PEC for AVX --- total failure

[saiph]

Here's one for y'all: what's a fellow stargazer to do to receive an answer, or any useful feedback, on a subject where are literally dozens of skilled and knowledgeable users?

 

I just returned from another failed attempt -fourth, and probably my last one - at PECing my AVX mount. This time i tried the inverted curve provided initially by Metaguide and PECprep, thinking that maybe last time it was wrong since the PEC graph and the PE graph looked pretty much identical; PECtool managed to load it snd upload it to the controller, but none of the curves tried so far seems to work. Nothing works!!..... HOW EXACTLY DO YOU, PEOPLE, EVEN PEC TRAIN YOUR AVX MOUNTS??

 

Seriously, i'm pretty desperate here, and horribly frustrated by this lame failure. Please, PLEASE, anyone (preferably knowledgeable), tell me what to do!  I see older topics where people flex their astro-muscles, with success stories. I find it hard to believe that absolutely everyone switched to guiding lately, or got into a mass amnezia.. Unless, of course, no one gives a rat's @$s about a fellow stargazer's trouble. In which case i might as well sell my mount, sell everything i have and quit astro for good, because i see no point in struggling like a mad man, with no help. I also tried private messages - got totally ignored. What more can i say....

 

 

 

 

 

[daz]

It's entirely possible that very few people are doing this - hence the lack of responses.

I can only find 1 thread dating back to '15 but this discusses PEC itself, not the how-to - here. Even CN only has a couple of references to it...

 

And, just my own take, but stamping your feet because you haven't had a response is pretty much guaranteed to ensure you don't get any!!

Why don't you post what you have tried, with the results - perhaps someone who knows about PEC in general may be able to help? If I did, I would, but this is not something I've done with my AZ-EQ6.

 

[michael8554]

I tried to improve the PEC of my Meade LX200 GPS by repeated training of the PEC.

What I didn't realise at first was that the curve displayed by most softwares is a curve of the amplitude of the corrections applied to the drive. Only PemPro type programs show the amplitude of the actual error, which you can then tweak and feed back to the drive as a correction. Simply inverting the corrections curve will effectively cancel the corrections , not improve them.

That's my understanding, but may well be wrong !!

Michael

[saiph]

Gee, thanks a lot Daz great help. Now I know exactly what (not) to do when another fellow stargazer will be in need. That's the spirit of the community: sit back, relax, do nothing, let the newbie bust his head, he will leave eventually.

 

Anyho'... getting back to my splinter here.

 

Just for the record, here's what I did so far:

 

 - leveled, aligned and polar aligned the mount, chose a suitable star for guiding (couldn't lock on to a proper Dec.~0 star, so I went for the next best thing - Rasalhague)

 - gathered PE data from the mount using a C80ED, a LifeCam webcam (I only adjusted it to fit into the focuser, no LE mod or anything) and Metaguide - which resulted in a log file; I do not have another more suitable guidecam with autoguiding capabilities, which is probably why PHD or GuideDog failed to get along with what I have..

 - I didn't exactly have all the time in the world to gather more than 4 RA worm cycles, so I'm guessing the data set is close to irrelevant, although it shows some variety; the opportunities to gather more data are also scarce, since my laptop would simply die after a few minutes unplugged;

 - also performed a one-cycle PEC training session using PECtool (just out of curiosity, and not having any information about how exactly this step needs to be done, nor what exactly happens when performing it), which resulted in a CSV file I downloaded from the mount, just to see the data structure inside it; unfortunately, there was only blank data, a pure -1..+1 sine wave data, which rather confused me in the end;

 - used PECprep and Excel to "hairstyle" the 4-cycle data set into a package that PECtool would recognize and upload to the mount's controller - that is, chop all 598 data points down to only 88 points (sampled more or less equally), and limited the values to a -1..+1 range (as I noticed from the default CSV file exported by PECtool after the default so-called one cycle PEC training session; big mistake); this resulted in more or less nothing when played back the PEC recorded in the controller; PE was still all there;

 - inverted this curve, as I figured maybe PECprep provides the PE instead of PEC, since they look incredibly similar (I find it hard to believe that something called PEC curve can cancel out a PE curve that goes the same way; unless the mechanism behind the PEC playback inside the controller is not addition but subtraction... no one would tell...); this, again, resulted in a big nothing;

 - then I realized "hey, wait a minute - this guy here shows his PECtool window with a graph spanning in arcsec, not -1/+1 values"; so I went back to PECprep, tuned the frequencies a little bit more, exported the smoothed correction curve arcsec data spanning roughly -20"..+25", prepared it for upload, played it back under the stars, and then something happened  -  the error increased, almost doubled! (oh, the joy, the happiness.....)

