PEC vs PPEC

[Ewan]

Hi SGL members.

I need some advice regarding the above.

Tonight i am 'using' PPEC on my AZEQ6 or so i think i am.

I setup as usual, calibrated PHD2, started guiding on a star & let it settle for 60 secs then start the PEC record option in EQMod, once its done it's 15 min cycle (i think that's the correct time) EQMod then sais it is doing the corrections, sure enough the red figures are changing.

The PHD graph is considerably flatter, i did not put the PEC captured file through Pempro.

I thought as the mount has a Permanent Pec option in the hand controller that the mount does the calculations / adjustments, is this correct or not ?

If so then am i right in thinking doing PEC on say a EQ5 or 6 you have to put the captured file through Pempro or other, then load the smoothed data file into EQMod & use that ?

Any advice would be most welcome as i hope i am not wasting my time using it, the graph & sub's show i am not but i am still not 100% sure.

[sharkmelley]

I don't know what the gear ratios are in the AZEQ6, but in the NEQ6 the ratio from the motor to the worm is 47:12.  This means that it takes a huge number of worm cycles before the gears are back in their starting position.  This makes any kind of PEC pretty useless if the gears themselves are a large contributor to the periodic error, which they normally are on the NEQ6.

As I say, I'm not certain if this applies to the AZEQ6.

Mark

[chrisshillito]

Hi Ewan,

EQMOD does not program the PPEC function of mount/synscan. What you have used is EQMOD's own PEC which is stored as a file on your PC rather than in the mount. To maintain sycronisation of the file with the mount it is essential tha tthe mount  is parked between sessions. For optimal results you should set EQMOD to record 4 or 5 worm cycles worth of data - I can't recall the worm period od the AZEQ6 off hand but 15 mins seems a little short compared with the EQ6 which requires some 40 minutes to complete 5 cycles.

Pempro use has never been a requirement for EQMOD's PEC and the AutoPEC mechnism now included in EQMOD aims to provide an easier/integrated method of PE capture and PEC curve generation without requiring a suite of external tools. However shloud you need more precise control of PEC generation or more detailed PE analysis there is an alterntive workflow that can use Pempro to capture and build a PEC file that can be loaded into EQMOD - the same is true using PHD/Metaguide/Maxim/AstroArt  logs and our free PECPREP tool.

EQMOD's PEC mechanism is the same across all skywatcher synscan EQ mounts - it doesn't matter wheter you have an HEQ5, NEQ6, AZEQGT or EQ8.

Mark is correct that the the standard gearing ratio of the EQ6 means that our form of PEC will never be 100% effective and there will always be some residual PE present - however in most cases I've seen (and I've seen 100s!) the PE introduced by the spur gears / pulleys is significantly smaller than that of the worm itself and so PEC is still worthwhile (on an EQ6 you would typically expect a 20 arcsec PE to be reduced down to 5 arcsec).

I can't see that you aren't wasting your time using AutoPEC, after all you can have it runing whilst your are actually guiding/imaging so no time is being lost. You can also run AutoPEC whilst guidng with autopec - so you can effective refine a previous PEC definition on the fly. Guiding, by its very nature, can onlly apply a correction after an error has already occured and has been measured. This means by the time the mount is able to respod to a PHD correcting your image will already have been compromised albeit at a sub-pixel level (hopefully!). If you are seeing a flatter PHD graph then I would suggest PEC must be having a noticeable effect and is reducing the error that PHD sees. This can't be a bad thing, but ultimately the proof of its usefulness lies in the quality of the final image subs and not in diagnositc graphs.

Chris.

[Ewan]

I don't know what the gear ratios are in the AZEQ6, but in the NEQ6 the ratio from the motor to the worm is 47:12. This means that it takes a huge number of worm cycles before the gears are back in their starting position. This makes any kind of PEC pretty useless if the gears themselves are a large contributor to the periodic error, which they normally are on the NEQ6.

As I say, I'm not certain if this applies to the AZEQ6.

Mark

Thanks Mark,

To be honest i know nothing about the gearing / ratio etc so i would be a little lost, seem's you were on the money though, apprecaited.

Hi Ewan,

EQMOD does not program the PPEC function of mount/synscan. What you have used is EQMOD's own PEC which is stored as a file on your PC rather than in the mount. To maintain sycronisation of the file with the mount it is essential tha tthe mount is parked between sessions. For optimal results you should set EQMOD to record 4 or 5 worm cycles worth of data - I can't recall the worm period od the AZEQ6 off hand but 15 mins seems a little short compared with the EQ6 which requires some 40 minutes to complete 5 cycles.

Pempro use has never been a requirement for EQMOD's PEC and the AutoPEC mechnism now included in EQMOD aims to provide an easier/integrated method of PE capture and PEC curve generation without requiring a suite of external tools. However shloud you need more precise control of PEC generation or more detailed PE analysis there is an alterntive workflow that can use Pempro to capture and build a PEC file that can be loaded into EQMOD - the same is true using PHD/Metaguide/Maxim/AstroArt logs and our free PECPREP tool.

EQMOD's PEC mechanism is the same across all skywatcher synscan EQ mounts - it doesn't matter wheter you have an HEQ5, NEQ6, AZEQGT or EQ8.

