Is it the mount, or is it - me?

[Steven Bellavia]

On my Celestron AVX mount, when you press the negative RA button (Left arrow) it drifts continuously until you press the positive RA button (Right).

https://www.flickr.com/photos/125134422@N06/29282782791/in/datetaken-public/ 

That is with no auto-guiding.

 

I suspect this is also what is causing my terrible auto-guiding in RA.  I tried changing every possible setting, in a very methodical way, changing each parameter a small amount and noting the performance, until I maxed out the setting, along with changing the mount's anti-backlash settings (from 0 to 100, on all four (+/- RA and +/- DEC) and also the autoguide rates, as well as the balance, going from perfectly neutral, to slightly East heavy, moderately East Heavy, very East heavy, etc

 

And as you can see from this plot made from the PHD2 guide log,  I had excellent Polar alignment, with less than 0.25 arc-second (RMS) DEC drift over 5 minutes.

What I think is happening is that the guide camera issues a negative RA motion, and then the mount just keeps drifting, long after the signal was sent, requiring a large positive RA correction, and it goes back and forth.

Any help is appreciated!

[Kenneth Crichton]

I had a very similar issue with my cpc mount. Firstly check that the button on the hand control isn't sticking which was my issue even though it was very subtle. Failing that it sounds mechanical although someone else may better explain it .

[Andy Galasso]

Hi Steven,

What you are seeing is the symptoms of backlash in the RA drive. The good news is that you should still be able to achieve good RA guiding with PHD2 despite the backlash.

• turn off the mount's backlash compensation in both RA and Dec.
  
• make sure your mount's guide rate setting is in the range 0.5x - 1.0x sidereal
  
• balance your scope slightly east-heavy so that the RA drive is always lifting against gravity. A very small weight bias is sufficient, like several ounces.  This will keep the RA gears engaged always on one side, never taking up the backlash during tracking or guiding.  When you do a meridian flip you'll need to shift the counterweights to re-establish the East-heavy weight bias.
  
• you may be able to adjust the mount mechanically to reduce the backlash though that is not necessary for good guiding results. Consult the user forums for your mount, or your mount manufacturer's tech support.

If you still see any problem guiding, please post your PHD2 Guide Log and your Debug Log (Instructions here)

If you haven't had a chance to do so, I'd recommend taking a look at the PHD2 user guide  where we talk about backlash and many other issues. Bruce's tutorial on analyzing guiding results is also a great resource.

Good luck.

Andy

[Steven Bellavia]

Thanks Andy, but, as I said in my original post, I have done exactly everything you outlined, below (except for physically adjusting the gear mesh).  No improvement whatsoever.   How can you explain the constant negative RA drift when there is no guiding?  That's not backlash.   I think the mount has an electrical / logic flaw.  I am not sure of the circuitry, but I believe that their implementation for anti-backlash involves a signal sent to the motor to a reverse direction for a brief period, except that somehow the implementation of this "advanced" feature causes the motor to completely stop tracking.  The drift seems to be almost exactly the sidereal rate, which means that after a negative RA signal is issued, the motor just turns off until a positive RA signal is sent.  I am trying to work with Celestron on this, but they are painfully slow to respond.

[Steven Bellavia]

Did you watch the video?

https://www.flickr.com/photos/125134422@N06/29282782791/in/datetaken-public/

On Wednesday, September 7, 2016 at 1:50:22 PM UTC-4, Andy Galasso wrote:

[Andy Galasso]

On Wed, Sep 7, 2016 at 2:05 PM, Steven Bellavia <stevenb...@gmail.com> wrote:

No improvement whatsoever. 

Hi Steve,

If you are concerned with the drifting after you nudge the scope with the hand control, I don't think that is a guiding question, so I'll defer to  Celeston mount experts to help you address that.

I'd be happy to look at your phd2 logs to help you diagnose the guiding behavior, but we'll need to see logs (guide log + debug log) from a session with backlash comp disabled in the mount, guide rate set in the 0.5-1.0x sidereal range, and scope balanced with a slight East-heavy bias.

Andy 

[Sandy MacPherson]

I have an AVX and it does the same thing..it drifts for a bit until the gears make contact again which is due to backlash.  It should not affect guiding as there is no reversal of gear direction in RA

[peter wolsley]

Steven,

I own a CGEM mount and I have seen this RA drift issue as well.  I hope it is as Sandy described...RA backlash.  I have roughly 60" of RA backlash so I see the star drift for 4 seconds before it start tracking again.  Just to clarify the "drifting" I see is actually the RA motor turning at Sidereal rate 15"/Sec but the gears are moving thru there backlash.  It takes 4 seconds for it to move thru it's backlash. The physical RA axis is not moving during this time so what I see as "drift" is actually the star slowing moving across the sky at Sidereal rate. There are two things I would like to know and see.

1)Can you repeat the test you conducted on the video.  Move the guide star to the right and then to the left and then let the star drift until you see it stop drifting.  Keep track of the seconds that it takes for the drifting to stop.  If you don't have any backlash compensation turned on in your AVX then the time in seconds multiplied by 15 equals the backlash in your RA axis.  I can tell from your video it is going to be a big number.  If it star stops drifting then you are dealing with excessive RA backlash and you are going to need to make mechanical adjustments to reduce your RA backlash.  If the star continues to drift then you either have a sticking button or the firmware in the mount or hand controller may be corrupted.  For my CGEM I use MCUpdate or HCUpdate windows application to do update firmware.  Another possibility would be if you have an ASCOM driver for your AVX you could try slewing your telescope using the ASCOM driver...the software you show on the video should be able to send slew commands to your AVX.  If the star stops drifting using ASCOM then, at least, you have a work around.

2)Can you provide a PHD Guiding log?

Peter

[Steven Bellavia]

Hi Peter (and Sandy and Andy),

First I want to say how awesome you guys are for giving this so much attention.  THANK YOU!