 - went back, prepared this file one more time, with the inverted arcsec values (logically, if a data set produces the opposite effect, you fix the data set accordingly. right?), but guess what - the error is still there, not doubled, but not exactly the same amplitude as with no PEC, however it won't even allow a lousy 30s exposure time, let alone a 2-3 min that would seem reasonable for a 480mm FL; residual error is reported to be around +/-4", with a <2" RMS.

 

So, all that being said, i'll stop stomping my feet so I won't scare the fish away   (pun intended), drink my tea, bow my head to all you fine gentlemen, and prepare the sale ad, just in case no useful feedback shows up. Should anyone care to give me a hint on what I've done wrong, or what else I need to do (except guiding, of course) to get it right, I'd appreciate it very much and I'm willing to give it one more try.

 

 

Clear skies!

 

[newbie alert]

I was reading the other day about a procedure that someone was doing to try to pec train their avx and thought ...STUFF THAT.... glad I'm using phd and that's enough on my little brain so far...maybe once I'm retired I might have the time for that but by the time that comes around my avx would have long gone.. Thought this was a hobbie to enjoy,dont sound like youre enjoying much to me...maybe divert your thoughts on guiding..must be far easier surely

[michael8554]

4 hours ago, saiph said:

residual error is reported to be around +/-4", with a <2" RMS.

Isn't that pretty good for an AVX ?

Are you now seeing RA drift in your 30sec exposures, or Dec drift due to PA error?

Michael

[saiph]

That is only, as I said, "reported", by PECprep, as in "this is what I should observe"    So, I presume I was supposed to see that when I take a long exposure picture;  although "long" is a pretty relative term, when it comes to AP.. let's say a minimum of 30s would be decent; although by other standards "decent" could mean 60s, or 120s, or 240s (anyway, not longer than the acceptable deviation at the chosen image scale I'm shooting at).

However, it hasn't been observed yet, in any sub so far. Which leads me to believe that something went miserably wrong; either in my understanding of the whole process, and of each step (most likely..  in the absence of any reliable information), or in the way the collected data is played back by the controller.

[Zakalwe]

Why are you so anti autoguiding? it seems to me to be a lot easier than the struggle that you are having so far?

[saiph]

Now, that I come to think about it, you might actually be right   lol..
It's the payload I have to carry with me every time I go out that keeps me away from guiding. And the fact that I don't have a solution for keeping my laptop powered up for long enough. Maybe with a more portable solution, such as a stand-alone guider, things would look differently..  I can't tell until I try it

[Zakalwe]

5 minutes ago, saiph said:

Now, that I come to think about it, you might actually be right   lol..
It's the payload I have to carry with me every time I go out that keeps me away from guiding. And the fact that I don't have a solution for keeping my laptop powered up for long enough. Maybe with a more portable solution, such as a stand-alone guider, things would look differently..  I can't tell until I try it

A finderguider and camera will weigh less than a kilogram. Get a cheap Windows tablet and a USBToGo adapter and it will run PHD  for hours. Heck, you could run a planetarium and your imaging software from the same tablet.

http://www.laptopsdirect.co.uk/a1-refurbished-acer-iconia-tab-8-intel-atom-z3735g-1gb-32gb-8-windows-10-t-a1-nt.l7gek.002/version.asp

[newbie alert]

when it comes to AP.. let's say a minimum of 30s would be decent; although by other standards "decent" could mean 60s, or 120s, or 240s (anyway, not longer than the acceptable deviation at the chosen image scale I'm shooting at).

depends on the focal length that you're using..

[tooth_dr]

47 minutes ago, Zakalwe said:

A finderguider and camera will weigh less than a kilogram. Get a cheap Windows tablet and a USBToGo adapter and it will run PHD  for hours. Heck, you could run a planetarium and your imaging software from the same tablet.

http://www.laptopsdirect.co.uk/a1-refurbished-acer-iconia-tab-8-intel-atom-z3735g-1gb-32gb-8-windows-10-t-a1-nt.l7gek.002/version.asp

How does that all work?  It looks so compact and bloody cheap!

[Zakalwe]

1 hour ago, tooth_dr said:

How does that all work?  It looks so compact and bloody cheap!

It runs Windows, so all your usual programs will work. A USBToGo adapter will turn the USB port into a host port so you can plug your gear in. Then just install the drivers as required.

 

https://www.amazon.co.uk/dp/B00OVWY7NY

[saiph]

5 hours ago, newbie alert said:

 

depends on the focal length that you're using..

yes, that is exactly what stands behind "image scale" - as in arcsec/pixel (given by focal and sensor size)

[newbie alert]

So what scope are you using on the avx?

