why you were getting a regular burst of problems, nearly all in RA:
These were happening every 8 minutes which I think coincides with the worm period on the mount. Is there anything in your imaging operation that happens every 8 minutes? Is the RA drive system making noises at these intervals? Is there grit or something in the drive system? If you make a long Guiding Assistant run, you can see if these problems are consistent.
In any case, these are not guiding problems, a fact that you pretty much figured out for yourself.
Hope you can track down the problems,
Bruce
On Thursday, July 21, 2022 at 10:05:33 AM UTC-7 arthur_...@btinternet.com wrote:
That was my first thought as well - was one of my cables catching something - but I have watched the scope whilst guiding and at no time do any of the cables catch or brush against something.
The main telescope is at F10, so 2800 mm focal length. The ASCOM platform used is 6.3 installed 4 Dec 2017. I believe the Starsense ASCOM driver came from the https://download.ascom-standards.org/drivers/ and at the time this was the latest Celestron version.
It has never reported side of pier - I have always manually switched the setting inside PH2D when I need a meridian flip. Very rarely a problem as I don't have a clear view to the west so invariably image on the east side anyway.
Usually I use auto find, but on this occasion I manually selected the guide star. I have never managed to get multi star guiding working, always ends up being single star guiding.
My cables are routed from a Pegasus power box attached to the dovetail bar on the underside of the scope. The cables to the telescope dew heaters, cameras and focuser are tied along the dovetail but do hang down of the back of the cameras as they are longer than needed, but I have checked they never actually catch on anything. The dew heater power cables are particularly long so they are folded as you can see in photo below. As the pier is inside a roll of roof observatory, not troubled by wind either. A single power cable and usb cable to Computer connect to the Pegasus box running down the side of the pier, but again loose so they don't catch.
The guide scope is attached using a two ring mounting system and whilst if i tap it, it does cause the image to oscillate but only the same as if I tap the main scope. I have rechecked the guide scope attachment to the main scope and they are tight as they have always been.
Whilst I accept that the cables are hanging of the guide scope, over a period of 1 or 2 hours the down ward pull on the scope would marginally change slowly (due to the weight of the cable and the movement of the scope - the downward force on the scope would change as the RA of the scope changes) but surely this wouldn't cause the repeated wayward DEC and RA movements as the change would be smooth.
Also as I have indicated in earlier post I sometimes can get whole sessions with no wayward DEC and RA movements or perhaps 1 or 2 over a 3 hour period, whilst as this one shows sometimes I get problems many spikes at various intervals - no obvious pattern to them in session or from session to session.
On Thursday, 21 July 2022 at 16:30:43 UTC+1 bw_m...@earthlink.net wrote:
Hi Arthur. I don't think this has anything to do with the mount at this point, these big guide star excursions are coming from an external source. I would look first at how you've done your cable routing - you can't afford to have dangling cables or have them arranged so they can pull on the guide scope assembly at all. Second, you should check all the fittings on the guide scope assembly to be sure nothing can move. These "huge" excursions you're seeing correspond to unwanted movement of only 5-10 microns at the guide camera sensor - so less than 10% of the thickness of a human hair. Said another way, if the guide camera shifts position by 10 microns - for whatever reason - you've just created a 30 arc-sec guide star excursion. What is the focal length of your main scope? If it's 2000mm or above, you should be using an off-axis-guider instead of whatever short focal length scope you have now. If you want, you can post some pictures of your setup.
On a couple of unrelated points:
1. Your mount driver isn't reporting side-of-pier. That means that PHD2 can't automatically adjust the internal guiding parameters when a meridian flip occurs. I don't recognize this "StarSense" ASCOM driver you're using, where did it come from?
2. Why are you forcing single-star guiding by manually selecting a guide star? You should be using the auto-find button and let PHD2 choose multiple guide stars for you.
Good luck,
Bruce
On Thursday, July 21, 2022 at 3:06:20 AM UTC-7 Lenny wrote:
Sounds like excessive backlash.
