PMX intial impressions

[gregbradley]

I am at my dark site observatory and have spent the first 2 nights getting the new mount working and everything communicating.

I have some initial impressions.

Firstly it is a very beautiful mount. The anodising, the machining, the stainless steel and even the little LEDs are all done to look fabulous.

It is neat and tidy and beautifully made.

Installing. The pier wedge mounting plate is also super well made which is surprising as it is $99. It is very well made and made adapting the mounting to my existing piers fairly painless. I got a 10mm drill and tap plus 10mm stainless steel allen key head bolts from Coventry fasteners and it was easy to drill and tap some holes so the mounting plate was in the right orientation. It took 10 minutes or so and worked well.

The supplied CD was a bit confusing. They supply serial numbers for all the varous software programs. The manual said it is on a sticker on the outside of the shipping box. It wasn't there. It was in an envelope hidden under the foam packing the mount came in. I nearly threw it out. Obviously they changed that for security reasons and don't let you know.

Also the manual says the driver for the electronics is on the CD and it is not. That causes confusion. You have to download it off the SB site.

So a bit of confusion there that no doubt they will improve as time goes on and not too bad considering its a new product and they have been in a rush to get lots of product out.

Another point of confusion is the mount will not work (at least as far as I can tell) using the Sky 6, only the SkyX.

If you use CCDsoft then you must configure CCDSoft to run on the Sky X or it will not communicate with it. You won't be able to autoguide.
That detail is also hidden in an SB support forum. Hmmm, a lot of users will have been using CCDsoft and some may have used a PME where it all works together seamlessly. Oh well another confusion from a new product.

So SB need to smooth out the instructions for using the mount as it is lacking at least 3 vital datums that will stop you dead.

However the manual it comes with is one of the better manuals I have seen.

Although it seems to skimp over T-point and says to use the T-point users guide.

I did the quick polar alignment routine in the manual and that is a cool feature. I don't think it would be accurate enough for good imaging but perhaps you could get away with it if your autoguding is setup well.

Still its a good feature.

The USB and power outlet on the part where the OTA fits on the mount (SB call it a versaplate) is fantastic and I only have one USB cable going to my computer. Nice and clean and less tangle. A great feature.

Getting used to the SkyX is not that hard and they have improved the ability to customise the display without resorting to hard to find menus as in the Sky 6. It is much easier and more user friendly than it was.

Go-tos are super fast and the mount is very quiet. It doesn't do the electronic chatter of the fax machine sounding PME. It has pleasant sounding chimes when it starts up and a pleasant LED light that changes colour to let you know its all working.

I did the quick polar alignment then drift aligned my mount. I then did some imaging last night under very adverse conditions (60kmh + winds).

Once the wind settled down it appeared tracking was quite good and about the same as my Tak NJP mount. I have not done a T-point model yet (it keeps failing to plate solve) nor a PE curve so I expect to improve on that. I was getting round stars at 5 minute exposures. I didn't do many 10 minute exposures because the wind was extreme. Perhaps tonight.


The 3 way locking system for the gearing is handy. It becomes very easy to balance. Not that its hard with the PME where you simply unwind a knob to disengage the gear.

But a few precautions. The manual says if you OTA is heavily out of balance then use the lock position for the 3 way switch. I removed my camera and reducer to change an adapter. The scope suddenly flipped due to the imbalance and the gears slipped on the tracking setting.

So a lesson learned there and a near disaster narrowly avoided.
The gears will slip unlike other mounts if there is too much imbalance.
My Proline, filter wheel and reducer probably weigh about 7kgs. So that is the point it could do a sudden flip.

Overall the mount seems to be a winner. Its too early to comment on tracking accuracy except that my tracking errors seemed to be similar to the NJP (perhaps not quite as good but then my NJP was probably more accurately polar aligned). I'd estimate roughly, without looking at a graph its probably about 4-6 arc sec PE. I did see the odd spike. So yet to be measured.

The overall impression is very positive and I am happy so far with this mount. You will need some patience as you get to know how to use it and the little confusions as noted above but then it is a very complex and hi-tech piece of equipment.

