PE Confusion

[Shelster1973]

Had this with my AVX and still am a wee bit befuddled now that I have moved up to my EQ6R-Pro.

The3 mount comes with PPEC and records the worm over a cycle so as to determine the periodic error of this which it can then use to compensate for when guiding.  This bit I get, but my confusion comes from many conflicting opinions and advice as to how to go about recording this for playback and then whether it is actually worth doing when guiding.

My feeling for how it should be is that you set-up the mount as per normal (ie fully loaded with kit, balanced correctly and polar aligned), you then set the mount to track and record the worm over a cycle.  This then will record the exact errors that the gearing has under load and tracking. 

The confusion in my head is that why do some say that you should be guiding?  Yes this will be giving you a nice guide on a star, but the point is to record the error of the gear.  By guiding you are not recording the true periodic error, but any error that is skewed by guide pulses which will not be constant due to many factors (alignment, seeing, wind etc).  This means that any errors in the period will be shifted in a slightly positive or negative way surely?

Also, with regards to the just guide with no PPEC running, surely if you have recorded the error, then the software will know when one of the peaks is coming and will not see it a 'OH MY GOD I NEED TO PUT IN BIG GUIDING PULSE AS THERE HAS BEEN AN EARTHQUAKE THAT SHOOK THE MOUNT' (some dramatic licence has been taken in the last comment).

Anyhooo....I think that when my mount makes it back from FLO, I will be going down the record unguided and then playback while guiding.... unless of course someone can clear the fog that surrounds this

[vlaiv]

I don't know how PPEC is implemented in EQ6-R, so what I'm about to write applies to PEC with EQMod. PPEC might be different as it should store PEC information in the mount itself.

There are two ways to record PE and create PEC curve.

First one is "traditional" way. You start your guiding session but disable guide output. This is done so that guiding application can create log of guide star position without altering any "regular" motion of mount. After you finish this PE recording session, you load resulting log in suitable application, analyze guide star motion and from this generate PEC curve.

Second one is new way of doing it. It involves less work, but I suspect is not as precise as traditional way. It's added as convenience for people that don't want to bother with above complex procedure (first recording than manually analyzing and creating PEC curve). It is implemented in EQMod - which is ASCOM driver for Synta mounds.

You start almost the same way with a slight difference - you don't disable guide output. In this instance you press "Record PEC" in EQMod control panel as you are guiding. EQMod is then monitoring your pulse corrections and from these it builds PEC curve in the end.

In some sense both methods are equivalent - meaning that they "suffer" from same issues. Both star position and guide corrections will be affected by seeing, wind, shake, whatever along with actual PE. This is where analysis step comes in - you can analyze collected data and effectively exclude random influences and keep periodic ones. This is same as regular imaging and stacking of multiple exposures - random gets "canceled out" while stuff that is really there gets "reinforced". Same thing happens with mount PE - random influences get canceled out while PE that is present in each period gets reinforced. This is why it is important to collect multiple PE cycles to get good PEC curve - same as with imaging you need multiple exposures stacked and more exposure you have - better the image.

Measuring star positions is equivalent to measuring corrections needed to move mount to that star position - it's just two different ways to measure same thing. One is just absolute and other is relative, but they add up to same thing.

There is also issue of "conflicting" PEC and Guide commands when guiding with PEC curve often mentioned. I'm certain that EQMod does not have this issue (look under VSPEC in EQMod manual for explanation), and I think that most modern mounts should not have this issue as PEC should be implemented in similar way to VSPEC - via variable speed of RA tracking.

Hope this clears things up a bit. Just make sure you understand how PPEC is generated, stored and applied in case of EQ6-R (like I said - I don't know details for that).

[Filroden]

48 minutes ago, Shelster1973 said:

then whether it is actually worth doing when guiding

If you are guiding you shouldn't need PPEC. If the periodic error is so large that guiding does not cope with it then there are bigger issues that need resolving

About the only time PPEC would be useful is trying to get longer subs on unguided mounts which suffer from a known periodic error. But why go to all that trouble when guiding is so easy and takes care of all tracking variances, not just PE?