On 7/21/2022 4:57 AM, 'Arthur Williams' via Open PHD Guiding wrote:
Hi, I have been using PH2D for successful guiding for many years without any real issues (except very poor polar alignment resulting in large DEC drift and field rotation) to guide my tripod mounted telescope.
However, since setting up a permanent mount for my telescope on a pier, I have started to experience periods where I get large spikes in both Dec and RA (upto 30" of deviation). Some sessions guide okay throughout 2 - 3 hour long sessions. Others will guide without issue for several minutes - sometimes over 40 minutes - but then suddenly it goes very erratic with large spikes in DEC and RA before settling back down for a period. In the attached log you can see in the 3rd session (2022-07-19 22:20:15) repeated periods of instability. This session was particularly bad in that there was only 6 - 10 minutes of stability between the erratic behavior. More usual is to get 30 - 40 minutes at the start of a guiding session with no problems and then it starts with the spiking at intervals that range from 2 upto 40+ minutes. I have checked my guide scope mounting and it is solid. I also have looked at my cable management and there is no obvious cable catching.
I had run the Guiding Assistant and accepted its recommendations and also recalibrated just before this session and the calibration looked slightly better than normal.
Looking at the logs, it doesn't appear to be a guide pulse that is initiating the spiking as the corrections follow the spike rather than being before it.
Any ideas as to what is going on and how to improve the guiding?
Equipment is Celestron C11 Edge CGX on a Altair Astro pier using an 60mm F4 guidescope with ZWO ASI120MC camera for guiding.
I have attached the log for the 19 July when it was particularly bad and a screenshot of the relevant example guiding log.
--
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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/be488f3b-06f0-402c-a7ba-a87cfeac8b7an%40googlegroups.com.
--
Murphy's Law of the Week: "If authority was mass, stupidity would be gravity"
--
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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/bac6a7d5-029a-4ffc-9bb0-5f4ad64de82cn%40googlegroups.com.
Bill Richards
bw_m...@earthlink.net
Bill, I think you’ve created a mess here by hi-jacking someone else’s message thread. You can’t just do ‘replies’ to incoming messages from the forum, you need to pay attention to the subject line. Please open your own thread so we can respond. I see now that you did include some log files so we’ll look at those. Once you post a new message, I will delete this one so we don’t leave things in such a mess.
Thanks,
Bruce
From: open-phd...@googlegroups.com <open-phd...@googlegroups.com> On Behalf Of Bill Richards
Sent: Thursday, July 28, 2022 6:34 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] Guiding Issue with large spikes in Dec and RA
OK, so I just finished writing a Python script to analyze the PHD2 Guide Log file from last weekend (uploaded here). I was specifically interested in how large RA spikes (>1.25") correlated with prior Dec guider adjustments. The results:
- 89 of 117 large RA spikes were immediately preceded by large DEC moves (76%)
- 65 of the 92 large DEC moves were reversals (71%)
So roughly 3/4 of the RA spikes can probably be attributed to Dec guiding corrections. And of those, over 2/3 were reversals.
If I increase the definition of an "RA spike" from 1.25" to 2.0", the numbers are even more dramatic:
- 41 of 46 large RA spikes were immediately preceded by large DEC moves (89%)
- 32 of the 46 large DEC moves were reversals (70%)
I think this leaves little doubt that suddent, large Dec movements are not only affecting my Dec guiding, but my RA guiding as well. Somehow, I need to find a way to tame PHD2 so it isn't jolting the Dec axis around so much. Two suggestions given to me so far:
- Disable FastSwitch - I thought I had done that but in looking at the log file I see it was enabled. My bad.
- Back off the aggressiveness which was set at 100%. I actually did reduce it to 80% last weekend and that did seem to improve things in the last guiding run of the night.
Should I also reduce the Minimum Movement from 0.25 to something lower to the Dec adjustments are so drastic? Or would that cause it to overreact?