Also having all the software bits and pieces is a real plus once you get to know them and can use them to get everything perfect.

I'll add to this as I use it more.

Greg.

[frolinmod]

Thanks for the update Greg. Good stuff. You have a small TheSky6 to TheSkyX transition learning curve to overcome. You'll get over that hump shortly.


Well, I for one would not expect SB to add support for new mounts to old obsolete versions of their software from years gone by. The Paramount MX comes with TheSkyX. Unless you've been running TheSky6 as a prior Software Bisque user from years back (as a few of us here have), you wouldn't have TheSky6 and wouldn't even know that it ever existed. I've been running TheSkyX for the last 14 months. I now don't miss TheSky6 for anything except the missing horizon draw editor (which alas I noticed right off the bat). If you notice any other missing important TheSky6 features, please go over to the SB support forums and request them.


Correct. It's a checkbox in CCDsoft->File->Preferences. However, a bigger issue is that CCDsoft doesn't come with the mount! In my opinion, it should. Especially when you read the manual and it starts talking about using CCDsoft to collect PE information for PEC programming. As a Paramount MX owner, if you don't already have a CCDsoft serial number, you can request a CCDsoft serial number and they will give it to you - free of charge for the asking.


Yup. I think it is hands down the best manual they have produced yet.


A complete PDF copy of which you can find in TheSkyX->Help. Use the supermodel feature any time you have more than six points mapped. Don't try to add terms on your own. Supermodel does this for you and does a better job than you can do manually. Don't even bother with the suggest button. It's no longer useful.


Works with PME mounts too. Usually gets me within six or so knob tics before even running the first Tpoint run to fine tune polar alignment. Not bad.


[Emphasis mine.]That is a particularly nasty and uncalled for thing to say. It's not just like they can drop in the old CCDsoft code and boom, make it so, or something. This is the fastest changing and improving area of TheSkyX. They have limited resources and are in fact working as hard as they can on integrating all of CCDsoft's features into the Camera Add On. Features which you will obtain 100% for free (because The Camera Add On has no subscription fee) as they are added over the next months and years. However long it takes, you will get them all eventually. In the mean time you can set the Camera Add On to use "CCDsoft's Camera." Or at least you can if you already have CCDsoft. I think they should give it to you with the mount. See above about requesting a CCDsoft serial number if you don't already have one.


Ouch. This is a rather sketchy third party anecdotal report. Please have the actual user's of the actual mounts come here and describe the problems and how they were addressed. That would be fair and I hope much more informative.

[gregbradley]

They were private communications to me. One had bad spikes in the PE that would knock the autoguider off track. He couldn't autoguide as a result. He was told PEC would correct it. It didn't. They eventually gave him a new worm and I didn't hear anything after that so I expect it was fixed. Another is on the PMX support forum currently regarding PE above the spec level and he was told to send the mount back.

Looking over images from last night I see I was achieving quite round stars very often when the conditions improved. Those that weren't were wind affected.

By the way I found T-point would fail doing plate solving again last night. I did what Terry said (at least I think I did, the only adjustment seems to be the top adjuster in source extraction setup/detection threshold) of increasing the number of stars it recognises.

Greg.

[frolinmod]

I started off by copying the normal image link extraction parameters (default.sxt) to those used by the automated calibration (AutoPointingRun.sxt). By default they're different for some reason. I think they're set to work well with whatever camera, binning and exposure the Bisques use most, or something. I adjusted from there. I had to make the same sort of minor adjustment. TheSkyX's image link is actually more robust than the one in TheSky6.

[cfranks]

Initially, I too had problems with Image Link not successfully doing Plate Solves. I was certain that it was my great inexperience (with TheSkyX etc) and so it proved. The two main parameters that seemed to be the culprits are the Image scale and the Detection Threshold (Image Link setup). Once I set those properly, the only time I get the dreaded 655 error is when the scope is trying to image through some trees.
My main problem now is I can't get CCDSoft to AutoGuide.

Charles

[gregbradley]

Not sure what you are copying from there.

