139 replies to this topic

[yzhzhang]

Is there a step by step guide to set up a new 10Micron mount in a remote observatory setting? That assumes the use of an image acquisition tool like SGP with a plate solving solution already in place. 

 

Thanks and CS!

For me it was pretty straight forward, and it works just like other mounts you use in the automation integrator workflow. You connect the mount via ASCOM just like others, and you forget about it. You probably want to prohibit sync in your Automation program. Like others said the model is saved and automatically called, so no worries on that. And a lot of us don't plate solve anymore. Even if you do it's centered on in the 1st try anyway (I set my tolerance to be something around 20arcsec).

 

On a bit broader scale, if you want to automating the power on/off, you probably don't want to handle that on the PDU side (so that you introduce a sudden power cut). If your automation system supports executing scripts like Voyager, you can achieve that via script version of MGPbox command. That's something I just learnt recently, and seems very useful.

[yzhzhang]

Hey guys, I have a few questions about MountWizzard 4. I finally (!!) got it to work with QSI cameras... and I updated to MW4 1.0.5. 

 

When creating the model points I use "golden spiral", and I generate maximum # of data points, then I got error message as obviously it exceeds allowed 99 point. I can delete the individual ones, but it's rather inconvenient. So question #1: how do you typically generate the data points to maximize the model but within accepted limits?

 

Also I don't find an option to "Sort E-W" so it is now flipping across meridian once every 2-3 points... Is that option available?

 

Finally, can I view the model results (RMS) during the model building, or it only spits out at the end?

 

Thanks!

[mworion]

Hi Yizhou,

 

if you finally made it with your QSI, could you tell, what the problem was and how you solved it. I would ike to add you experience to the documentation.

 

Michel

[yzhzhang]

Hi Michel,

 

So it's still partially, as the QSI filter wheel and QSI camera cannot be connected in the same time. Only thing to get the camera work is you have to use the latest driver (v2020.07.30) from their page. The lagecy driver will get you connected, but it won't read the pixel size etc info, and exposure won't work for me.

 

Please double confirm before putting it "official" in the document

 

Thanks,

Yizhou

[mworion]

THat's wired as I have the Camera and the filter Wheel connected at the same time with the newest driver as they named now camera and filter wheel different....

[yzhzhang]

Uh, I have my camera connected first, then set up the ASCOM FW. I got the error below; however I continue to "connect ASCOM device" anyway, and now it seems working....

 

I saw the same error message in the past. Key is to ignore them and proceed as-is I guess?....

 

Edited by yzhzhang, 24 February 2021 - 01:29 PM.

[Tonk]

That's a generic late binding Remote Procedure Call error - loads of causes unfortunately

1. Check you are actually running the right version of ASCOM - one that is at least compatible with the QSI ASCOM driver

2. Reinstall the relevant QSI ASCOM driver

3. ???

[Tonk]

Model building survey time. As there are a reasonable number of 10Micron mount owners that visit here I thought it would be useful to survey how you build a model and what the unguided results you are getting. I'll list what parameters might be useful then I'll state my results with my two mounts.

A) 10Micron mount model (1000, 2000, etc)

B) Type of scope (refractor, reflector, SCT etc)
C) Mount firmware
D) Software used to assist building model (with associated software) - or if hand built using keypad/astrometric eyepiece
E) Typical number of stars in model
F) Typical polar align error
G) Image scale in arc sec per pixel

and for results

H) Typical pointing RMS
I) Typical length of reliable unguided imaging


For my GM1000HPS I get:

 

B) Type of scope: Refractor (TV85)
C) Firmware: 2.16.12
D) Model building method: Hand built model
E) Star/point count: 20 - 25
F) Typical polar error: 15"
G) Image pixel scale: 1.5"

H) RMS: 10" - 15"
I) Reliable Unguided time: 10 minutes - very rare to discard frames. Not really tried longer due to light pollution limits at home.

For my GM2000HPS I get:
 

B) Type of scope: Refractor (Tak TOA130)
C) Firmware: 2.16.12
D) Model building method: Mount Wizard 3 with SGPro for image acquisition and ANSRV for plate solving and the Per Frejvall 10Micron mount ASCOM driver
E) Star/point count: 85+
F) Typical polar error: 10" - 12"
G) Image pixel scale: 1.3"

 

H) RMS: 5" - 8"
I) Reliable Unguided time: 10 minutes - very rare to discard frames. Not really tried longer as this is sufficient for LRGB frames at my dark site and my narrow band frames typically start at 20 minutes up to 2 hours so I use hybrid guiding (unguided (dual tracking) with very low rate guiding on top)

These are my virtually bomb proof unguided limits (99+%). I also find that my unguided images (10 minutes and under) have slightly sharper detail than if I guided so I preferentially go unguided under 10 minutes.