On Wednesday, July 27, 2022 at 8:10:44 AM UTC-7 bw_m...@earthlink.net wrote:
It's good to hear you're making progress, hope you can continue to get improvements.
Bruce
On Tuesday, July 26, 2022 at 2:31:07 PM UTC-7 arthur_...@btinternet.com wrote:
Finally got a clear night and have tracked part of the problem down to an issue with the worm gear mesh adjustment on the Dec drive. There was no slack between the motor and the worm gear. Slackened of the limit adjustment a little and it sorted most of the problems with spiking in DEC and RA - there was no movement at all before hand and there should be a little. Also did some work on the cabling - could still be better but my guiding is vastly improved. Still getting some erratic behavior on DEC so still more work to do.
On Thursday, 21 July 2022 at 20:37:51 UTC+1 bw_m...@earthlink.net wrote:
To come back to the main point, you are almost surely not going to be successful guiding a 2800mm scope with a separate guide scope - the differential flexure will kill you unless you're willing to live with only 1-2 minute main camera exposures. You should be using an off-axis-guider of some kind, that's a well-accepted fact.
Setting that aside, I would say you don't really have a cable-routing solution - it's just a mess. All of those hanging cables have the potential to pull on the guide camera, even more so because you've apparently fastened some of them together. Here's a recent post from another skeptic on this topic:
The details of the cable mess will be completely non-deterministic, depending on the recent history of where and how the scope has been slewed. So it's not surprising that your results vary from night to night. Beyond that problem, there are other weaknesses:
1. While there are two mount points on the finder-scope, they are close together, leaving a long, unsupported overhang of the guide camera.
2. The finder-scope is fastened in the rings with Delrin-tipped thumb-screws. While you might be able to get these tight enough at the start of the evening, they usually won't stay tight through slewing operations and temperature changes. These are a well-known source of problems
3. You've got a cable-fed dew shield on the front of the finder, adding yet another opportunity for a cable to tug on the optical assembly, this time in front of the mounting rings.
On another front, we need to ask why you were getting a regular burst of problems, nearly all in RA:
These were happening every 8 minutes which I think coincides with the worm period on the mount. Is there anything in your imaging operation that happens every 8 minutes? Is the RA drive system making noises at these intervals? Is there grit or something in the drive system? If you make a long Guiding Assistant run, you can see if these problems are consistent.
In any case, these are not guiding problems, a fact that you pretty much figured out for yourself.
Hope you can track down the problems,
Bruce
On Thursday, July 21, 2022 at 10:05:33 AM UTC-7 arthur_...@btinternet.com wrote:
That was my first thought as well - was one of my cables catching something - but I have watched the scope whilst guiding and at no time do any of the cables catch or brush against something.
The main telescope is at F10, so 2800 mm focal length. The ASCOM platform used is 6.3 installed 4 Dec 2017. I believe the Starsense ASCOM driver came from the https://download.ascom-standards.org/drivers/ and at the time this was the latest Celestron version.
It has never reported side of pier - I have always manually switched the setting inside PH2D when I need a meridian flip. Very rarely a problem as I don't have a clear view to the west so invariably image on the east side anyway.
Usually I use auto find, but on this occasion I manually selected the guide star. I have never managed to get multi star guiding working, always ends up being single star guiding.
My cables are routed from a Pegasus power box attached to the dovetail bar on the underside of the scope. The cables to the telescope dew heaters, cameras and focuser are tied along the dovetail but do hang down of the back of the cameras as they are longer than needed, but I have checked they never actually catch on anything. The dew heater power cables are particularly long so they are folded as you can see in photo below. As the pier is inside a roll of roof observatory, not troubled by wind either. A single power cable and usb cable to Computer connect to the Pegasus box running down the side of the pier, but again loose so they don't catch.
The guide scope is attached using a two ring mounting system and whilst if i tap it, it does cause the image to oscillate but only the same as if I tap the main scope. I have rechecked the guide scope attachment to the main scope and they are tight as they have always been.
