"TheSkyX->Telescope->Tools->Bisque TCS->Periodic Error Correction->Compute PEC Curve" pretty much is Precision PEC.
Yes I imagine that its the guts of Precision PEC. But it may be handy having an external piece of software that does that in case there are still bugs in Sky X PEC.
I got a replacement worm sent for free bySoftware Bisque several weeks ago. I haven't been able to get to the mount due to work.
I did last night and successfully changed the worm and readjusted the cam thanks to the helpful video from Chris Venter.
I did a 10 minute image and got elongated stars, quite badly elongated too. It was windy and that was not helping for sure but the elongations were from tracking/PE issues not the wind as I strategically position the slide off flat roof to minimise wind and I was using my AP140 which is not partcularly wind prone.
I also did several automated t-point runs and that is now working quite reliably (I got on top of that last trip and it remains working - yeah!).
I adjusted the polar alignment until I got a perfect report from t-point. Odd behaviour occassionally with t-point. It can be slightly erratic.
I did a 24 point run and it told me to raise the altitude by 8.8 ticks. I did that. I ran it again and did a 48 point model and it told me to lower it by 8.5 ticks! I did. I ran about another 44 point model and it said it was perfect.
I also did a new PEC curve and the resulting curve did not seem to affect guiding.
I did another PEC using 2x2 binning instead of 3x3 binning. It gave a better curve and one that looked more like the one on my PME that works so well. Its more subtle and only small sections above and below the middle line rather than large.
So I conclude from this that 2x2 binning is way more accurate than 3x3 is on a 1050mm focal length refractor. Also seeing was not great due to the wind. I wonder if I should try one at 1x1 binning to get even higher resolution.
So now I have a new PEC and perfectly polar aligned it was time for the acid test - more exposure runs.
Oops, more elongated stars and very very odd autoguiding where the guide star oscillated between 2 points back and forth and I was getting 2.39 pixel errors in guiding which is massive (I want below about .3).
I immediately thought this must be some sort of backlash issue. Perhaps I did not tension the drive belt correctly but I thought it was done the same as Chris did on his video.
I slide the counterweight further down the shaft to put more weight bias (the RA shaft was nearly horizontal and I wondered if it was too balanced). I also recallibrated the autoguider just in case something was way off there.
I also went for much longer guide exposures and reduced the aggressiveness down a lot (to 3). Bingo - guide errors routinely low - between 0.00 and .4 mostly.
I also slid the dewshield down from full extension to keep it out of the wind stream. That helped a bit.
5 minute subs were now showing perfect guiding. I was using a MMOAG and an SBIG STi guider. Woohoo!
I think the PE was so low that the autoguider was causing its own oscillations and I needed to turn the guider way down. I may need to turn it down even further as the polar alignment is very very spot on.
No sign of a spike in PE (there was a bit of black rubbery thread on the belt, I wonder if that was part of the spike I was getting, the worm did not appear to be damaged in any way although perhaps scratches may only need to be tiny to affect guiding and grease hides some of that).
So it looks like I have a working PMX and most likely a working PEC curve (to be verified tonight).
Next step is a 300 point t-point model with Protrack corrections as well as PEC corrections and 15 minute subs. That's tonight if its clear enough.
Gee, I even did some imaging for once!
Perhaps I could even whack a 12 inch OOAG on this baby now or a CDK 12.5 inch (probably not, too wind prone but OOAG only has half the tube sticking up in the air).
Glad to hear that u seem to have it all under control Can u do a test for me as i've been having trouble with trying to get 5 min unguided round stars with my fsq and try it without guiding at all These mounts are supposed to do that for 2000mm FL's.
From what i've read 1x1 bin for pec is the best, yet to do that myself, Perth's weather's not been the best.
Also do u have a pic of your setup ? Any leads off the end of the scope from the camera etc or did u route them thru the mount ?