Hmmmm.  Giant backlash.  It is possible.  I think the drifting does eventually stop, and when I set the negative RA backlash setting to 20 (I just tried this), with all others at 0, it almost stops completely.  But anything greater than 20, and when you press the neg RA button it actually makes a huge jump in the wrong direction.

I am going to try this again tonight (if the clouds pass, as it has gotten cloudy, but they predict another clearing), and get a good, steady guide log.  I have literally hundred of guide logs, but I want a nice, clean steady run, where I haven't changed any of the settings for a while.

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guiding+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Steven Bellavia]

There was also an occasion where even pressing the right (+RA) button would not stop the drift, and I had to turn off the mount and start over.

https://www.flickr.com/photos/125134422@N06/28910772123/in/datetaken-public/

And another occasion, when I had sub-arc-second guiding, both axes, on the very same night, without changing mount balance or any other settings.

https://www.flickr.com/photos/125134422@N06/29423951792/in/datetaken-public/

[peter wolsley]

Steven,

The more I look at your videos...the more I can't believe it's just RA backlash.  I would think that if your backlash is hugh then you should see visible backlash in your mount if you lock the RA axis and then try to rotate it with your hand.  With my CGEM I can only see the very slightest movement if I try to rotate RA with my hand on the end of the counterbalance bar.

It really is beginning to look as if when you press the RA button to move the star to the left that as soon as you release the button, it stops the RA motor...i.e. disables tracking.  As far as I know the tracking/slew sequencing logic is in the hand controller.  You should try reloading the firmware into the hand controller.

I use a Celestron application called Nexremote which emulates my hand controller...I virtually never touch my hand controller.  I can align/calibrate/goto using Nexremote.  Do you have a windows application that emulates your hand controller?  If so, you should repeat the test using this program.  If it works ok then I would suspect your hand controller.

Good Luck

Peter

[Steven Bellavia]

I just did another test, but indoors.  I took the gear cover off and watched, and also set the hand controller to readout the axes position.

(anti-backlash at 0, all directions, all axes). If you set the motor speed at 3 or 4, and press the neg RA button quickly, the gear moves the wrong direction, briefly, but every single time.  It is very apparent, both looking at the gears and the axes position readout.  If you set it to 2, the motor stops, completely.  No noise at all.  No motion in RA at all.  You can hold the button forever, and it never moves,  And if you set it to speed 1, it never reverses, and keeps tracking in + RA.

According to the manual, 2X sidereal is the slowest it should go, in any direction, any axis.  Not zero or less than zero.

On Wednesday, September 7, 2016 at 10:42:50 PM UTC-4, Steven Bellavia wrote:

[Steven Bellavia]

So at speed setting = 2,  even if stopping the motor was correct for negative RA (like the older mounts used to do), that would be 1X sidereal in the wrong direction, not 4X as the manual says?

[Ray Gralak]

Hi Steve,

Maybe you are looking at the wrong manual or the firmware has changed? Attached is the screen shot of what I found. The rates seem to match the behavior you are describing. That is:

1 -> 0.5x (apparent motion is 0.5x sidereal when a button is pressed). Gears will always move forward, never any backlash.

2 -> 1.0x (apparent motion is 1x sidereal when a button is pressed). Gears stop against RA motion, or add an extra 1x rate to motion. Never any backlash.

3 -> 4x . Gears should reverse at 3x sidereal rate, or add an extra 4x sidereal rate forward. Backlash could be evident.

BTW, the manual where I found the attached screenshot is at this URL (page 23):

http://www.celestron.com/media/795887/1235085036_cgem11097inst.pdf

-Ray Gralak

[Steven Bellavia]

Oh!  That was the manual that came with the mount.  But yours makes much more sense!

And getting slightly off topic (PHD2 stuff), why do they have the motor gear turning another gear of the same exact size, attached to the worm gear?  This is where all the backlash is, and could be eliminated if instead they:

1. Made the same motor turn the worm directly (where there is very little backlash from worm to worm-gear)

 - OR -

2. Use a belt instead of gears.  The load is very small, so it would probably never slip.

--

[peter wolsley]

Steven,

In my CGEM, the matching 2-gear design allows me to install/remove the motor/gearbox assembly easily.  A belt design probably would have been difficult for the CGEM.  I believe, on the AVX mount these two gears can be easily accessed so designing them as a belt drive is a "missed opportunity" for Celestron to make improvements.  I suspect the two-gear design has been around for a while.

I removed my RA motor/gearbox assembly a few times to examine them.  I noticed that I can rotate the large output gear almost 1/2 tooth without turning the motor...so there is 1/2 tooth backlash just in the motor/gearbox assembly.  I remember doing some calculations and I had figured out that each tooth on this gear represented roughly 120" to 140" of rotation.  It's my opinion that this backlash in the motor/gearbox assembly is roughly 60".  On the CGEM, the RA and DEC motor/gearboxes are identical.  I believe the motor/gearbox in my CGEM is different than in your AVX because the AVX features integer gear ratios.  Hopefully Celestron was also able to reduce the gearbox backlash while they were at it. The total backlash in the RA axis includes this motor/gearbox backlash and then you would add the backlash caused by these two large gears and then finally the wormgear/ringgear interface.  Replacing the two large gears with a belt drive would help but there is still lots of gears with backlash to account for.  Fixing all of these problems is surprisingly easy...just add money.

Peter

To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.

[CyScape]

There are 2 places for backlash to exist. 1.) The gear to gear mesh and 2.) Worm gear to ring gear. If you remove the gear attached to the motor and then rotate the worm gear back and fourth, you'll see and feel the slop.

[Steven Bellavia]

I think there is much less backlash there then those two same-sized gears, but more importantly, east-heavy really helps the worm-ring gear backlash, but it doesn't make its way "back" to those two gears, so it has almost no effect in improving that situation.

On Thu, Sep 8, 2016 at 2:59 PM, 'CyScape' via Open PHD Guiding <open-phd...@googlegroups.com> wrote:

There are 2 places for backlash to exist. 1.) The gear to gear mesh and 2.) Worm gear to ring gear. If you remove the gear attached to the motor and then rotate the worm gear back and fourth, you'll see and feel the slop.

