AVX Guide Rates

[Juno16]

Hi All,

I know that there are several Celestron AVX imagers out there.

Has anyone changed the default guide rates for DEC and RA?

Since I have been guiding (about a year), I have had pretty decent guiding success with my mount. No, I don’t have perfect guiding by any stretch, but it is good enough with sometimes worse than others.

Since I have been dithering (maybe six months), I have enjoyed its great benifits.

However, when I get a large dither that rocks the DEC axis, my mount (and PHD) really struggle to get things under control. Huge DEC backlash in is the culprit. It might take over 30-40 pulses from PHD to get the DEC axis under control.

I have read online about changing the autoguide rates from the default 50% to 99% (the highest I can go) to help the mount recover from dithers and guiding upsets more quickly.

I made the changes and have high hopes that tomorrow will be the first opportunity of the year to get out under the stars.

Of course, I plan to re-calibrate PHD and run the GA.

Have any other AVX imagers tried changing the autoguide rates and what was the outcome?

Thanks,
Jim

[JayTee]

Based on knowledge obtained from the old forum I had that info before I got the mount. So from day one I've been using 99%. And it still takes a while to recover from the huge backlash.

Cheers,
JT

[Juno16]

Based on knowledge obtained from the old forum I had that info before I got the mount. So from day one I've been using 99%. And it still takes a while to recover from the huge backlash.

Cheers,
JT

Thanks JT,

So, I guess its worth a try. If going to 99% from 50% will increase the mounts response time by 2x, it surely can’t hurt.

Hopefully I will get a try tomorrow!

Thanks,
Jim

[SkyHiker]

Did you calibrate the AVX backlash? It can compensate for backlash in ways that PHD2 can't, by cranking the motors real fast. If the DEC backlash causes your problem maybe that could help.

[bobharmony]

Thanks for that reminder, Henk. Yep, that's what I did after trying to mesh to DEC worm gear better. It does work, Jim!

Bob

[Juno16]

Did you calibrate the AVX backlash? It can compensate for backlash in ways that PHD2 can't, by cranking the motors real fast. If the DEC backlash causes your problem maybe that could help.

Thank you Henk!

Dang, dang, dang (only that isn’t what I said!).

I did calibrate the DEC backlash in the mount some time ago (probably after Bob’s [mention]bobharmony[/mention] post), so I thought that I was good there. BUT, I ran into some trouble about a month or so ago and did a factory reset! Wow, I forgot about the backlash settings!
I didn’t record the settings, but I am pretty sure that both DEC positive and DEC negative were 20 or close to it.

I might set them at 20 and leave the autoguide rate at99 and see how it goes.

Thanks again for the reminder Henk!

Jim

[JayTee]

Did you calibrate the AVX backlash? It can compensate for backlash in ways that PHD2 can't, by cranking the motors real fast. If the DEC backlash causes your problem maybe that could help.

Not exactly true Henk. This quote is from the PHD2 User Guide and Best Practices.

Backlash settings in the mount are generally not going to work for auto-guiding because they lead to over-shooting. I think they are generally intended for visual use where the tolerances are much greater. With PHD2, you should disable all backlash settings in the mount. Then run the Guiding Assistant and let it measure the backlash directly. If the backlash is not too large, the GA will recommend a backlash compensation option – give that a try. The difference here is that PHD2 tunes the size of the compensation pulse based on the results it’s actually getting – it’s not a one-time, constant value. Of course, you need to look at your guiding performance in totality to know whether Dec backlash is limiting your overall results.

I do not use Backlash Comp in my AVX, I let PHD2 do it.

Cheers,
JT

[SkyHiker]

Did you calibrate the AVX backlash? It can compensate for backlash in ways that PHD2 can't, by cranking the motors real fast. If the DEC backlash causes your problem maybe that could help.

Not exactly true Henk. This quote is from the PHD2 User Guide and Best Practices.

Backlash settings in the mount are generally not going to work for auto-guiding because they lead to over-shooting. I think they are generally intended for visual use where the tolerances are much greater. With PHD2, you should disable all backlash settings in the mount. Then run the Guiding Assistant and let it measure the backlash directly. If the backlash is not too large, the GA will recommend a backlash compensation option – give that a try. The difference here is that PHD2 tunes the size of the compensation pulse based on the results it’s actually getting – it’s not a one-time, constant value. Of course, you need to look at your guiding performance in totality to know whether Dec backlash is limiting your overall results.