[vlaiv]

7 minutes ago, Filroden said:

If you are guiding you shouldn't need PPEC. If the periodic error is so large that guiding does not cope with it then there are bigger issues that need resolving

About the only time PPEC would be useful is trying to get longer subs on unguided mounts which suffer from a known periodic error. But why go to all that trouble when guiding is so easy and takes care of all tracking variances, not just PE?

PEC in general is useful when guiding. You certainly don't need PEC when guiding, but why not get better results if you can?

Guiding is supposed to correct for tracking errors. Less error to correct - the better, and PEC is good way to lessen mount tracking error.

With large PE, for example, you need faster guide cycles because of speed of PE change. If you can lower speed of PE change, you can use longer guide cycles and still have enough "time to react". This in turn helps with seeing - as longer guide exposures average seeing effects out - less chance of chasing the seeing.

[Shelster1973]

Guys

Many thanks for the input on this.

Has cleared things up a bit.  Form what I understand now.....the EQMOD method 'looks' as though it is doing a PPEC but in the EQMOD side of things and not the mount, which makes sense.  Being on the computer side and not stored in the mount would allow PHD2 access to the data, so it can anticipate when the periodic errors are due to happen.  If it was just in the mount, then PHD2 would have no knowledge that the mount is about to/has corrected and will react accordingly with large compensation.

Also feel that it is best to run with both if this is the case as will reduce the input from the guiding side when the error is due to happen.

Not sure what the error of my mount is also.  Will check it when I get it back.  Has been returned to FLO as was refusing to track and was showing a power level error on the LED despite trying 3 different PSU's and 2 different cables.

[wimvb]

Just to confuse things. I read just yesterday that PHD has a PPEC guiding algorithm, where the first 'P' does not stand for Permanent, but for Predictive.

Anyhow, weather permitting I will try to record PPEC on my AZEQ6, because my guiding results show an oscillating error with a time period of 2 minutes and an amplitude of 0.6". And no matter what I try, PHD won't guide that out.

[vlaiv]

7 minutes ago, wimvb said:

Just to confuse things. I read just yesterday that PHD has a PPEC guiding algorithm, where the first 'P' does not stand for Permanent, but for Predictive.

Anyhow, weather permitting I will try to record PPEC on my AZEQ6, because my guiding results show an oscillating error with a time period of 2 minutes and an amplitude of 0.6". And no matter what I try, PHD won't guide that out.

PPEC guiding algorithm seems like really good idea in some cases. Regular PEC can only work on harmonics of worm period. These are usually long cycles (order of one minute or more). Some mounts suffer from shorter period errors that are not harmonics of worm period. HEQ5 for example can have issue with 13.8s motor gear tooth period issue. PPEC being predictive in nature can "learn" this short periods. It can do it quickly, because it only takes couple of minutes to get enough "short cycles" to learn them. In case of said 13.8s period it will gather 8 full periods in under two minutes - so one does not need long time to train PPEC guiding algorithm - just let it guide for 2-3 minutes before you start imaging and it should pick up any short term oscillations in mount drive.

These two are going to work well together - regular PEC and PPEC guiding - one will take care of long period error and other will do the same for short term period errors if mount suffers those.

[wimvb]

11 hours ago, wimvb said:

Anyhow, weather permitting I will try to record PPEC on my AZEQ6, because my guiding results show an oscillating error with a time period of 2 minutes and an amplitude of 0.6". And no matter what I try, PHD won't guide that out.

Well, weather permitted. But still no luck with programming ppec. I could start the sequence from the indi version of eqmod. The mount responded with the power led blinking, indicating that it was waiting for the index on the ra worm. But for some reason, never detected it. So I interrupted it and in stead used phd's Predictive PEC guiding algorithm with default values. I have a feeling that it brought the 120 s oscillation down, as guiding rms was a steady 0.88" and the oscillation was less obvious. But I'll know more tomorrow after I've analysed the guide log.

For now I'm inclined to say: forget about the mount's permanent pec, and use phd's predictive pec in stead.