Glad to hear that u seem to have it all under control Can u do a test for me as i've been having trouble with trying to get 5 min unguided round stars with my fsq and try it without guiding at all These mounts are supposed to do that for 2000mm FL's.
From what i've read 1x1 bin for pec is the best, yet to do that myself, Perth's weather's not been the best.
Also do u have a pic of your setup ? Any leads off the end of the scope from the camera etc or did u route them thru the mount ?
Thanks Greg
Thanks for the tip. I think I'll do another one at 1x1 and save both log files so I can revert if it worsens things. I think also per the manual a night of good seeing is really required. 1x1 versus 2x2 probably needs excellent seeing to get that advantage but you may as well max it out anyway.
I don't use through the mount cabling. I do use the PMX usb and power port on the dovetail though. Very handy.
Per the manual you need to use ProTrack for long unguided round star images. I was just reading through the ProTrack info as I plan to do that tonight. 50 -200 point model is required first and super model used.
You need perfect polar alignment as well so you'll need to do several t-point automated runs anyway to get to that point. What threw me off last night I think was the fact my polar alignment was more perfect than I am used to and perhaps this new worm is considerably more accurate than the old one. So autoguiding, especially with ProTrack enabled may only need to be the occassional minor correction once every 15 seconds or longer. I have to trial and error to get the best exposure length and aggressiveness setting.
I'll let you know how I go with ProTrack to see if I can get 10 minute pinpoint star images with no autoguiding. I'll give that a go tonight now everything seems to be working.
So your ideal PMX setup is a very stable platform, levelled off, T-point model done and Polar alignment tweaked until you get the report
no further corrections needed. 200 point t-point model with Protrack enabled. A PEC curve done exactly per the instructions and on a night of good seeing, an offaxis style gudiing setup to prevent flexure, a sensitive and clean autoguiding camera. Scope protected from wind.
Quote:
Originally Posted by DavidNg
Congratulation Greg. Whats you have achieved in one night most probably took us many weeks and pulled half of our hair out.
Its been a while really delayed mostly by the fact I haven't been able to use it much for a while due to work.
When it all works its an impressive engineering /software meld which is why I got it in the first place. I guess its well known that early adopters of new technology often have problems. Nikon seems to be having trouble that way recently with quality issues with some of their high end cameras.
Always doing at least 42 sample Tpoint runs and Supermodel helps with the polar alignment consistency. Supermodel uses different algorithms with fewer points. 42 or so is where it appears to change to a better more consistent algorithm. Either that or I'm deluded.
A windy night is never a good night to test anything. Been there, been frustrated.
Great to hear Greg!
With a good model I don't think you will have much trouble going 10 min unguided with the focal length you are using.
I measured and programed my PEC a couple of nights ago, and my uncorrected PE was around 2.8", after the correction it went down to 1.4". The seeing here has not been very good. I think with better seeing I should be able to get sub arc second tracking.
Here is a shot from the other night. This is 15min unguided, focal length of 2541mm, .6 arc seconds pixel, PEC and ProTrack on.
This is not perfect but maybe with a better model it might be possible to get round stars. 15 minutes unguided is pushing things to the limit at this focal length.
Always doing at least 42 sample Tpoint runs and Supermodel helps with the polar alignment consistency. Supermodel uses different algorithms with fewer points. 42 or so is where it appears to change to a better more consistent algorithm. Either that or I'm deluded.
A windy night is never a good night to test anything. Been there, been frustrated.
That's good to know. Thanks for that. Also I had the hand controller plugged in for most of it and then I detached it. I noticed the raw data showed a big drop perhaps after I unplugged the hand controller so that step of unplugging the hand controller may be important. I thought it would only affect results if you accidentally used it.
Windy night is not good - true.
Quote:
Originally Posted by CDKPhil
Great to hear Greg!
With a good model I don't think you will have much trouble going 10 min unguided with the focal length you are using.
I measured and programed my PEC a couple of nights ago, and my uncorrected PE was around 2.8", after the correction it went down to 1.4". The seeing here has not been very good. I think with better seeing I should be able to get sub arc second tracking.