I do not use Backlash Comp in my AVX, I let PHD2 do it.

Cheers,
JT

That is just plain wrong. The way in which the AVX compensates is equivalent to a system that has less backlash, looking at it from the outside as PHD2 does. Note the "if". The DEC backlash of an AVX (well mine at least) is horrendous. PHD2 will not spin the motors faster so it will just suffer from any delay due to backlash, and there is a lot of that. I always have the AVX backlash compensation enabled and it seems to work. It must be tuned carefully so there is no overshoot, even with PE.

[Baskevo]

I don't think you are supposed to use the backlash compensation with PEC... Which I found to be a huge benefit! When I trained and turned on PEC in CPWI, and then used the PPEC algorithm, it made a huge difference.
I was getting less error messages in PHD2 about DEC backlash and my guiding was about 50% better.

Just a tip, you have to turn on PEC in CPWI or however you're doing it every time you turn on the mount.

[JayTee]

That is just plain wrong. The way in which the AVX compensates is equivalent to a system that has less backlash, looking at it from the outside as PHD2 does. Note the "if". The DEC backlash of an AVX (well mine at least) is horrendous. PHD2 will not spin the motors faster so it will just suffer from any delay due to backlash, and there is a lot of that. I always have the AVX backlash compensation enabled and it seems to work. It must be tuned carefully so there is no overshoot, even with PE.

Well, I can only go by what the guys who wrote the software have to say on the subject. Here is the entire section on Dec Backlash Compensation out of the PHD2 v2.6.6 User Guide

Declination Backlash Compensation
Most commonly-used mounts have some amount of backlash in declination. This causes delay whenever there is a change in the direction of the
Dec guide commands. During this interval, the declination gears aren't fully engaged and the axis doesn't move in response to the guide commands.
Many mounts have settings for backlash compensation but these should not be used for guiding - they are typically intended for visual use where
high precision isn't required. Because the actual amount of compensation needed at any given time may depend on the pointing position and the
mechanical load on the system, a fixed value will usually result in oscillations that never stabilize. The backlash compensation implemented by
PHD2 is adaptive, meaning that the compensation amount is adjusted up or down depending on the measured results. Before enabling this feature,
you should run the GuidingAssistantand measure the declination backlash - the time delay required to fully reverse direction in declination. Keep
in mind, the higher the guide speed setting in the mount (e.g. 0.9x sidereal), the smaller this delay will be. If the measured amount is 3 seconds or
less, the GuidingAssistant will recommend trying backlash compensation. If you apply that recommendation, the backlash compensation settings
will be handled for you automatically. The UI controls for backlash compensation include settings for 'minimum' and 'maximum' compensation
amounts. These effectively limit the range of the adjustments that are made to the starting compensation value. If you'reexperienced with your
mount's behavior, you can adjust these settings manually to be sure that the compensation stays within a range that you know works well.
Otherwise, you should just leave these at their default values. The backlash compensation algorithm will generally work well if the backlash is less
than a few seconds and the mount doesn't have other significant mechanical problems. You should expect a short period of instability when
guiding starts because the initial state of the Dec gear system is unknown - just let it stabilize before you actually start imaging. If you see recurring
periods of Dec oscillation or the axis won't settle down, disable the compensation feature and submit your debug log file to the PHD2 support
forum. It's important to note that the correct amount of Dec backlash compensation will often change depending on the scope pointing position.
This can be caused by uneven gear wear on the Dec axis or by differences in the gravitational load being applied to the axis.

Cheers,
JT

[Graeme1858]

Since I have been guiding (about a year)

Jim

I'm about 11.5 months behind you on the guiding front, so thanks for asking the question and generating the replies. It's hugely helpful.

Regards

Graeme

[bobharmony]

That is just plain wrong. The way in which the AVX compensates is equivalent to a system that has less backlash, looking at it from the outside as PHD2 does. Note the "if". The DEC backlash of an AVX (well mine at least) is horrendous. PHD2 will not spin the motors faster so it will just suffer from any delay due to backlash, and there is a lot of that. I always have the AVX backlash compensation enabled and it seems to work. It must be tuned carefully so there is no overshoot, even with PE.