[Steven Bellavia]

I have another question. (like Winnie-the-pooh, I am a bear of very little brain):

If the signals sent from the guidescope are always less than sidereal rate (I have tried 0.25, 0.5 and 0,75X sidereal autoguide settings, 1.0 is the max, which just stops the motor - you can hear it when  on this setting), than how can there be backlash in RA?  There would never be a signal that would cause the gear to go in the opposite direction (Of course for DEC this is not true, but I am talking  about RA only).  It is still the gear loading up in the same direction, just slower when a negative RA pulse is sent.

Unless this is the scope "bouncing" from the change in speed, and, by definition, any change in speed is an acceleration, and since the system has mass, that means there is a force in the opposite direction (F=ma), for a negative acceleration, which I guess could cause a bounce.  But I am still not sure this is backlash of the gears?  It might just be the "bounce" of the entire equatorial assembly and scope, while all the gear teeth are still loaded up, all in the same direction, as they were before the negative RA signal?

So maybe I just answered my own question?  But any help is appreciated!

Steve

P.S.  I am asking this because I think I am gong to try replacing the two RA drive gears with a set of anti-backlash gears (one anti-backlash, one mating spur).  It's about $170 with shipping, so not a whole lot of money, but now I am not sure it will help.  But I am 99% sure  it won't make it any worse.

On Thu, Sep 8, 2016 at 4:07 PM, Steven Bellavia <stevenb...@gmail.com> wrote:

I think there is much less backlash there then those two same-sized gears, but more importantly, east-heavy really helps the worm-ring gear backlash, but it doesn't make its way "back" to those two gears, so it has almost no effect in improving that situation.

[peter wolsley]

Steven,

You started off in this topic talking about the RA drift you see when slewing.  Are you now convinced that it is caused by backlash?  You are sure it's not a software bug in the HC firmware?

Your latest post is focused on guiding and I think that RA backlash is really a very very small issue during guiding.  In my opinion, the ringgear/wormgear interface has such a high gear ratio that any inertial F=ma effects will never get transmitted back thru the wormgear to the two gear set.  I do think that F=ma does enter into DEC guiding.  When the DEC motor is stopped and then pulse guided at 50% (7.5"/Sec) this may kick the ringgear/telescope assembly a little too far and open up the ringgear/wormgear backlash.  The guide star might travel a little too far in DEC which would then cause the guiding to command the DEC motor to reverse direction.  Unfortunately this would mean reversing all of the gears...opening up all of the backlashes in all of the gears until the backlash is taken up in the reverse direction so that the wormgear can finally begin moving in the reverse direction.

Just my two cents

Peter

Saturday, September 10, 2016 at 8:34:00 AM UTC-4, Steven Bellavia wrote:

I have another question. (like Winnie-the-pooh, I am a bear of very little brain):

If the signals sent from the guidescope are always less than sidereal rate (I have tried 0.25, 0.5 and 0,75X sidereal autoguide settings, 1.0 is the max, which just stops the motor - you can hear it when  on this setting), than how can there be backlash in RA?  There would never be a signal that would cause the gear to go in the opposite direction (Of course for DEC this is not true, but I am talking  about RA only).  It is still the gear loading up in the same direction, just slower when a negative RA pulse is sent.

Unless this is the scope "bouncing" from the change in speed, and, by definition, any change in speed is an acceleration, and since the system has mass, that means there is a force in the opposite direction (F=ma), for a negative acceleration, which I guess could cause a bounce.  But I am still not sure this is backlash of the gears?  It might just be the "bounce" of the entire equatorial assembly and scope, while all the gear teeth are still loaded up, all in the same direction, as they were before the negative RA signal?

So maybe I just answered my own question?  But any help is appreciated!

Steve

P.S.  I am asking this because I think I am gong to try replacing the two RA drive gears with a set of anti-backlash gears (one anti-backlash, one mating spur).  It's about $170 with shipping, so not a whole lot of money, but now I am not sure it will help.  But I am 99% sure  it won't make it any worse.

On Thu, Sep 8, 2016 at 4:07 PM, Steven Bellavia <stevenb...@gmail.com> wrote:

I think there is much less backlash there then those two same-sized gears, but more importantly, east-heavy really helps the worm-ring gear backlash, but it doesn't make its way "back" to those two gears, so it has almost no effect in improving that situation.

On Thu, Sep 8, 2016 at 2:59 PM, 'CyScape' via Open PHD Guiding <open-phd...@googlegroups.com> wrote:

There are 2 places for backlash to exist. 1.) The gear to gear mesh and 2.) Worm gear to ring gear. If you remove the gear attached to the motor and then rotate the worm gear back and fourth, you'll see and feel the slop.

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.

To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.

[Steven Bellavia]

Hi Peter,

Thanks again for responding.

I am not convinced of anything at this point.  I still do not believe it is backlash, but until I do more extensive testing, I don't want to rule anything out.  There are times when I get a steady negative RA drift that never stops and I have to turn the mount off and start over, so that is obviously not backlash, but that only happens once in a while.  But that makes me think it might be related to the other negative RA drift I get while slewing, and maybe even the huge RA swings while guiding, so electrical, or firmware/software, but not gears.

[peter wolsley]

Steven,

Next time you have a situation where you want to turn off the mount because of a steady RA drift you might want to determine whether the mount thinks it is tracking.  Maybe even tell the mount to track using the hand controller to see if it "wakes up" and starts tracking.  It seems to me that either the mount has turned off tracking or it may still say it is tracking but the signal to the motor is inhibited some how?  If the motor is inhibited by some logic bug it might be convinced to "snap-out-of-it" by simply telling it to track.