[saiph]

it's the one in the signature - C80ED.  And it's also reduced by a 0.8x FF/FR

[Zakalwe]

Do many people use PEC at all? I always thought that it was a bit of a waste of time on mounts with clutches. As so as you open the clutch then the relation between where the scope is pointed and the geartrain is lost.

Anytime that I tried it with either an EQ6 or an EQ8 I got worse results than with guiding.

[newbie alert]

1 hour ago, Zakalwe said:

Anytime that I tried it with either an EQ6 or an EQ8 I got worse results than with guiding.

And the eq8 has encoders to say where it is if you undo the clutches

[Droogie 2001]

I have seen you struggling with this for some time and I have even posted a few responses to your threads. I can understand your reluctance with carrying all the gear around with you as I have had the same concerns / issues. When I used a DSLR I would setup my laptop and guiding kit on my remote site. When I went to CCD I have decided to largely only image from my back garden, this is not the mounts fault but the logistics of even more kit and power issues. These could be resolved but I am happy to image from my garden for a while.

Having had an AVX for a couple of years I have had a real love hate relationship with it.
Positives are that its reasonably cheap, lightweight and can offer decent results if you are willing to spend the time.
Negatives are that it is fussy with being balanced correctly and does not have a big imaging pay load. I have also had issues with the Firmware in that I have had to reset to factory default many times. Recently the main board and handset failed which cost the best part of £350 to fix. I was close to ditching it due to it not being cost effective but then I would have nothing to sell so had to get it fixed. Plus lots of other niggles I won’t go into here…

My other main issue was when guiding with a WO Star71 and a 50mm Orion Guidescope.

When I upgraded to a CCD the weight caused a big flexure issue. So where as guiding with a DSLR I could get 10mins I was only just able to get 6mins with the CCD. I spent nearly a year trying to fix this issue but never managed until I switched to an OAG. I can now take 25min exposures something which I never thought I could do with an AVX! Everyone says you don’t need a OAG for such a small refractor but on an AVX and if you want to do Narrowband then in my case it was essential.

Anyway, I am getting ahead and slightly off topic.
As others have said you have spent a lot of time in trying to avoid guiding you will be better off just looking at going down that path.
If you want to get long exposures without guiding then you would need to look at buying a high end mount which probably start at £5K+ (a 10 Micron is £7K and I believe this does not require any guiding!)

Good luck and let us know how you get one.

[Zakalwe]

5 minutes ago, Droogie 2001 said:

If you want to get long exposures without guiding then you would need to look at buying a high end mount which probably start at £5K+ (a 10 Micron is £7K and I believe this does not require any guiding!)

^^This^^

Even if you get PEC working, it is not a cure all. PEC will reduce periodic errors but not eliminate them. Periodic error isn't the only source of trailing in images though- bad polar alignment, gear backlash, flexure, focuser sag, cables dragging, improper balance and a myriad of other things can also affect tracking and PEC will do nothing to correct for these. Autoguiding will go some way to eliminating or reducing many of these things, but itself is not a magic "cure-all". However, it is probably the single biggest thing taht you can employ to both increase the quality of your data and the exposure length.

If you are adamant that you want to be autoguiding free then you need to invest heavily in a mount, ideally one with expensive encoders (Renishaw or the like). A Celestron Advance VX is not the mount that you are looking for. I use a Mesu 200.  I wouldn't dream of using it without guiding.

[Zakalwe]

The other thing that I would add is that aiming for 30 second exposures when using a DSLR will not get you very good results. 30 seconds is very short, unless you are using a sensitive CMOS astro-camera. You really need longer exposures than that with a DSLR.

I feel that the best solution to getting longer AND better subs is not PE correction- there's just too many other variables involved (as my post above indicates). There's a reason why the vast majority of imagers use autoguiding - it's fairly easy and, most importantly, it works. There's also a reason why there are relatively few imagers using unguided mounts- they are very expensive and there's not that many mounts capable of taking multi-minute exposures without trailing.

I think that you are painting yourself into a corner here, chasing a solution that, in my eyes, isn't a solution that's employed by many people and one that (again IMHO) will not deliver you the results that you are looking for. Even if you get PEC working I think that you will get immediately frustrated as the results of all that hard work will not match your expectations.

When I was starting out I delayed using autoguiding as I though that it would be expensive and complicated. Once I actually got the correct kit and jumped in I was kicking myself for not doing it from day 1. The quality of the data took a quantum jump, which meant that I discarded fewer sub-exposures and the post processing became a lot easier as I had better star shapes and the signal-to-noise ratio was better with higher-quality data. Moving to autoguiding as early as possible is one of the recommendations that I make to anyone that asks me. The second is moving to a dedicated astro-camera, though that takes a much higher money investment.