There's the source extraction setup;
defaults are:

detection threshold 10
minimum number of pixels above threshold 10
number of deplending thresholds 32
analysis threshold 1.5
minimum contrast for deplending 0.005
mask type 0
aperture diameters 5 in pixels
photometry Kron factor 2.5
saturation level 65535
magnitude zero point 10
background mesh size 32
background filter size 1
detector gain 0
size of pixel (arc secs) 2.35 (in my case at 1x1 although I image for plate solving at 2x2, previously in T-point you had to enter 1x1 binning scale)

seeing 1.2 FWHM (as if - I'll change this to 3.2, I don't think I've ever seen 1.2)
cleaning efficiency 1.0
clean (1=yes) 0



A lot of these terms are meaningless to me. But the ones I understand are the first 2 (they are explained) so it looks like I may need to lower both of them.

The arc secs/pixel should be 2.35 which is the 1x1 binned image scale of this setup.

Seeing should be changed to 3.2 which would be more accurate for this area.

Do I need to change the other settings?

Greg.

[frolinmod]

By default only the first two differ between default.sxt and AutoPointingRun.sxt. Back when I first upgraded to TheSkyX (14 months ago now) I found the ones in default.sxt represented a better starting point, so I set the ones in AutoPointingRun.sxt the same and then made minor adjustments from there. That immediately made a huge difference in success rates. Tweaking then improved things to 100% success.

I do not recommend doing automated calibration runs at 1x1 binning. There's no point in doing 1x1 for automated calibration runs. I bin 4x4 with 5-10 second exposures. Binning will NOT affect your pointing accuracy. I always let the automated calibration first do an image link to compute and set my image scale and position angle before proceeding on to the run.

Also note that any time I need to do a sync, I take a picture, do an image link, then sync on linked image. I do not do them manually.

[gregbradley]

Do you use windows explorer to locate these files and copy the default one and paste it over the other one? I can locate them but windows won't open them.

I set binning to 2x2 but now you say that I used 3x3 before. I'll try 4x4.


The take image link photo button on automated callibration run was the one I clicked and it would fail and give an error message.

I reduced the detection threshold, I had already set image scale as calculated in CCDcalc.

Greg.

[frolinmod]

Windows explorer works fine. It'll open them if you tell it to use Wordpad to open them. They're just text files.


4x4 if you can. Not all cameras can. 3x3 works fine if you can't do 4x4.


I would not rely on calculations. Take a picture and open it up in Tools->Image link. Then have Tools->Image link do the plate solve. Let it determine the image scale as well, don't attempt to give it a known image scale. If it can't plate solve the image, then the automated side won't be able to either. Tweak until you can get Tools->Image link to successfully plate solve it. Then use whatever source extraction parameters worked there over in the automated side. You should only have to do this experimentation once with maybe very slight minor tweaks if any after that. I haven't touched my source extraction parameters in the last 14 months after first setting them.

[gregbradley]

Thanks very much for your help there. Much appreciated.

I think I am on top of it now. I studied the manual and I did a plate solve on an image I took last night and it worked. I now have the position angle etc.

Because it is a 16803 chip I wonder if the FOV is too large. The manual says no larger than 10 degrees. I am not sure how wide the 16803 is at F5.25 and 140mm aperture but perhaps it is close.

I suspect it will all work fine now.

There are a couple of cool features in this new T-point add on. I am not sure of the earlier T-point will do this but you can do a quick recallibration of an existing model when you setup newly using the old large model. That is great as imagine if you had a 300 point model and you had to do it newly each time you setup. You'd never do it.

Also the optimization of the model automatically instead of the earlier tutorial of trying out different terms to see which ones worked and which didn't.

Its really very sophisticated.

Cheers,

Greg.

[gregbradley]

Another question.

With Precision PEC you had to enter the image scale for the camera at 1x1 binning and if you used 2x2 etc it automatically calculated the new image scale.

I did the PEC run with the Sky X and it seems you need to enter the image scale at the binning you just shot the PEC data at.

Its not mentioned in the manual. It says to do an image link to get the image scale but I am still finding it failing. Perhaps it doesn't like the large chip.