I have to go with Henk on this one. I read that portion of the manual when I first started with the ASGT and PHD2 and resisted any sort of mount configuration to control the DEC backlash. I tried to address the issue with mechanical adjustments only, but continued to have issues with backlash and dither settling time.

As a matter of desperation after I had to shut dithering off while doing a session early last year, I decided to try the mount backlash adjustment in addition to PHD2 - it reduced my settling time, and helped with overall guiding. I rarely have to throw a sub away anymore, and I am able to spend more night sky time collecting data, and less time waiting for the mount to settle. I report my experience here, which means as much to me as what's written in the manual.

Note: the ASGT is the predecessor to the AVX - YMMV.

Jim, sorry for sidetracking your thread.

Bob

[Juno16]

That is just plain wrong. The way in which the AVX compensates is equivalent to a system that has less backlash, looking at it from the outside as PHD2 does. Note the "if". The DEC backlash of an AVX (well mine at least) is horrendous. PHD2 will not spin the motors faster so it will just suffer from any delay due to backlash, and there is a lot of that. I always have the AVX backlash compensation enabled and it seems to work. It must be tuned carefully so there is no overshoot, even with PE.

I have to go with Henk on this one. I read that portion of the manual when I first started with the ASGT and PHD2 and resisted any sort of mount configuration to control the DEC backlash. I tried to address the issue with mechanical adjustments only, but continued to have issues with backlash and dither settling time.

As a matter of desperation after I had to shut dithering off while doing a session early last year, I decided to try the mount backlash adjustment in addition to PHD2 - it reduced my settling time, and helped with overall guiding. I rarely have to throw a sub away anymore, and I am able to spend more night sky time collecting data, and less time waiting for the mount to settle. I report my experience here, which means as much to me as what's written in the manual.

Note: the ASGT is the predecessor to the AVX - YMMV.

Jim, sorry for sidetracking your thread.

Bob

No problem Bob. It is all very good information. If you don't mind, can you post a link to your thread a while back when you talked about setting your backlash compensation using the hc. I tried, but can't find it.

I also set my backlash compensation, but I really don't recall if it made much difference with guiding. This is a very good discussion with strong evidence for both methods (with and without backlash compensation in the mount).

Unless I change my mind, I will approach this in two steps as I don't want to include too many variables if I am lucky enough to get out tonight.

I set the autoguide rates to 99% for DEC and RA.

I will not set backlash compensation in the mount yet.

I will be trying out my new idas lps-d1 light pollution filter also.

Thanks for the great information and please continue to comment if thoughts arise!

Jim

[JayTee]

Out of curiosity Bob which version of PHD2 are you using? The current version is 2.6.6 Dev 4.

I ask this because dec backlash compensation is an evolving feature. They keep making the algorithm better and better.

JT

[bobharmony]

Out of curiosity Bob which version of PHD2 are you using? The current version is 2.6.6 Dev 4.

I ask this because dec backlash compensation is an evolving feature. They keep making the algorithm better and better.

JT

JT - I had to check because it's been so long since I've had a night out. I am at 2.6.5 dev5. If they come up with a version that can tame the AVX's backlash I am all for that!

Bob

[Juno16]

Hey All,

I got a chance to get out tonight! First time this year.

Well, it is very early into the evening, but so far, the experience has been very interesting.

I created a new profile in PHD2 with the new autoguide setting of 99% for both DEC and RA.

Got home and got things rolling about 7:30. First off the PHD calibration was my best yet.

Then I ran the Guiding Assistant. The first time that the DEC backlash was successfully measured. I always knew it was huge, but now it was measurable as such! Yes, 13,000ms!!!!

Focused on Capella, then plate solved to M78. Marius's image the other day inspired me. [mention]MariusD69[/mention]

Besides being a test on the new autoguide rate, this was also a first light on my new IDAS LPS-D1 filter.