[Droogie 2001]

1 minute ago, Zakalwe said:

When I was starting out I delayed using autoguiding as I though that it would be expensive and complicated. Once I actually got the correct kit and jumped in I was kicking myself for not doing it from day 1. The quality of the data took a quantum jump, which meant that I discarded fewer sub-exposures and the post processing became a lot easier as I had better star shapes and the signal-to-noise ratio was better with higher-quality data. Moving to autoguiding as early as possible is one of the recommendations that I make to anyone that asks me. The second is moving to a dedicated astro-camera, though that takes a much higher money investment.

100% agreement from me. I wish I went down the guiding route earlier even though it has its own set of issues for sure!

[saiph]

People, people. come on.. really? 

 

I know it's not a "cure-all" method and I never claimed it to be, of making the mount long exposure friendly - by which, btw, i didn't mean ONLY 30s (I was giving it as an example of what it could be, just as the other exposure times were an example - 60s, 120s, 240s - which, as I mentioned, are much closer to what I'm aiming at. not 30s.. seriously..?)

But, on the other hand, I don't see myself spending big K of money, on super high-end mounts, just to enjoy taking some 2-3min pictures, come on.. you realize too this is plain ridiculous, right? That is why I was after a quick solution for a simple (in my mind) problem.

And, what's with this "clutch loses PEC" thing? In my understanding, all a clutch release could possibly do, mechanically speaking, is shift the larger period curve caused by the RA shaft bearings with respect to the worm period (which is the one we are trying to reduce, anyhow; there aren't enough PEC data cells to cover a full RA shaft turn); since there is no way one could turn the RA worm, or its bearings, or the gear accidentally during a clutch twist.. am I right? Btw, ever since I collected the PE data I never touched the RA clutch, just to make sure no "wonder" happens; but, after seeing that none of my efforts has no effect whatsoever, I gave up on that caution, and played freely with that knob.

The portable solution proposed by Stephen (?) [Zakalwe] seems pretty interesting, provided PHD can work well under Win10 and it will manage to communicate with my cam (which I think I already mentioned it pretty much failed to, so I may have to upgrade that too). So far I'm waiting for the reply of another fellow stargazer who has a Synguider for sale - I just hope he won't "miss" my PM too, as another fellow stargazer did, with another ad, earlier this year... I'm willing to give it a try this way - if it works, it works; if it doesn't, i'll explore more options, but for sure, it won't be a very pleasant endeavor. I am this close to giving up on this wonderful hobby (the photo part, mainly) due to excessive expenses, for my pocket at least.

 

Anyway, if PEC is such a hostile thing to deal with, y.t.h. did Celestron even implement it in the first place, and how on Earth could those people, bragging about their success, make it through?.. It's puzzling me big time, to say the least

[Droogie 2001]

38 minutes ago, saiph said:

But, on the other hand, I don't see myself spending big K of money, on super high-end mounts, just to enjoy taking some 2-3min pictures, come on.. you realize too this is plain ridiculous, right? That is why I was after a quick solution for a simple (in my mind) problem.

The portable solution proposed by Stephen (?) [Zakalwe] seems pretty interesting, provided PHD can work well under Win10 and it will manage to communicate with my cam (which I think I already mentioned it pretty much failed to, so I may 

 

Anyway, if PEC is such a hostile thing to deal with, y.t.h. did Celestron even implement it in the first place, and how on Earth could those people, bragging about their success, make it through?.. It's puzzling me big time, to say the least

You dont have to spend thousands to get pictures that you are happy with, though of course AP can get expensive. 

Do you have any examples of the images you have taken, what exposure times are you using? How is your post processing going? Post processing images is where your AP really starts to come through. You could spend thousands and still end up with lousy images. I only have to look at the early abominations I created before I started understanding the post processing side of things. This is true whether you capture 2min exposures or 30minutes exposures so definitely give this some time and effort.

BTW PHD v2.6.3 works great with Windows 10 as I use it with this.

I have a feeling Celestron implement PEC more as a marketing point than actually being of much help. It would appear that most people who have an AVX do not bother with PEC, well that's my opinion anyway.

[newbie alert]

Maybe you need to speak to the guys that brag about their pec training success .. I'd guess that there's a good few thousand people out there that are using a guiding system such as phd..and maybe have a handful of people that have pec trained their mounts and even less that have successfully trained them..whats the sort of sub length that's attainable from a fully pec trained avx? Maybe a pretty average guided sub length..

To get 2-3 min subs from a AVX you only need decent PA with a 80ed if you didn't want to use a guiding system...

Interesting post that Droogie said about the OAG... thanks for that..

If I was you I'd forget about the pec training..return it to factory settings and enjoy some imaging..after all that's what it's all about..enjoying this frustrating hobbie...best of luck..