Just for fun I search AVX Firmware and found this info regarding the motor control firmware:

MC Versions

• 5.20 - Support for "Custom Rate 9": the ability to set the maximum slew rate during a GoTo (requires hand control version 4.20 or higher).  The final GoTo approach is now faster and quieter.
• 5.16 - Corrects problems in RA Limits (previously named Slew Limits) and improves precision of RA Limits to 1 degree (improved precision only available when using hand control version 4.16 or higher).
• 5.14 - Corrects a potential problem with the cordwrap prevention feature.  Re-enables the RA Slew Limits.
• 5.07 - An official version.  Minor changes - rate 1 is now 50% of sidereal while rate 2 is 100% of sidereal.  This makes manual guiding much easier.
• 5.06 - First official version available for download from Celestron.  Improved anti-backlash, fixed runaway slews during normal use and when when autoguiding, general improvements to autoguiding, fixed occasional slow motion slews that never seem to reach their destination, improved sidereal tracking rate.
• 5.04 - Original version on the AS-GT mount.
  
• 
  

This may not apply to you but it's interesting that slow motion slews that lasted forever are not a new issue for Celestron.

Peter

[Steven Bellavia]

I did a very long series of tests last night.

One of the tests was without guiding, just to record the mount's motion while tracking. A huger periodic error was apparent.

So I then used the mount's Periodic Error Control and recorded and played back the PEC for most of the remainder of the tests (which went to 3AM).

This is a link to the GuideLog:

https://drive.google.com/open?id=0B_Ml755QWNOecWtYUGtTWlM2YTA

And the larger debug log:

https://drive.google.com/open?id=0B_Ml755QWNOeWTg0VW93NzVYOU0

And this is an image of the performance graph before-and-after a much better drift align (using PHD2) and performing the PE recording / playback:

https://drive.google.com/open?id=0B_Ml755QWNOec0lCbnJyNEU4aWM

Which got my RA RMS error from 3.5 down to around 2.0 arc-seconds

Note that many things happened along the way that may not be apparent in the guide logs:

Before 11:30 PM it was too windy and I had to start over when it calmed down

 After 2AM I turned the mount off and turned it on again, and lost the PEC control

The most representative data is with the PEC turned on, between 12:42 and about 2AM

This is a 3-minute image without guiding, but with PEC on:

https://drive.google.com/open?id=0B_Ml755QWNOeazdTamp6alV4WjQ

And this is with guiding, PEC on:

https://drive.google.com/open?id=0B_Ml755QWNOed0FNSzl0dFg3clE

The stars are still slightly oblong left-to-right (my RA direction) because of the 2:1 (sometimes  3:1) ratio of RA guide error vs. DEC:

Steve

P.S.  The moon was out, hence the bright background.

[Andy Galasso]

Steven,

The most interesting part of the session was the section that starts at 23:32:38 and runs for 31 minutes.   In that time period you had dec guiding disabled and RA aggressiveness set to 10%, which for practical purposes effectively disabled RA guiding.   In effect, you were doing what the guiding assistant does -- disabling guiding so we see how the mount behaves un-guided.

I'm not sure whether you had PEC enabled at that time, but we can see two primary components of PE. The slowly varying 620 second cycle, 15" peak-to-peak component can be guided out by PHD2 (set your RA aggressiveness back to the default value).

The 11-second oscillation (~4" p-p) is going to be more difficult to guide out. You may have to see if there is anything you can do locate the mechanical source of this oscillation and correct it. We have had other AVX users report this same problem.

Andy

[peter wolsley]

Steven,

Well it looks like Andy beat me to it.  I have been working on a reply for the past few hours.  Everything that Andy said is spot on.  That 11 second oscillation has now become a dominant complaint regarding the Celestron AVX mount.  Andy said it is 4"P-P...I'm tending to say it's more like 2"P-P.  Regardless, it's a big oscillation that needs to be corrected if you want to take photos requiring high magnification.  I know you said that you recorded/playback PEC.  I suspect you did this with the hand controller.  The PECTool application supplied by Celestron can do a much better job and can automatically average the data from several wormgear revolutions which results in a more accurate PEC.  You need to get your mount communicating with your computer.  The following is the E-mail I had started.  It has some additional info you might fine useful.

Peter

Steven,
If at all possible you should invest the time getting your mount's ASCOM interface up and running.  PHD takes full advantage of the ASCOM connection.  I assume that,from the 163mm guider focal length, that you are using a separate guide scope and not an OAG arrangment. If so then PHD can eliminate the need to recalibrate every time you move from one DSO to the next...it knows how to compensate for changes in DEC.  It also allows you to guide via ASCOM which PHD prefers.

Unguided PEC 11:30PM
The 31 minutes of unguided data starting at 11:30pm tells alot about your mount.  First of all the PE of your mount looks like it is roughly 12"P-P.  On my CGEM it is 18"P-P.  What is very significant about this 31 minutes of guiding is that there is also a very fast oscillation happening.  I calculate it to be 2"P-P repeating every 11 seconds. This is a very big, fast and bad oscillation.  PHD cannot do anything about this fast oscillation.  It's my opinion that this oscillation is originating in your mount's RA gearbox. It is most likely the tooth mesh frequency of one of the gears.

Your AVX mount should be able to achieve guiding with a total RMS less than 1".  With my CGEM I can routinely achieve 0.8" and better.  This 2" oscillation is going to cause your stars and DSOs to appear slightly out of focus.  This effect with be most obvious when you are using high magnification.  You wide field (low magnification) photography will look sharp.

Here are some ideas.
-If the AVX is still within warranty then I would talk with whomever you bought it from.  It's not performing as well as it should.
-If it's out of warranty then you can first try to balance the mount perfectly in RA.  Lightly loading the gearbox can sometimes reduce the magnitude of oscillation like this.  I perfectly balance my CGEM in RA and I have good results.
-If the motor/gearbox assemblies are identical for RA and DEC then you can consider swapping the two assemblies.  On my CGEM I can do this and I have done this as a test.  On my CGEM, the RA assembly has a mounting plate that holds a proximity switch but this mounting plate is interchangable.  The large gear on the RA assembly has a special index disk but, again, these large gears are interchangable between RA and DEC.  If you can do this then you might find that the DEC motor/gearbox doesn't have as large an oscillation.