Here is a shot from the other night. This is 15min unguided, focal length of 2541mm, .6 arc seconds pixel, PEC and ProTrack on.
This is not perfect but maybe with a better model it might be possible to get round stars. 15 minutes unguided is pushing things to the limit at this focal length.
Cheers
Phil
That's impressive Phil. It looks like you are almost there.
I noticed on the 2nd night with the new worm if my mount is well balanced I am getting oscillating autoguiding errors. If I slide the counterweight down a bit more it turns off and goes accurate. But if I slide it down too much it can stall the slews. I turned down the slew rate.
Do you think I may not have gotten the correct tension on the belt that drives the worm?
It did not oscillate before when in close balance. I mean its not too big of a deal as I can simply keep some weight on the balance to the west which is a common advice but this is belt drive not gear drive.?
Do you think I may not have gotten the correct tension on the belt that drives the worm?
It is possible that your belt is not tight enough. I read on the SB forum that you could not over tighten the belt. It was Steve or Daniel Bisque that said this.
I would be reluctant to put a lot of pressure on the belt, but when you push in on one side the other side should not move.
I don't know if this would cause the oscillation but it would be worth trying.
Keep in mind this may alter your PEC and TPoint Model.
I have only done some quick tests with the guide function on Sky X, but I have found at my image scale, I had better results unguided.
I did a log and found there was an oscillation also, +- 2 arc seconds. I put this down to the settings. I haven't had a chance to explore this further.
It is possible that your belt is not tight enough. I read on the SB forum that you could not over tighten the belt. It was Steve or Daniel Bisque that said this.
I would be reluctant to put a lot of pressure on the belt, but when you push in on one side the other side should not move.
I don't know if this would cause the oscillation but it would be worth trying.
Keep in mind this may alter your PEC and TPoint Model.
I have only done some quick tests with the guide function on Sky X, but I have found at my image scale, I had better results unguided.
I did a log and found there was an oscillation also, +- 2 arc seconds. I put this down to the settings. I haven't had a chance to explore this further.
Cheers.
Thanks, that is very helpful. So perhaps I can tighten it more. I used the Chris Venter video and Daniel commented it had sage advice about the belt tightening and Chris demonstrated that the belt should not bow when reversed which I checked for but my tension may have been a taf more than he seemed to go for. It was approx the same tension as when I removed the casing on the original worm.
2 arc sec error in guiding is massive.
With a bit of weight on the counterbalance I am getting under .4 and most corrections are under .3.
This morning with even longer guide exposures of 10 secs I was often getting 0.0 and then under .3 so its getting pretty close to being spot on. But the weight effect worsens when the counterweight arm is at a higher angle and may be producing a drag causing slightly worse guiding.
Some subs the round stars are breathtakingly perfect. Others very good and close to perfect but needs a tad more.
I am hoping getting this little issue adjusted and a better PEC on a night of good seeing (last night seeing was spectacular) and a new large model and it could be perfect.
Greg.
Quote:
Originally Posted by Mighty_oz
I'm pretty sure they suggest No hand controller connected when doing t-point, pec, etc
I suggest leaving the hand controller in a box and only connecting it when you absolutely have to for some reason (which should be a rare event). In the history of Paramount mounts, the hand controller is the #1 source of oddball problems. The guider ports are #2. Read through the old paramount forum on yahoogroups.com. I place dust cover plugs in all three of these ports. (I realize the hand controller has gone through multiple revisions and that the current one is by far much better than the old one.)
I suggest leaving the hand controller in a box and only connecting it when you absolutely have to for some reason (which should be a rare event). In the history of Paramount mounts, the hand controller is the #1 source of oddball problems. The guider ports are #2. Read through the old paramount forum on yahoogroups.com. I place dust cover plugs in all three of these ports. (I realize the hand controller has gone through multiple revisions and that the current one is by far much better than the old one.)