Shot a test image at my usual 150 seconds at iso 800. Way underexposed.
Shot a test image at 240 seconds. Still underexposed.
Shot a test image at 300 seconds and the histogram looks okay. Definitely not overexposed. Maybe could use a bit more, but since I have never pushed my AVX to 300 second images, I left it there. Strange seeing subs with a blue cast instead of the modded dslr red cast.

With only about 10 images so far, I see very good response to large dithers in DEC. At least so far anyway. With the 50% mount autoguide settings previously, sometimes after a large DEC dither, guiding would take a very long time to stabilize. I would sometimes increase the range of the guide graph to see if I could find the DEC red graph lines. Even with the long stabilization time, sometimes the next exposure would start before the DEC had completely recovered.

I am sure that I will see some upsets as it is early in the evening and that is usually how things go!

The new filter is letting me get 2x the exposure time which greatly increases the ratio of imaging time/total time (including dithers).

Very interesting night so far!

Jim

[bobharmony]

No problem Bob. It is all very good information. If you don't mind, can you post a link to your thread a while back when you talked about setting your backlash compensation using the hc. I tried, but can't find it.

I also set my backlash compensation, but I really don't recall if it made much difference with guiding. This is a very good discussion with strong evidence for both methods (with and without backlash compensation in the mount).

Unless I change my mind, I will approach this in two steps as I don't want to include too many variables if I am lucky enough to get out tonight.

I set the autoguide rates to 99% for DEC and RA.

I will not set backlash compensation in the mount yet.

I will be trying out my new idas lps-d1 light pollution filter also.

Thanks for the great information and please continue to comment if thoughts arise!

Jim

Jim - I initially set this up and tested it in April, and I didn't join TSS until May, so the post is probably over at the old place. My imaging night in April was 4/6/2019 and my target was M51.

If you are daring, feel free to take a look over there. When I try to get Firefox to load the page, it warns me about security risks due to a bad cert, and strongly suggests avoiding the site.

I did discuss the effects a little in a post here near the end of September while imaging the Iris nebula. IIRC correctly the post has something about APT in the title, as that was my first attempt with that software.

My experience in April was that I got my best calibration ever (orthogonality off by .3 degrees) and dither settle times were reduced from 90-120 seconds down to about 45 for large dithers. I was not able to measure backlash in the Guiding Assistant at that time.

Bob

[JayTee]

JT - I had to check because it's been so long since I've had a night out. I am at 2.6.5 dev5. If they come up with a version that can tame the AVX's backlash I am all for that!

Bob

Hi [mention]bobharmony[/mention]

So the unfortunate thing for me is that the way I read those guidelines, is if your backlash is greater than 3 seconds (>3000ms) you're SOL until you can tame it and bring back to less than 3 seconds. Right now, my AVX Dec backlash is around 14.5 seconds (HUGE and unmeasurable!), so I'm in the SOL category until I fix it or forsake this mount. I'm leaning toward forsaking!

Cheers,
JT

[JayTee]

Very interesting night so far!

Jim

Congrats on the newer better calibration numbers!

Since you are experimenting, don't hesitate to use ISO 1600 and see what your noise levels really are. I know you are using a Nikon and not a Canon otherwise, I would unequivocally say use ISO 1600. Who knows, 1600 may be the way to go with the new filter in place.

Cheers,
JT

[Juno16]

Very interesting night so far!

Jim

Congrats on the newer better calibration numbers!

Since you are experimenting, don't hesitate to use ISO 1600 and see what your noise levels really are. I know you are using a Nikon and not a Canon otherwise, I would unequivocally say use ISO 1600. Who knows, 1600 may be the way to go with the new filter in place.

Cheers,
JT

Yes, JT. I feel your pain with my 13.1 second backlash. At least I could measure it!

I am experimenting, but I will leave the new autoguide setting and the filter as tonight's variables. Maybe I will give iso 1600 a shot another night though!

For my Nikon, unity gain is just a hair over iso 200. I thought that I would strive to get lower in iso to reduce the noise, but I am actually having pretty good guiding at 300 seconds even near the meridian (for my mount). I definitely don't want to push higher than 300 seconds just yet.

I have read that Canon's do well at high iso's. If I can get good performance at 300 seconds at iso 800, what would be the advantage of trying iso 1600?

Thanks,
Jim