Note that I have decided to use medium focal length rigid refractors as my workhorses. I guess the real test comes with much longer focal lengths and with mirror scopes so I'm interested in those results

Also I've never really studied at what point unguided imaging starts to break down and become unreliable - for my purposes, and with site/system limits, unguided works for me reliably up to 10 minutes

Edited by Tonk, 26 February 2021 - 06:57 AM.

[SimonIRE]

For my GM2000HPS I get:

 

B) Type of scope: Dual mounted FSQ106 and LZOS 152mm (with reducer)
C) Firmware: 2.16.11
D) Model building method: Model Creator
E) Star/point count: 99
F) Typical polar error: 15-35" in either scope
G) Image pixel scale: 2.1" (FSQ), 0.8" (LZOS)

H) RMS: 2-4" (FSQ), 5-8" (LZOS)
I)   Reliable Unguided time: It really depends on the criteria for discarding images. I am beginning to realise I am throwing out images based on numbers and not what they look like visually and it has been ruining my experience (stupid). I reliably have imaged up to 15mins unguided in my FSQ at 2.1" or my LZOS at 0.8" with no visual defects. 

 

Results for my GM4000HPS II are pending although I can give this:

 

B) Type of scope: Triple mounted FSQ106, TSA120mm and LZOS 152mm (with reducer) - side by side on Baader Planewave plate
C) Firmware: 3.0
D) Model building method: Model Creator
E) Star/point count: 99
F) Typical polar error: I got 16" in my FSQ and 38" in my LZOS - last night
G) Image pixel scale: 2.1" (FSQ), 0.8" (LZOS)

H) RMS: 3.1" (FSQ), 4.6" (LZOS) - last night

 

No consistent unguided imaging results yet

Edited by SimonIRE, 27 February 2021 - 07:41 AM.

[SimonIRE]

Note that I have decided to use medium focal length rigid refractors as my workhorses. I guess the real test comes with much longer focal lengths and with mirror scopes so I'm interested in those results

 

 

I have tried to image with an Edge1100; it consistently breaks my heart although to a degree this is related to my skill rather than the scope. But mechanically its a dog. I always end up defaulting to my refractors. I know that some here are keen SCT users - I just don't have the clear nights or patience. 

Edited by SimonIRE, 27 February 2021 - 07:58 AM.

[SimonIRE]

I) Reliable Unguided time: 10 minutes - very rare to discard frames. Not really tried longer due to light pollution limits at home.
 

 

Tony,

 

On what basis do you discard subs? Is it visual inspection?

 

Cheers

[Tonk]

On what basis do you discard subs? Is it visual inspection?

DSS score and visual inspection. However discards are mainly because of clouds photo bombing the exposure - very rarely on star shape (the "rare" bit of my statement)

Edited by Tonk, 27 February 2021 - 12:02 PM.

[SimonIRE]

DSS score and visual inspection. However discards are mainly because of clouds photo bombing the exposure - very rarely on star shape (the "rare" bit of my statement)

Apologies; newbie alert - DSS score?

Edited by SimonIRE, 27 February 2021 - 12:03 PM.

[Tonk]

Deep Sky Stacker - it grades your images according to a number of measured criteria - roundness of stars, background sky level (correlates to clouds, lens fog, dawn, dusk etc) etc etc ....

[yzhzhang]

Tony,

 

On what basis do you discard subs? Is it visual inspection?

 

Cheers

Pixinsight SubframeSelector + visual check via Blink.

[SimonIRE]

Pixinsight SubframeSelector + visual check via Blink.

 

I used to do this, but I am beginning to find it unreliable - going by the numbers, in my experience, can result in good subs being discarded. 

[lambermo]

A) 10Micron mount model: GM2000HPS Ultra
B) Type of scope: SCT 25cm-LX200. Focal length 2970 mm.
C) Mount firmware: 2.15.1
D) Software used to assist building model: MW4 0.240.1 on Linux.
E) Typical number of stars in model: 84
F) Typical polar align error: 851 arcsec # good survey, I bumped into the mount weeks ago and forgot, I need to adjust :-D
G) Image scale in arc sec per pixel: 0.26"
H) Typical pointing RMS: 6.5
I) Typical length of reliable unguided imaging: Not tested recently. I use 30s subs. Stars get elongated when I guide, stars are nice round when i do not guide

 

Also, I still do not yet use temperature or pressure compensation. I will someday

 

I have had successful unguided subs at 15 minutes years ago at this nearly 3m focal length when I tested for that, but it was not consistent.