Guided PEC 12:42AM
This first thing I noticed about the guided run started at 12:42AM is that your RA aggression setting was 10 which is very small.  You should be running with aggression values of 60 or higher.  I am only guessing here but...are you using low aggression values to deal with the 11 second oscillation.  I would suggest you try using 70 to 90 aggression combined with 70 to 95 hysteresis.  In your case the large hysteresis values with help to stop PHD from chasing this oscillation. Just be careful not to try 100 hysteresis...this will cause PHD to stop guiding.

I know you believe that PEC was turned on for the 12:42AM guiding session but it looks to me that PHD was still having to move the RA axis quite a bit. Typically when you use PEC you have to record PEC for several cycles and then use the averaged results.  PECtool (by Celestron) is designed to do this but it needs to communicate to your mount via a serial port.  Another reason to figure out how to get ASCOM and Nexremote up and running. 

So, as far as your original topic  "Is it the mount, or is it - me?"  I think it's a combination of both.  The 11 second oscillation is going to restrict how much magnification you can use for astrophotography.  Visual observing will be fine.  You need to get ASCOM, Nexremote and PECtool up and running.  You also need to use higher values for aggression.

[Steven Bellavia]

Thanks Andy!  I meant for it to be 0.  I must have typed the wrong  number, but I think it still showed the problem well enough.

PEC was off for that run.

I am going to try some more tests tonight.  :)

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.

To unsubscribe from this group and all its topics, send an email to open-phd-guiding+unsubscribe@googlegroups.com.

[Ken Self]

If you want to do an unguided run you should go to the Guiding tab on the brain and untick "Enable mount guide output".

[Steven Bellavia]

Oh!  That would have been easier!  Thanks!

[Steven Bellavia]

Hi,

(If this is posted twice, I apologize. The original disappeared from my screen)

Thanks for such a well thought out diagnostic, and well written reply.

I did another test last night (and likely the last of these test-only sessions, though I will continue to tweak every time, as needed).

Guide Log

https://drive.google.com/open?id=0B_Ml755QWNOecmc0UDJsVWVVaWc

Guide debug

https://drive.google.com/open?id=0B_Ml755QWNOebWZrTWJaS1VoLTQ

Summary:

https://drive.google.com/open?id=0B_Ml755QWNOeckFBbnVqMV9PQmc

As you can see in the summary, the best I got was slightly under 2.0 arc-seconds RMS by driving it hard (90%-110% aggression)  and fast (0.5 seconds).

The mount is under warranty, and I received a UPS label to return the equatorial head.  But I was holding off, as I didn't want to get a rebuilt or new head with the same exact issues.  

I ordered a RS232-to-USB adapter cable and downloaded the latest Celestron driver for ASCOM communication.

I'll keep you posted as to what happens next, but thanks again for an immense amount of help.!

Steve

[peter wolsley]

Steven,
If you can return your mount then you certainly have plenty of options to pursue. 

-Looking at your latest guiding log I would guess that your mount has about 60" to 70" of DEC backlash.  At 50% autoguide gain that translates to roughly 8,000mS.  What does the PHD Guiding Assistant say your DEC backlash is?  My CGEM has roughly 60" of DEC backlash.  These very large backlashes are too big for PHD to accommodate and I have found that the backlash logic in the mount has it's issues as well.  My solution was to tell PHD to use it's hysteresis algorithm for DEC.  I use 80% aggression with 90% hysteresis. The net result is that the DEC pulse guiding always starts off with small pulses which will grow in size if DEC doesn't respond.  It's my opinion that what triggers a backlash event is when DEC is given a significant guide pulse that's in the same direction of the overall DEC drift.  This typically happens due to the noise caused by seeing. The DEC axis moves a little farther than it really needed to.  If the DEC drift was always to the North then this "a little farther than it really needed to" will have caused the guide star to now be to the South.  PHD will begin sending south guide pulses which will cause DEC to start moving across it's backlash.  Depending on your Polar Alignment you will see your guide star's DEC position look like a square wave(perfect PA) or a saw-tooth(PA slightly off) as the DEC backlash event takes place.  Using the hysteresis algorithm with lots of hysteresis causes PHD to initially apply small guide pulses which I find greatly reduces the number of backlash events.

-Cranking up the RA aggression and using short exposures is going to be a difficult battle to win.  While you may be able to reduce the 11 second oscillation you are now going to be vulnerable to chasing your local seeing which is completely random and can be fast acting. With the equipment we have, PHD is meant to provide slow trim adjustments with our mounts providing PEC.  I don't know anything about AO (Adaptive Optics) other that it is designed to cater for fast sub-second seeing disturbances and also oscillations like you are dealing with.  It's an expensive option.

-Your 27 minute guiding session starting at 10:27PM demonstrates a really good PEC compensation.  Your mount's PEC was trained very, very well...maybe.

-Your 34 minute guiding session starting at 11:30PM shows unusual events that seems to be occuring every 600 seconds.  You were making adjustments during these events so I can't be sure. They could be a large offset in your mount's PEC data that shows up every wormgear period or they could just be a co-incidence or maybe just my clumsiness in interpreting the guide log .  During each of these three events
23:37:09,  23:47:15 and 23:57:10 PHD sent out guide pulses totally roughly 35"...always in the same direction.  Each time after these brief events PHD quieted down completely and didn't seem to need to make any corrections for PEC. I could be completely wrong about this but it's something to look for next time you are out.

Good luck

Peter

[Steven Bellavia]

Hi Peter,

Yes fast and aggressive is not a good path to follow.

According to PHD2, my DEC backlash 38.0 px (x4.75 = 180") , 21197ms (yikes)

And yes, towards the end of the test session, I was making many changes, just trying to see what happens.

Here is another thought regarding the 11-second oscillation:  

Yes, most  likely the "fast" gears, but could it be cogging of the servo motor?  It is a "Lo-Cog" motor not "No-cog".  But I would think that would be an even faster oscillation.

As I said, I might, eventually, buy anti-backlash gears, just to see if it improves anything, but now I am wondering if it is the motor itself?

Thanks again!

