PEC training confusion

[Anthonyexmouth]

So I have my pier mounted HEQ5 controlled with EQMOD, I guide using PHD2. Whats the deal with PEC training, do I need it, How do I do it and does it help?

[Ronclarke]

Not sure, I've had an EQ3 Pro, EQ5 Pro, HEQ5 pro and now NEQ6 pro and I've never done any PEC training but still get good images? Maybe I've just been lucky??

Ron

[Tomatobro]

If you set Ascom to record six cycles while PHD is doing its stuff when the sixth cycle ends  PEC will kick in and you do see a reduction in the error graph but even though I always park the scope when finished turning on PEC correction at the next session using the file created results in PHD complaining of unexpected movements.

As each graph overwrites itself it gives you something to watch while waiting for the next image.

I clearly do not yet have a good handle on PHD and PEC but perhaps someone can come in and advise as when it does work it does help.

Edited February 19, 2019 by Tomatobro
pelling

[Anthonyexmouth]

1 minute ago, Tomatobro said:

If you set Ascom to record six cycles while PHD is doing its stuff when the sixth cycle ends  PEC will kick in and you do see a reduction in the error graph but even though I always park the scope when finished turning on PEC correction at the next session using the file created results in PHD complaining of unexpected movements.

As each graph overwrites itself it gives you something to watch while waiting for the next image.

I clearly do not yet have a good handle on PHD and PEC but perhaps someone can come it and advise as when it does work it does help.

Yeah, its confused me quite a bit. Watched plenty of youtube vids but none seem to explain it from the ground up or just discuss it on the handset. 

[Star101]

.

 

Edited February 19, 2019 by Star101

[Tomatobro]

 I will read the write up on PHD again to be sure but it does say you can run both together. My experience is that once you have created a PEC file the tracking improves ( the numbers reduce for Dec , RA and RMS).  I only get the PHD error when trying to use the recorded PEC file at the next imaging session which points to  a syncing issue.

[vlaiv]

Here is how I do it:

Disable any PEC if you already have it loaded in EQMod (there are controls for PEC in one of the tabs).

Fire up PHD2 and do calibration near equator. Disable guide output for PHD2 and start "guiding" session (mount won't be guided at this point because you disabled guide command output). Make sure you have logging enabled in PHD2. Do about 1 hour and 30 minutes of recording. HEQ5 has worm period of 638s and you want as much full periods as you can. If you want for example 8 full periods, that is going to be 1:25 or something like that. Make sure you record for at least 1 minute past complete cycle just to make sure you got it. While you are "guiding" at some point - press PEC timestamp button in EQMod.

Once you are finished, just stop guiding and close PHD2. Next start PECPrep on same computer, load phd2 log and analyze PE curve. Choose only harmonics of Worm period in main window - generate PEC curve and save it.

Load it into EQMod and make sure you have PEC gain set to 1.

From that point on, it should be always loaded when you start EQMod - and default tracking rate should be Sidereal+PEC.

That is it.

[Anthonyexmouth]

the reason im interested now is because i have the permanent pier and was wondering if i should be doing something like the PEC training and if its more useful for me now.

[Anthonyexmouth]

2 minutes ago, vlaiv said:

Here is how I do it:

Disable any PEC if you already have it loaded in EQMod (there are controls for PEC in one of the tabs).

Fire up PHD2 and do calibration near equator. Disable guide output for PHD2 and start "guiding" session (mount won't be guided at this point because you disabled guide command output). Make sure you have logging enabled in PHD2. Do about 1 hour and 30 minutes of recording. HEQ5 has worm period of 638s and you want as much full periods as you can. If you want for example 8 full periods, that is going to be 1:25 or something like that. Make sure you record for at least 1 minute past complete cycle just to make sure you got it. While you are "guiding" at some point - press PEC timestamp button in EQMod.

Once you are finished, just stop guiding and close PHD2. Next start PECPrep on same computer, load phd2 log and analyze PE curve. Choose only harmonics of Worm period in main window - generate PEC curve and save it.

Load it into EQMod and make sure you have PEC gain set to 1.

From that point on, it should be always loaded when you start EQMod - and default tracking rate should be Sidereal+PEC.

That is it.

thanks,

1. where do i disable guide output, PHD or EQMOD?

2. are more periods better, would running it longer help or not?

[vlaiv]

I think it's worth doing.

PEC works very well with guiding if you are using EQMod. There are couple of benefits of it. First is that you get a chance to understand how your mount behaves. When you load PHD2 log into PecPrep you have a chance to analyze mount behavior. One of more important things to note is after you generate PEC curve there is "residual" shown - note max RA change rate. It will be expressed in arc seconds per minute or per second. This will tell you maximum guide cycle that you can use. Longer guide cycles are something you want to use as it smooths out the seeing. In order to use longer guide cycles (like 5-6 or more seconds) you need smooth mount and you need your periodic error after correction not to change rapidly. If max RA rate is something like 0.06"/s - and you use 5 second guide cycle - at worst your drift between corrections will be 0.3".