Whilst I accept that the cables are hanging of the guide scope, over a period of 1 or 2 hours the down ward pull on the scope would marginally change slowly (due to the weight of the cable and the movement of the scope - the downward force on the scope would change as the RA of the scope changes) but surely this wouldn't cause the repeated wayward DEC and RA movements as the change would be smooth.
Also as I have indicated in earlier post I sometimes can get whole sessions with no wayward DEC and RA movements or perhaps 1 or 2 over a 3 hour period, whilst as this one shows sometimes I get problems many spikes at various intervals - no obvious pattern to them in session or from session to session.
On Thursday, 21 July 2022 at 16:30:43 UTC+1 bw_m...@earthlink.net wrote:
Hi Arthur. I don't think this has anything to do with the mount at this point, these big guide star excursions are coming from an external source. I would look first at how you've done your cable routing - you can't afford to have dangling cables or have them arranged so they can pull on the guide scope assembly at all. Second, you should check all the fittings on the guide scope assembly to be sure nothing can move. These "huge" excursions you're seeing correspond to unwanted movement of only 5-10 microns at the guide camera sensor - so less than 10% of the thickness of a human hair. Said another way, if the guide camera shifts position by 10 microns - for whatever reason - you've just created a 30 arc-sec guide star excursion. What is the focal length of your main scope? If it's 2000mm or above, you should be using an off-axis-guider instead of whatever short focal length scope you have now. If you want, you can post some pictures of your setup.
On a couple of unrelated points:
1. Your mount driver isn't reporting side-of-pier. That means that PHD2 can't automatically adjust the internal guiding parameters when a meridian flip occurs. I don't recognize this "StarSense" ASCOM driver you're using, where did it come from?
2. Why are you forcing single-star guiding by manually selecting a guide star? You should be using the auto-find button and let PHD2 choose multiple guide stars for you.
Good luck,
Bruce
On Thursday, July 21, 2022 at 3:06:20 AM UTC-7 Lenny wrote:
Sounds like excessive backlash.
On 7/21/2022 4:57 AM, 'Arthur Williams' via Open PHD Guiding wrote:
Hi, I have been using PH2D for successful guiding for many years without any real issues (except very poor polar alignment resulting in large DEC drift and field rotation) to guide my tripod mounted telescope.
However, since setting up a permanent mount for my telescope on a pier, I have started to experience periods where I get large spikes in both Dec and RA (upto 30" of deviation). Some sessions guide okay throughout 2 - 3 hour long sessions. Others will guide without issue for several minutes - sometimes over 40 minutes - but then suddenly it goes very erratic with large spikes in DEC and RA before settling back down for a period. In the attached log you can see in the 3rd session (2022-07-19 22:20:15) repeated periods of instability. This session was particularly bad in that there was only 6 - 10 minutes of stability between the erratic behavior. More usual is to get 30 - 40 minutes at the start of a guiding session with no problems and then it starts with the spiking at intervals that range from 2 upto 40+ minutes. I have checked my guide scope mounting and it is solid. I also have looked at my cable management and there is no obvious cable catching.
I had run the Guiding Assistant and accepted its recommendations and also recalibrated just before this session and the calibration looked slightly better than normal.
Looking at the logs, it doesn't appear to be a guide pulse that is initiating the spiking as the corrections follow the spike rather than being before it.
Any ideas as to what is going on and how to improve the guiding?
Equipment is Celestron C11 Edge CGX on a Altair Astro pier using an 60mm F4 guidescope with ZWO ASI120MC camera for guiding.
I have attached the log for the 19 July when it was particularly bad and a screenshot of the relevant example guiding log.
--
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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/be488f3b-06f0-402c-a7ba-a87cfeac8b7an%40googlegroups.com.
--
Murphy's Law of the Week: "If authority was mass, stupidity would be gravity"
--
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 on the web visit https://groups.google.com/d/msgid/open-phd-guiding/bac6a7d5-029a-4ffc-9bb0-5f4ad64de82cn%40googlegroups.com.