Steve

[Steven Bellavia]

Sorry to bother everyone (especially Peter and Andy) again,

But did I mention, that ever once in a while, I get sub-arc-second guiding / tracking in RA and DEC?

These are the guide logs, and some more screen snapshot,s where on two separate nights I had very acceptable guiding/tracking, and took, perhaps by best astrophoto to date.  And just before that, without changing a single thing, and even pointing in about the same part of the sky, it was all over the place, and I was about to give up for the night.

Guidelog: Good guiding, early AM, Aug 28th, 2016:

https://drive.google.com/open?id=0B_Ml755QWNOeR3d4ZEtPSE5Ucnc

Guidelog: Good guiding, early AM, Aug 30th, 2016:

https://drive.google.com/open?id=0B_Ml755QWNOeS3BPVzh1ZXZoaHc

Performance Summary / screen snapshots:

https://drive.google.com/open?id=0B_Ml755QWNOeZi02MlQ2c0FTRjQ

So now I wonder if this could be gears and not something electronic?  Gears can't fix themselves.  Very interesting (at least for me)

[CyScape]

You mentioned buying "anti backlash gears" in your previous post. Are there after market gears available?

[Steven Bellavia]

There are gears available, but they are not a "direct replacement".  The bore might need to be opened up a little, and you will need to readjust the mesh-distance (but not by much at all, and I believe within the range on the mount - but I haven't tried this yet). :

https://catalog.nordex.com/pc2772_17089

And, of course you'll need the mating spur:

https://catalog.nordex.com/pc595_14197

1/8" is a little thin.  PIC has better choices, but it says 75 day lead time:

http://pic-designcatalog.com/images/pdfcat/section_12.pdf

page 37 for anti-backlash and page 13 for the matching 3/16" face spur.

If I do this, I will let everyone know how it goes!  :)

On Tue, Sep 13, 2016 at 4:05 PM, 'CyScape' via Open PHD Guiding <open-phd...@googlegroups.com> wrote:

You mentioned buying "anti backlash gears" in your previous post. Are there after market gears available?

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.

To unsubscribe from this group and all its topics, send an email to open-phd-guiding+unsubscribe@googlegroups.com.

[CyScape]

Thanks Steve,
 I've been watching this thread closely because I also have an AVX mount. This is my first "decent" mount capable of auto guiding and I've found it can be challenging. I bought my mount in February of 2015 and I've already sent the mount back once... about this same time last year. I was doing pretty good until a few weeks ago my dec in PHD was dropping and then picking back up, do good for awhile and then do it again. Looking at the subs during that exposure you can definitely see where it moves... ruining the sub. Working days and very few good weekends to get the scope out makes it difficult to work out problems. I've noticed when I'm not tracking and I slew my scope in dec, the motor will stop as normal but sometimes continue to quietly buzz. When I bump either direction (slightly) it becomes quiet. Not sure if it correlates to the dec behavior in phd when guiding or not? I've been waiting to get a decent night to run the guiding assistant, get a backlash measurement and do some guiding with default settings for the guys to look at. Basically I'm just trying to access the mount before the warranty expires.

I sometimes look at my graph and wonder...is it me or was this built the day before the Chinese New Year?

Btw...do you have a lot of dew at night when you're out?

To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.

[peter wolsley]

Steven,

I believe that cogging is a DEC issue.  The RA motor is in constant motion and the position regulation of the RA axis is very smooth on my CGEM.  I believe, for your mount, that the RA motor is turning as desired...quietly following it's tracking, PEC and PHD guide pulses.  Unfortunately, there is a gear downstream of the motor that has a very slight issue with the shape of it's gearteeth.  These slight imperfections have a larger effect on guiding the closer these gears are to the wormgear.  On my CGEM I have a 1"P-P oscillation with a period of 15 seconds.  This 15 second oscillations works out exactly to 32 x the wormgear rotation period.  Sure enough I have a gear that has 32 teeth that turns once for every revolution of the wormgear.  I have had nights where the oscillation virtually disappeared and others when it was 1"P-P.  I am still trying to nail down why it happens but I am encouraged that it can be modified.  My current guess is to balance the mount in RA to offload the gear as much as possible. 

I included a photo I took of my CGEM gearbox that indicates the 32 tooth gear that it giving me my oscillations.  I suspect that something similar is causing your issue.

Peter

[Sandy MacPherson]

This sounds consistent with the 20 sec peak to peak i notice on my AVX...600 sec cycle by 30 teeth

--
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-guiding+unsubscribe@googlegroups.com.

[peter wolsley]

Steven,

I have had quiet guiding randomly and yet I believe I have a "suspect" gear giving me grief.  I hope to nail down what's causing this but it's difficult.

Peter

[peter wolsley]

Sandy,

Do you have a guess where this 11 second oscillation is coming from?

Peter

[Steven Bellavia]

Seems to me it is those first two gears in the gear train,as they move the fastest.  The 11 seconds might be one tooth's engagement at sidereal rate.(??)  (I am guessing here - I have to measure a full rotation as soon as I can - with 30 teeth, that would be one full rotation every 33 seconds, which seems fast to me(?))

I am thinking that "fancy" anti-backlash gears may be overkill.  Maybe just replace the two same-sized/pitch gears with higher quality ones ?  And a finer pitch couldn't hurt as it slightly reduces backlash - though not by much - see page 141 of this engineering guide from Boston Gear:

https://www.bostongear.com/pdf/gear_theory.pdf

This is a less expensive option to try.  About $60 plus shipping for two 64 pitch stainless steel gears, and a one week lead time.  I am going to try this first.  All I have to lose is $60 and about 15 minutes to swap them out.