It will enable smoother guiding as well because there will be less corrections issued - and corrections are not ideal so you want as few corrections as possible to keep the mount on track.

[Tomatobro]

Ah...I see what you mean.  The error I am making is that PHD is "guiding" while creating the PEC file which is clearly wrong. Will do your method when the clouds clear....

Thanks vlaiv

[vlaiv]

2 minutes ago, Anthonyexmouth said:

thanks,

1. where do i disable guide output, PHD or EQMOD?

2. are more periods better, would running it longer help or not?

Guide output is option in PHD2 - there is check box (somewhere, I can't remember exactly, but I'll look it up on line now).

Uncheck that particular checkbox and guide output will not be sent from PHD2

It's generally better to capture as much worm period cycles as you can - it's like stacking subs - more subs you stack - better the image. Each captured worm cycle will suffer all sorts of "noise" - wind, seeing, large period error (period of more than half hour or hour - which is not of particular concern, but you want to "cancel" it out with multiple worm cycles).

I usually go for 8 - 10 cycles, I think I never had patience to do more than 10 cycles

[Tomatobro]

When you look at the PHD guide log the periodic error really jumps out of the data. Its unmistakeable.

Just clear one point for me. I do not attempt a visual guide but just let the mount run...right?

Edited February 19, 2019 by Tomatobro
pressed key too soon

[vlaiv]

4 minutes ago, Tomatobro said:

Ah...I see what you mean.  The error I am making is that PHD is "guiding" while creating the PEC file which is clearly wrong. Will do your method when the clouds clear....

Thanks vlaiv

Depends.

I was describing "old" manual way of creating PEC curve. I think that one offers more flexibility and allows you to see PE data and choose how to create PEC curve. You can do all sorts of analysis of mount behavior with PECPrep. For that one you need "clean star motion" data in PHD2 log, so no guide corrections, only star position.

There is new way of doing pec in EQMod - autopec. This is when you just press record in EQMod control panel. For this to work you need your guiding turned on. EQMod can "listen" to guide pulses and based on that reconstruct PEC curve, but in order to do it, actual corrections need to be sent. This process is "automatic" and it does not allow you to either see the data nor to analyze it.

http://eq-mod.sourceforge.net/docs/eqmod_vs-pec.pdf

check page 9 - autopec for how to record with new/simple procedure

[Anthonyexmouth]

Right, next clear night is dedicated to this. 

[vlaiv]

Forgot to mention - best time to record PEC is full Moon

 

[Anthonyexmouth]

On 19/02/2019 at 20:28, vlaiv said:

Guide output is option in PHD2 - there is check box (somewhere, I can't remember exactly, but I'll look it up on line now).

Uncheck that particular checkbox and guide output will not be sent from PHD2

It's generally better to capture as much worm period cycles as you can - it's like stacking subs - more subs you stack - better the image. Each captured worm cycle will suffer all sorts of "noise" - wind, seeing, large period error (period of more than half hour or hour - which is not of particular concern, but you want to "cancel" it out with multiple worm cycles).

I usually go for 8 - 10 cycles, I think I never had patience to do more than 10 cycles

ok, got a clear night so going to give pec training a go, first problem, cant find pec training in my eqmod window

[SamAndrew]

Click on the "screen and green +" icon under the eqmod logo at the top left a couple of times to cycle round to the autopec.

Edited February 21, 2019 by SamAndrew

[Anthonyexmouth]

1 minute ago, SamAndrew said:

Click on the "screen and green +" icon under the eqmod logo at the top left a couple of times to cycle round to the autopec.

Thanks, so do I just click on record there? 

[SamAndrew]

there is a settings spanner on the RHS, you can adjust the number of cycles to record and tell it to automatically play back the pec once done. Its not clear to me how the pec curve gets synced with the worm position if you save it out to a file and then import it back in, so I used the auto playback assuming that would sync it.

[SamAndrew]

As I have an az eq5 I can then start recording ppec on the mount while playing back the pec to load the curve into the mount. It took a few tries as the initial auto pec recording had a few unexpected jumps in it. Haven't seen an improvement in my guiding however, so I'm not convinced it's any use. Need to experiment with it some more.

[Anthonyexmouth]

So it's recording while I'm guiding with phd2? 

[Anthonyexmouth]

So this is 5 cycles. odd looking curve. thought it was meant to be a sin wave like curve

[vlaiv]

That is "perfectly normal" PEC curve. It's not going to be simple sine wave because you have at least couple of components added together.

[Anthonyexmouth]

Just now, vlaiv said:

That is "perfectly normal" PEC curve. It's not going to be simple sine wave because you have at least couple of components added together.

Ah, ok. Just watched a YouTube video and he had a nice looking curve. 

So what do I do with this now?