Trouble with Backlash on Rebuilt EQ6R Pro Mount

[Tim Stevenson]

I have been struggling with getting the backlash correct on a rebuilt EQ6R Pro mount.  I made several adjustments last night and measured each with the Guiding Assistant.  My logs are attached as is my last Guiding Assistant results (1310ms backlash) and my first result (5399ms).  I started making backlash adjustments and was sneaking up on a good backlash until I think I went too far and created binding.  I think it was binding because calibration and guiding assistant failed (well maybe not failed but the mount didn't act right).  So I backed off some and started the process over.  The last guiding assistant can be found in section 25 of the log file with timestamp 01:23:43.  Two calibrations follow and each one produced DEC guiding rates less than expected.  Note that other calibrations earlier in the evening were normal.  Frustrated, I decided to set up an imaging session and see how the guiding turned out.  The first guiding session is in section 31 of the log, then a meridian flip followed by section 32.  Guiding is not perfect but seems reasonable (~0.75" Total RMS).  BTW, I am making VERY small adjustments in backlash, ~ 1/16th of a turn.

Along with your recommendations for my current situation, I would also like to know the telltale signs of backlash vs binding.

Thank you for all your support to me and everyone else over the years.

https://openphdguiding.org/logs/dl/PHD2_logs_ZpCs.zip

Final Guiding Assistant

First Guiding Assistant

[Bruce Waddington]

Hi Tim.  I'll get back to you on this later today but I think you're actually in pretty good shape here.  It's good to see you took an organized approach to the problem.

Regards,

Bruce

[Bruce Waddington]

It looks like you've done a good job here, I think you should leave the Dec adjustment alone now.  Looking at your two guiding sessions, the Dec RMS is lower than RA so the Dec mechanics aren't limiting your results.  Since you have such an accurate polar alignment, PHD2 didn't need to guide in Dec at all except for handling the frequent and rather large dithers.  Some of these are 30 arc-sec dithers - why do you need to make them so large?  Remember, you need to calculate the dither size you need on the main image, not on the the guider image.  Your guider image scale is quite coarse, nearly 6 arc-sec/px, so I question the need for such large dithers.  Even with all that, the Dec guiding recovery after dithers is quite good and you aren't getting long settling periods or settling timeouts.  Just looking at these two guiding sessions, I would say you were probably guiding at the limits of your seeing conditions at the time.  

Stiction on the Dec axis is most easily seen during normal guiding sessions as opposed to the GA backlash test.  Here's an example of stiction - not from your mount. 

You can see that a reversal in guiding direction results in a delayed mount response followed almost immediately by an over-shoot.  The behavior is more pronounced when the initial guiding correction is larger.  This is usually an indication of static resistance on the Dec axis.

Regards,

Bruce

[Tim Stevenson]

Bruce ... Thank you for your response.  Yes, the dither requirement is too large.  I had a previous conversation on this forum about high dithers back when I was first starting to dither.  I went to change all my NINA profiles to a 1 pixel dither and that would translate to a 1 to 5 pixel dither on the guide image depending on which telescope I was using at the time.  This NINA profile slipped through the cracks and had a 5 pixel dither yielding 25-30 pixels in the guide image.  It has since been corrected.

In a separate question, what are your thoughts about PPEC and the best way to handle it?

Thank you

Tim

[Bruce Waddington]

I'm not real clear on what you're asking.  There's not much to "handle" with PPEC, and you have nothing to lose by trying it inasmuch as your setup is pretty well-behaved.  I expect it will improve your RA results a bit.  The frequency spectrum on your mount is pretty clean:

This is a fairly short guiding session so the frequency measures aren't exact.  The larger peak to the right shows a 4 arc-sec peak-peak tracking error at about 480 seconds, which is quite good.  I assume the native worm period is probably 480 seconds.  I would be inclined to attack the second highest peak, the one showing a 1.8 arc-sec peak-peak error at 119 seconds.  If that's in reality a harmonic of the worm period. it's probably a period of 120 sec.  So you could enable PPEC, set the period length to 120 sec, and un-check the "auto-adjust period" checkbox.  Then run it for a few nights and compare the results.

Regards,

Bruce

[Brian Valente]

PS - if you're looking for specific PPEC settings, you can try setting the period to 480 seconds (the public worm period, i believe) and disable auto-adjust period. 

it takes a couple cycles to get fully up to speed, after that you should see some improvement in RA.

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/d50eaf18-0ee2-429b-a3a0-d7a0b1ea94c6n%40googlegroups.com.

--

Brian 

Brian Valente

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

[Tim Stevenson]

Bruce and Brian ... sorry, I should have been more specific on my ask about PPEC.  As I understand it, there are two ways to handle PPEC, one in with the mount (EQMOD or GSS) and the other is with PHD2 using the PPEC algorithm.  So my question is which method would you recommend?  I'm using GSS at the moment (just recently switched to GSS).  I did use EQMOD and had a correction curve built on the data of several PEC recordings (not an easy task).  I get the impression that GSS will actively keep learning to improve PPEC after the initial recording. 

Regards,

Tim

[Brian Valente]

Tim

Generally periodic error correction can be handled either by the 1. mount, 2. by phd, or 3. both. How well the mount handles the first or third depends on the mount's software implementation. 

PPEC (Predictive PEC) is a specific algorithm in PHD (that's the #2 mentioned above). I think that's what Bruce was suggesting, partly because it's the easiest way to apply a periodic error correction without having to go into your mount or learn about its capabilities. You can just apply the algorithm and see if it helps your periodic error. If it does, then you can decide if you want to use the mount's PEC or a combination to further improve. 

But it's a very simple way to test for improved RA tracking

Brian 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/e2f481d6-4ce6-4827-9b7e-57040cbaea27n%40googlegroups.com.