Mark is correct that the the standard gearing ratio of the EQ6 means that our form of PEC will never be 100% effective and there will always be some residual PE present - however in most cases I've seen (and I've seen 100s!) the PE introduced by the spur gears / pulleys is significantly smaller than that of the worm itself and so PEC is still worthwhile (on an EQ6 you would typically expect a 20 arcsec PE to be reduced down to 5 arcsec).

I can't see that you aren't wasting your time using AutoPEC, after all you can have it runing whilst your are actually guiding/imaging so no time is being lost. You can also run AutoPEC whilst guidng with autopec - so you can effective refine a previous PEC definition on the fly. Guiding, by its very nature, can onlly apply a correction after an error has already occured and has been measured. This means by the time the mount is able to respod to a PHD correcting your image will already have been compromised albeit at a sub-pixel level (hopefully!). If you are seeing a flatter PHD graph then I would suggest PEC must be having a noticeable effect and is reducing the error that PHD sees. This can't be a bad thing, but ultimately the proof of its usefulness lies in the quality of the final image subs and not in diagnositc graphs.

Chris.

Chris i am a total novice where all this comes into play BUT i am learning by using EQMod more & more (weather permitting) which, like for anybody else, is a good thing.

Before using EQMod i only used the HC, from what i understood from the manual you selected the Pec training option in the HC, it recorded for what i remember was approx 15 mins (a default setting i presume) then 'applied' the corrections.

I was a little sceptical as it seemed 'too simple', using EQMod seemed more inmformative & visually it actually shows in real time what is happening, which is why i like it.

I forgot how to take a screenshot using a windows function,sorry or i would have posted a before / after shot.

You could clearly see once things had settled & the PEC corrections were being sent it really does improve things, to the point i was doing 1800 sec subs.

I am nowhere near as experienced with EQMod as others are but i can tell you i wouldn't be without it now.

[chrisshillito]

Before using EQMod i only used the HC, from what i understood from the manual you selected the Pec training option in the HC, it recorded for what i remember was approx 15 mins (a default setting i presume) then 'applied' the corrections.

Yes, handset PEC is 'different' and probably much cruder in its implementation. With EQMOD we like to take a few more samples due the fact that the PE waveform does vary between cycles (for the reason Mark mentioned) and taking an average is the best approach. Also, the fast fourier transform filtering and linear regression algorithms applied to build the PEC curve benefit greatly from having a load of data to play with.

I was a little sceptical as it seemed 'too simple', using EQMod seemed more inmformative & visually it actually shows in real time what is happening, which is why i like it.

Believe me an aweful lot of complexity went into making it so simple!

I forgot how to take a screenshot using a windows function,sorry or i would have posted a before / after shot.

Your keyboard printscreen button will dump the screen to the clipboard - or ALT Printscreen just to dump the active window. Then all you have to do is "paste" from the clipboard to get the image back.

Chris.

[Thalestris24]

I forgot how to take a screenshot using a windows function,sorry or i would have posted a before / after shot.

If you have Windows 7 the snipping tool (in Accessories) is excellent! I keep a link on my desktop

[Ewan]

If you have Windows 7 the snipping tool (in Accessories) is excellent! I keep a link on my desktop

I normally use my mobile but thought i lost it camping, turned up in the packed tent :-)

Anyway having another go & so far utilising Pec does help.

This is the lowest RMS i have had so far with this mount, i used the EQMod polar align function + 3 stars either side of the meridian for the GoTo.

Subs are good at 1200 secs & 1800 but settling for 1200's for tonight.

I will post an O3 sub tomorrow to see how it's faired.

Btw i changed the Pec cycle to 3 times which improved it from the other night.

I am now getting somewhere after long last.

Thanks Chris & others for the advice so far, it's very much appreciated.

[Fred_76]

Hi!

Here is a list of the main periods associated to the AZEQ6 mount :

Belt (71 teeth, T2.5) : 11’48’’ (2^nd harmonic : 05’54’’)

Worm gear and its pulley (48 Teeth, 2.5 mm pitch) : 07’59’’ (2^nd harmonic : 03’59’’)

Stepper motor's pulley (12 teeth, 2.5 mm pitch) : 02’00’’ (2^nd harmonic : 01’00’’)

Teeth period : 00’10’’ (2^nd harmonic : 00’05’’)

Period of return of the same alignment between belt and pulleys : 09h26’27’’

Worm ball bearings (608 ZZ or 608 2RS)

- inner track defect : 01’48’’

- outer track defect : 03’07’’

Stepper motor ball bearings (625 ZZ)

- inner track defect : 00’24’’

- outer track defect : 00’39’’

The other needle bearings periods are too long to be noticed :

Needle bearings (NK 40/20 NT)

- inner track defect : 01h23’54’’

- outer track defect : 01h36’29’’

- needle defect : 01h40’41’’

Needle thrust bearing (AXK 6590)

- tracks defect : 55’15’’

- needle defect : 54’35’’

PEC should give better results if one replaces the 71 teeth belt by a 72 teeth belt. It does not change the reduction factor of the mount.

When using a 71 teeth belt, the belt shall turn 48 times (and the worm gear pulley 71, and the motor pulley 284) so that the three elements are back to the same configuration : it needs about 9h30.

When using a 72 teeth belt, only 2 turns are needed : the worm pulley will turn 3 times and the motor pulley 12. Only 24 minutes... far better than 9h30 !

However, T2.5 Lg:180 mm 72 teeth  h:6 mm belts are a bit harder to source.

Fred