[steve]

Hi
Sorry to jump in late. EQ6 but with similar observations. Sometimes I
have 1" guiding with hardly any PE, sometimes not. Are we perhaps
forgetting the seeing on these occasions? I run the guiding assistant
and accept the values but still it's either bad or good. Just trying to
think of an easy explanation; there must some factors attributable to
the seeing changing between sessions no? Also, it may change during the
night. E.g. here, 10km from the coast, if the wind direction changes
from west (dry: inland) to east (from over the sea) it gets worse. And
vica-versa.
Just a €0.02 worth...
Cheers and clear skies,
Steve

**Rarely bad enough to affect the images though. My long telescope is
760mm.

[Steven Bellavia]

Hi again everyone,

This is a movie I made of the RA motor gear driving the gear on the worm shaft.

I am not sure this explains the 11 second oscillation, but a full tooth movement seems to be about 22 seconds:

https://drive.google.com/open?id=0B_Ml755QWNOeTUtmTjBXc2ZzZHM

Note the bottom left gear is driving the upper right gear.

I am going to try a set of Boston gears, with a 14.5 degree pressure angle, and a AGMA quality rating of 10. (I am not sure what the pressure angle is on the stock Celestron gears - they might be 14.5 or 20).

I also tried meshing the stock gears just slightly tighter, but I can't tell if I actually changed anything, and I haven't tested it outside yet.

Steve

[CyScape]

I watched your video. Question, is that an optical illusion or is the worm gear shaft hole really that egg shaped?

[Steven Bellavia]

I had a light source coming from the right, so I believe you are seeing the shadow.  The bores seemed fairly circular to me.

On Fri, Sep 16, 2016 at 12:44 PM, 'CyScape' via Open PHD Guiding <open-phd...@googlegroups.com> wrote:

I watched your video. Question, is that an optical illusion or is the worm gear shaft hole really that egg shaped?

--
You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guiding+unsubscribe@googlegroups.com.

[Steven Bellavia]

Maybe the gears need to be meshed LOOSER, not tighter?  It looks like I get 3-point contact at around 20 seconds (10 second half-period)

https://drive.google.com/open?id=0B_Ml755QWNOeajU0R2VlMzlFVHc

On Friday, September 16, 2016 at 1:02:37 PM UTC-4, Steven Bellavia wrote:

I had a light source coming from the right, so I believe you are seeing the shadow.  The bores seemed fairly circular to me.

[Steven Bellavia]

Or maybe even 4-point contact, of which on is in the wrong direction, perhaps explaining the oscillation.

[Steven Bellavia]

Hi again,

A friends was gracious enough to invite me to her home and do more testing on the Celestron AVX mount, but this time with all my equipment (scope, guide scope, camera, etc),on one of her AVX mounts.

The results are very interesting.

The good news is that my mount is no different than hers.  The bad news, is the same news.  So they are all like this. Nothing is "wrong" with my mount.  It's just not great at RA guiding/tracking. 

But more good news! We found a way to guide it fairly well.  Against the common thinking of other PHD2-guiders, this mount responds much better to "fast" and "hard" RA guiding.  That is, very short exposure times, and high proportional gain ("aggressiveness").

See my  report:

https://drive.google.com/open?id=0B_Ml755QWNOeS3ZKLWdnT19PNE0

and here is the guide log for the entire night's worth of testing:

https://drive.google.com/open?id=0B_Ml755QWNOeZWgyZXJEZDUzcnM

Steve

[bw_msgboard]

Well, if you’re happy, I guess that’s all that matters.  Personally, I wouldn’t be too excited by a guiding RMS of 1.8 arc-sec, which is what you got in the last 52 minutes of your log.  But I suppose if you are doing wide-field imaging, maybe you can get away with it.  I think the results you’ll get will look much like those of a better mount operating under truly hideous seeing conditions, something most people avoid.

 

Good luck with it,

Bruce

 

---

--

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.

[Steven Bellavia]

Bruce,

I never said "happy" I said "acceptable"  there is a big difference!!  :)

But seriously, Celestron has been very good with all this, as they have asked for all my test results and truly believe I have a bad motor, and that my friend does too. Perhaps a bad "lot"?  I will be sending the head back for a replacement motor this week.

I should not have stated that they are "all" the same.  It is possible that some are much better and there are a few bad ones out there.  It is not my place to recommend or bad-mouth Celestron at this point.   I would love to see other AVX user's logs / plots (in arc-seconds, not pixels).

And at 600mm focal length, I would like 1.0 arc-second RMS error total for a 5-minute sub.  I was "almost" getting that with the "fast-and-hard" settings, with 1.4" for a 5-minute duration ( I think?)

Steve

[bw_msgboard]

I’m glad to hear Celestron is working with you on this.  Since they presumably ship these things in large quantities, there’s probably no way for an end-user to really know what’s typical for performance.  But if they think there’s a problem with the motor, they must also think your results aren’t typical.  If it proves to be a fixable hardware problem, you may be able to dial back the settings to something more reasonable.  If you’re imaging at 600mm focal length, you may have a sporting chance with the mount. <g>  I hope so.

 

Bruce

 

---

[Steven Bellavia]

Well, at least now I know it's not the gears  :/

The performance with and without guiding seems just about identical.  Maybe a tad bit smoother, but the same RMS, and the same peaks, more or less.

I had better polar alignment with the original metal gear test, and had DEC guiding on for the non-RA guiding.  I also had less points (100 versus 400) for the metal gear test, but otherwise, the RA tracking performance is the same for either gear set, as far as I can tell.

So it must be the motor itself.  (The frequency of the oscillations is much too fast to be the worm gear, unless the individual grooves have burs or something).  It is supposed to be a "lo-cog" motor.  But that does not mean "No-cog", so perhaps this is cogging of the motor (?)

But as I said, Celestron is taking the head back and putting in another motor.  So we shall see.

https://drive.google.com/open?id=0B_Ml755QWNOeY0xhcHRxWFgtUTg

Steve

[Steven Bellavia]

Hi again.

I have now tested a 3rd Celestron AVX mount (I borrowed it, while mine is being "repaired").

The performance seems about the same as the last two I have tested. (which was kind of disappointing to me).

But one thing they all seems to have in common is getting better as the night goes on. 

Many people have told me that it is because of balance, and that as the scope moves, it gets to a better balance point.  The "test" I did Thursday night definitively disproves that theory, as the scope was re-pointed to nearly the identical direction, and the guiding was noticeably different later.  So it is definitely not balance.