 

-- Hans

[mccomiskey]

Follow up to an earlier discussion. 

 

I made a series of changes to my system a couple of months back in order to try to improve the pointing accuracy (RMS) of my models.  I did them all at once, and ultimately brought the RMS down from 7"+ to ~5".  One of those changes was to extend the tripod legs.  For a variety of reasons I reversed that last week, pulling the tripod legs in to being as short as possible, while changing nothing else.  I have since built four models, and the RMSs have all been less than 3" (for reference, in my system the sampling is 3.5").  Seems to be a consistent and substantial improvement.  Sharing for the benefit of the group still trying to optimize the physical aspects of their systems.

[Tonk]

pulling the tripod legs in to being as short as possible, while changing nothing else

This is what I do on my mobile rig when using a Losmandy tripod. I only extend individual legs up to a couple of centimeters for leveling only.

Tripods have a twisting flex mode around a vertical axis through the head of the tripod. The amount of flex is a direct function of the length of the legs and reduces as the leg length is shortened

Edited by Tonk, 30 March 2021 - 02:16 PM.

[durak]

I finally got a decent night to try to create a model with my new 10 micron 1000hps. Things started out poorly as I was unable to select NINA as my camera/platesolver in ModelCreator2. My only option for camera were ASCOM and my options for platesolver were PlateSolve2 and ASTAP.

 

"Ok, whatever. I will just use my camera's ascom driver and astap". Well after several failed solves I completely deleted ASTAP and reinstalled. Finally it worked. At this point, It was about 11:00 and I did not want to deal with trying to PA using the mounts integrated software so I fired up Sharpcap for a quick, single-iteration PA. After my PA, I tried my first model. I used 30 points (later culled down to 27) and it only took about 11 minutes! I fired up NINA and took two 10-minute shots. The attached image is a stack of those two frames. The stars are not perfect but I was pretty pleased with the result. Next time out, I will work on getting a better PA and creating a larger model. Now that I have done it somewhat successfully one time, I hope that it will become much easier and better with each successive attempt. 

 

tak fsq85

2600mm pro

L filter

2x10' unguided

 

[Tonk]

I did not want to deal with trying to PA using the mounts integrated software

Pity - because all you do is slew to a suitable star via the PA align menu function and center the star using alt/az knobs on cross hairs - its that easy. And the 10Micron knows you are doing this so automatically wipes the now invalidated model.

Edited by Tonk, 18 April 2021 - 01:35 PM.

[psandelle]

Pity - because all you do is slew to a suitable star via the PA align menu function and center the star using alt/az knobs on cross hairs - its that easy. And the 10Micron knows you are doing this so automatically wipes the now invalidated model.

Super super easy. And visual. No knob-turns to calculate (or remember which direction to turn in).

 

Paul

[durak]

Pity - because all you do is slew to a suitable star via the PA align menu function and center the star using alt/az knobs on cross hairs - its that easy. And the 10Micron knows you are doing this so automatically wipes the now invalidated model.

I don't really view not using the mount's PA routine as a loss. I had a fairly successful night considering it was my first go.

 

I will definitely give it a shot the next time I'm out though. 

[Tonk]

I don't really view not using the mount's PA routine as a loss.

No prob - but as it tied to the current model its an accurate method - especially if you pick the alignment star as one close to both the celestial equator and the meridian (for max accuracy on *both* Alt and Az axis).

To speed it up you initially build just a small model (more than 12 points but less than 20) to get a close initial PA, then rebuild a small model again and PA one more time to improve it - should be sub 1 arc minute now (I would be around 12" - 15" with my rig at this stage i.e 1/5 to 1/4 arc min error) - this will now be very good and easily good enough for dual tracking.

Now go and build the uber pointing/tracking model with as many points as you can get. 

[durak]

To speed it up you initially build just a small model (more than 12 points but less than 20) to get a close initial PA, then rebuild a small model again and PA one more time to improve it - should be sub 1 arc minute now (I would be around 12" - 15" with my rig at this stage i.e 1/5 to 1/4 arc min error) - this will now be very good and easily good enough for dual tracking.

 

So both PA models should be between 12 and 20 points? I see you explicitly mentioned that number for the first model but didn't for the second model. 

 

For the final model, how many points do you recommend for 600s unguided with a rigid fsq-85? I was thinking between 50 and 60.

 

Currently I set up every night but soon the mount will be housed in a backyard observatory.