Greg.

[cventer]

Strange Greg. Maybe is the chip too big. I used QHY10 which is reasonably large chip and image link works every time.

I did mess with some parameters in the solve settings. I will check tomorrow what mine are set to.

I find if you have a model or at least a sync first it works best as mount knows its close to object its solving.

You have to manualy enter the image scale into TCS within skyX for PEC

I still have not got my PEC to actualy work. It either makes it worse or does nothing. keen to see your CCDsoft log files from your mount.

I have tried so many combinations now with west of mount box ticked or not ticked. Just seems like PEC sends wrong commands to mount our they are out of phase some how.

Im hoping its a southern hemisphere bug. I have all my logs but no time to send to bisque borthers yet.

Mind you mounts tracks very smoothly without it so it seems to guide out reasonable easily.

[gregbradley]

I might just swap over the camera to the ML8300. I wanted to see how that imaged anyway with the AP.

Its cloudy now.

Depending if I entered the correct image scale (well its correct for 4x4 binning) it appears PE is around 3.5 arc secs. I tried to attach the file but error message came up invalid.

I can email it to you.

The east/west button leaves the curve as it is or inverts it depending on which one you use. I had it back to front on my PME for a few months and found also the PEC only worsened or did nothing. Getting east/west right made it improve tracking noticeably and gave round stars more reliably.

If tomorrow night is clear I can tell you if PEC improved guiding or not for my mount.

I got quite round stars last night without a T-point model and using a CCD drift alignment and no PEC. I did a short t-point model tonight and it T-point has PA adjustments. So tomorrow night I can do that, do another t-point model and get it closer plus use the PEC.

The resulting PEC curve looks similar to my PME one except a bit wider in amplitude.

It did seem to me that the curve inverted after it was fitted and loaded to the mount. I wonder if something odd is happening there? Perhaps that is usual as the inverse of the recorded error curve is the correction needed.

Greg,

[gregbradley]

Chris,

As I recall Pempro has a button to invert the curve. I just looked at Precision PEC and it does not have that. But it does have this east/west button plus an auto button to find out what it is on your mount.

If no go with the SkyX PEC perhaps try a Pempro curve and see if you get the same results. It has that added function that could make all the difference.

I should be able to tell tomorrow if the PEC commands are being sent to the mount and correcting the PE or not for us southerners.

Greg.

[frolinmod]

Be sure to run BOTH TheSkyX and CCDsoft "Run As Administrator" and that CCDsoft->File->Preferences is set for TheSkyX integration otherwise the worm phase data will always be recorded as zero. You can check your tracking log file from CCDsoft to see if they're zero or not.

The East/West button inverts the curve. If PEC makes things worse, try it the other way. Hey, you have a 50/50 chance of getting it right.

I know binning is okay with Tpoint mapping, won't affect the results, but I'm not sure about the effect of binning in PEC training. I'd be leery to bin there. When you record the data, you use a smallish 64x64 or so tracking box anyway, so you should not really need to bin there, I'd think.

Note that unless Ray has fixed PEMPro very recently, it can neither compute a correct curve for nor can it program a Paramount MX mount. You have to use CCDsoft and the built-in PEC curve computation tab in TheSkyX->Telescope->Tools->Bisque TCS->Periodic Error Correction->Compute PEC Curve.

[cventer]

Did all of that.

No binning, Both run as Admin. CCD Tracking log is showing the worm phase no problem. When I upload to mount using Sky X TCS the mount tracking stays the same or gets worse. (Yes I do clear it each time as well). I have tried it with Data on east box ticked and not, No difference. Actualy with it ticked PE stays about the same. As in same magnitude, but PE cruve looks different, so PEC is making corrections just not sure its making the right ones. Without it ticked it gets worse.

[gregbradley]

It really sounds like something else doesn't it. Perhaps a firmware adjustment and as you mentioned Chris quite possibly a southern hemisphere issue.

I'll be able to run my PEC shortly and I'll let you know the results.
Frolinmod, is your PEC working OK?

Greg.