Any thoughts why three of the mounts I have tried and tested, get better towards the very wee hours of the morning?  Maybe grease warming up?  There was no change in wind or seeing.  It was still with no wind, and the seeing was the same throughout the night.

Note this AVX is older than mine, or the other one I tested. (The hand controller required me to go to scope setup to enter Lat-Long, versus the current version where you just press "back" when starting up).

My "report":

https://drive.google.com/open?id=0B_Ml755QWNOeSUt6TDYtZ20zRDA

and this is the guide log for the entire night (I imaged from dusk to dawn):

https://drive.google.com/open?id=0B_Ml755QWNOeVmpuSnI2OFNRbUk

Steve

[Andy Galasso]

Hi Steve,

I agree that balance is not the issue, but I would not rule out seeing as the explanation.  Here are two 6-minute sections of your guide log, one at 21:30 and another at 04:20, same vertical scale.  The Dec RMS was 1.20" early in the evening but improved to 0.85" in the early morning sample.  The magnitude of the the high frequency oscillations is a direct measure of the seeing conditions.

I noticed that you are using 0.5 second exposures. I think this is much too short and pretty much ensures you will be chasing the seeing. I would not recommend less than 1.0 second exposure. The longer the exposure the better to avoid chasing seeing. Here's some data showing how increased exposure time reduces Dec RMS error (un-guided).

We see that longer exposures decrease the seeing effects. However, your exposures need to be short enough to be able to correct the mount's RA periodic error. The Guiding Assistant will provide a recommendation for exposure time that is as long as possible, yet still short enough to correct for PE.

Andy

[Steven Bellavia]

Hi Andy,

I have tried 1 and 2 second exposures and it just doesn't work as well as the 0.5 second ones.  This has been a fairly consistent result.  It seems the mount has short period issues that the longer exposures can't control.  If you look further up on this thread , I found that "fast-and-hard" was the only thing that works for me (so far). 

--

You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/DP0oEhp4_gk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guiding+unsubscribe@googlegroups.com.

[Andy Galasso]

Steve,

Fair enough, sounds like you are between a rock and a hard place: too short exposures and you chase seeing, too long and you cannot tame the rapid PE.  I think this will always be a struggle unless you can fix the PE problem. Then again, if you are imaging at shorter focal length (wider field imaging), the guiding difficulty may not matter so much.

Good luck,

Andy

[Steven Bellavia]

Thanks Andy!  I am always amazed at how knowledgeable and helpful everyone in this group is, and it is so much appreciated.

I am thinking of upgrading to an Orion Atlas, or a CGEM II, but want to try to get the AVX to work, because I actually really like it.

[Steven Bellavia]

Hi all,

I got my re-worked Celextron AVX equatorial head back from Celestron already. (A very fast turn-around).

Quick summary:

- It sounds much better at high speed slews (smoother, quieter)

- RA tracking did not change all that much (?)

- PEC makes a significant difference, and certainly worth doing.

- "Fast-and-hard" still seems slightly better, but maybe now I can go back to 1.0 second instead of the risky 0.5 second exposures I was doing.

and finally,

- On a night were the seeing was "iffy" (chart said 4/5, but data shows otherwise), I was able to get to 1.3 arc-second total RMS error for 10 minutes, on a $799 mount.  I guess I should not complain (?)

So I am still not sure if it is all that much better or different, but I think this concludes all my testing.  I will either accept this performance, and that 600 mm is my maximum focal length on this mount (the RMS guiding error is about the same as my single pixel resolution), or I may consider going to a "better" (and likely heavier) mount, like the CGEM II or Orion Atlas, though now we are talking 2X the cost and perhaps 2X the weight too.

A PDF summary is here:

https://drive.google.com/open?id=0B_Ml755QWNOebWpBeVVpZ3U5Ymc

And my guide log for the night here:

https://drive.google.com/open?id=0B_Ml755QWNOeeExZLWZNQ2h6b1E

As always, I want to reiterate how awesome everyone on this group is, and how much fun I had doing this.  I will continue to read and post here as often as I can.

I am sincerely grateful to all of you! (Andy, Bruce, Peter, and I apologize for anyone else I skipped - feel free to add your name to the list!).

Steve

[Andy Galasso]

Steve,

Thanks for following up and letting us know how it turned out.  It's great to hear that Celestron was able to re-work the mount for you and it seems to be at the point where it should be a decent platform now for short to moderate focal-length imaging.

Andy

[Steven Bellavia]

Hi again,

I know I already posted that my AVX mount is working well since it was re-built by Celestron, but I just wanted to reiterate that I am now getting 1.0 arc-second total RMS error on many occasions, for 5 minute and longer exposures.  Of course, this is seeing dependent, but I don't have to do "fast-and-hard" anymore, and can use a "normal" 1.0 second exposure and 80 to 90% aggression.  A typical (well, maybe a "good") guiding sessions screen capture is here:

https://drive.google.com/open?id=0B_Ml755QWNOeLUJFWE9TamlFZUk

And it seems things are working well enough, so no diagnostic needed, but if anyone is interested, the guide log is here:

https://drive.google.com/open?id=0B_Ml755QWNOeTUIxZ0pxZ3FQY1U

Also, the mount's PEC is great.  You can really see the difference before and after playback (though PHD2 handles it well enough if I don't use it)

And these are two images from that night, where the seeing wasn't ideal, but I still had good guiding (and awesome drift align), thanks to PHD2  :

https://www.flickr.com/photos/125134422@N06/30169604543/in/album-72157670340137352/

and

https://www.flickr.com/photos/125134422@N06/30806549155/in/album-72157670340137352/

And yes, I need a field flattener, and to do flat frames.  That was first light for my ED80 APO.  But stars are tight and round near the center, showing that the guiding was adequate.

:)

Steve

[Andy Galasso]

Steven, Thanks for the update, glad to hear the setup is working well for you.  Congrats on the images.

Andy