I use it all the time. For framing the image and lining up the next nights images with the ones already taken the night before. Its a fantastic little unit with the built in joystick and the adjustable sensitivity. You must use the move keys to frame your images in CCDSoft or Sky X?
I had an odd thing happen only once last night. I updated my Sky X to build 6249. I started CCDSoft to run my cameras as usual (not quite ready to add the step of Sky X running the cameras fully - soon) and
also connected the camera using the camera tab (I think this needs to be done to use automated t-point runs) and it all worked well for a while then I suddenly got an error message saying CCDSoft was no longer working and CCDsoft started flashing on and off the screen and I had to pull the battery to stop it.
I ran a clean up pgm to increase available ram etc and restarted it and all was well. But I did not connect the camera again in Sky X. Hadn't happened before with build 6248.
You shouldn't be able to connect both CCDSoft and the TheSkyX->Camera Tab to the camera at the same time - unless the Camera Tab is set to use "CCDSoft's Camera" which should work. I do (did until recently) that all the time. The only time I've had CCDSoft or TheSkyX "hang" on me is when I power down the camera without disconnecting them from it first.
I control everything from the comfort of my dining room table inside the house. I frame objects by slewing to them, use the telescope tab's jog control for small movements, then rotate the FOV (I have a Pyxis rotator).
Re the hand controller, I'm with Greg. I have never used the MX without the HC connected (I must have missed the directive to remove it) and have yet to experience a T-Point problem attributable to it. I always use it to centre the initial calibration stars or the initial sync star if I haven't disturbed the OTA. I did a 350 point model 3 days ago with only one 'not enough stars' error when a cloud intervened.
Quote:
With a good model I don't think you will have much trouble going 10 min unguided with the focal length you are using.
I'm not very knowledgable about imaging etc but I don't understand how a good model can affect unguided tracking. Polar alignment and PEC, definately, but doesn't the model only allow greater accuracy for the slews? Sorry if it is a noob question but these forums are my only source of education in this area.
OK, I see another thread that suggests Protrack is derived from the model!
I wonder is it worth to update every SB daily builds and take unforeseen risks? or just to wait longer for tested "latest update". On SB forum, it appeared many users have issues with daily builds.
I'm not very knowledgable about imaging etc but I don't understand how a good model can affect unguided tracking.
Turn on ProTrack. With a good model, Tpoint can model and correct for flexure and other stuff in real time thus improving tracking.
Quote:
I wonder is it worth to update every SB daily builds and take unforeseen risks? or just to wait longer for tested "latest update". On SB forum, it appeared many users have issues with daily builds.
The last "latest update" took over a year to be updated. Do you really want to wait over a year for a bug fix? Daily builds come out often. No way would I ever wait around for a latest update. I might sooner drop dead. Also, in my experience the latest updates are not magically less buggy than the daily builds nor are the daily builds any more buggy than the latest updates (and if they are, then they are quickly fixed).
[QUOTE=frolinmod;919056]You shouldn't be able to connect both CCDSoft and the TheSkyX->Camera Tab to the camera at the same time - unless the Camera Tab is set to use "CCDSoft's Camera" which should work. I do (did until recently) that all the time. The only time I've had CCDSoft or TheSkyX "hang" on me is when I power down the camera without disconnecting them from it first.
Yes that is what I have been doing. I think that is necessary to do automated t-point runs if you are controlling your camera in CCDsoft which I have been.
It only happened once but I didn't connect it again in SkyX so it may have been a once off thing. Who knows or I may have been running low on RAM.
I wonder is it worth to update every SB daily builds and take unforeseen risks? or just to wait longer for tested "latest update". On SB forum, it appeared many users have issues with daily builds.
You know I thought of this point before I updated. If a system is working don't do anything is really the golden rule! Who knows what complications a subtle change can have on complex software.
Can u do a test for me as i've been having trouble with trying to get 5 min unguided round stars with my fsq and try it without guiding at all 
