First light for a new mount: GM2000 from 10Micron

A place to post informational content, reviews, and links on observing gear/hardware
User avatar
mark.m
Posts: 86
Joined: Fri Feb 03, 2017 1:13 pm

First light for a new mount: GM2000 from 10Micron

Unread post by mark.m » Fri Feb 24, 2017 9:23 pm

My new mount arrived two weeks ago. This is the GM2000 from 10Micron, an Italian company (see http://www.10micron.eu/en/product/gm200 ... -monolith/). It replaces my old MI 250 mount from Mountain Instruments (see http://www.mi250.com/), a one-man company that has since closed up shop and shut down.
My motivation for the upgrade grew out of the realization that my variable star observing was limited by my mount. Two things matter to me: avoiding wasted time during an overnight session (caused either by images thrown away due to tracking errors or by time spent repeatedly trying to properly frame the desired variable star), and image quality (which affects photometric – brightness measurement – accuracy). The MI 250 is a rock-solid mount, and mine initially had good periodic error (+/- 7 arcsec). But over the years the periodic error worsened. I experimented with autoguiding using the embedded autoguiding CCD chip on my SBIG ST-9E camera, but never got the reliability to the point where I could trust it to work during unattended overnight sessions. Further, that guiding chip was behind my photometry filters, and finding a sufficiently bright guiding star was impossible for some variable star fields, particular through the blue filter. And so I've ended up sorta committed to unguided exposures.
What made the GM2000 so attractive is that it uses absolute encoders on both the declination and RA axes, which virtually eliminates periodic error. The company claims that tracking error is routinely less than 1 arcsecond. What appealed to me as an engineer is that 10Micron doesn't sell any version of the GM2000 without absolute encoders, which has permitted them to optimize the entire control system around the use of the encoders. (The downside to the absolute encoders is cost; encoders that can measure subarcsecond rotation are pricey.)
Installing the GM2000 onto my pier was straightforward, just requiring a few holes and bolts. The most difficult part of the installation was wrestling the 66 lbs of mount up onto the pier. The image below is a picture showing the new mount, telescope, and camera.
DSC_0068.jpg
GM2000 mount, 10" SCT, SBIG ST-9E camera
DSC_0068.jpg (121.61 KiB) Viewed 2114 times
It then took a couple of evenings to finish upgrading my software to handle the computer interface to the GM2000 and to learn how to build a "mount model" in the GM2000 firmware, which is a mathematical representation of the imperfections in the mount and the alignment of the telescope/camera.
The firmware has a very nice polar alignment tool, which now tells me that I'm only 18 arcseconds away from perfect (after just two alignment cycles). Using a software-driven alignment cycle, I was able to get 62 alignment points across the sky. The mount reports a residual pointing error of 30.5 arcsec RMS, which is 20 times better than I ever achieved with the MI-250.
The general "feel" of the mount is wonderful. I'd become tolerant of the quirks of the MI-250 and its Gemini-1 controller. The GM2000's firmware seems solid. When you execute a goto, the mount does it quickly and accurately, the same every time. When things go wrong, you don't need to cycle power to get the mount working normally again; instead, just fixing the problem makes the mount happy again.
The mount performs "two-axis tracking," with both the declination and RA motors involved in the tracking process. The mount's pointing model is translated by the firmware into both a declination tracking rate and a RA tracking rate. Thus, the two-axis tracking is able to compensate for all of the known elements of misalignment. In my case, the biggest misalignment source is non-orthogonality of the optical boresight to the RA axis (19 arcminutes), which I think is due to the way my Losmandy dovetail bar is bolted to the telescope optical tube.
I've run the mount on three overnight sessions. The first two were picture-perfect. Pointing error was pretty much always within 1 arcminute, no matter where I was looking in the sky. For my exposures (up to about 3 minutes), there is no visible tracking error. Typical star images have FWHM widths of about 2.6 pixels (compared to around 4 pixels for MI-250 mount images under good conditions). On a typical night with the MI-250, I would throw away 15% of my images due to trailing. Over the first two nights with the GM2000, I only threw away 3 images due to trailing out of 1,514 total images recorded (and I'm still trying to understand what's going on in those 3 images – it seems repeatable and might be caused by the mount reporting that a slew is finished before tracking has become stable at the new spot).
To my delight, photometric accuracy is markedly better. On those two nights, my average photometric measurement error was around 0.038 magnitudes – during the prior 6 months with the MI-250 (and everything else the same), photometric error was right around 0.075 magnitude. The improved performance is caused by the tighter star images, which reduces measurement error caused by nearby stars smearing into the measurement zone of the desired star.
Here's an example of a typical variable star field, using the GM2000:
ax-and_rgb-a-small.jpg
Variable star AX And
ax-and_rgb-a-small.jpg (103.94 KiB) Viewed 2114 times
The deep red star in the center of the image is the variable star AX Andromeda. It has a brightness of magnitude 15.540 +/- 0.05 in this image (green, or visual, wavelength). (And I note that the stars in this image aren't particularly round, but most of that is due to my own incompetence trying to merge RBG images into a composite.)
So that was the first two nights. Murphy joined me for the third night. I set things up relatively late, just before I went to bed. I woke up to find that the software had crashed within a half hour of when I went to bed. Without any further input from the computer, the mount merrily continued tracking the star it had been pointing at when I went to bed, except that this star was just barely circumpolar, so it never went (mathematically) below the horizon and didn't trip the mount's internal tracking limits. Instead, with the star just above the local horizon, the scope's dew cap eventually caught on the edge of the observatory wall upper sill. When I walked outside at 6am, the scope was wedged against the wall, the mount's motors had stopped moving, the mount's hand controller was displaying an "internal hardware error" message, and a dense fog had rolled in, so everything was dripping wet. The mount wouldn't respond to hand controller motion commands, so I loosened the RA clutches, grabbed the counterweight shaft, and moved the scope away from the observatory wall.
Within two seconds, the mount gave a happy beep, the hardware error message went away, and a message appeared to tell me that the RA clutch was loose. I tightened the clutch back up, issued a Park command, the the mount happily went perfectly to its preprogrammed park position. That recovery is something that's only possible with a mount that has absolute encoders, so that it can keep track of where it's pointed, even when moved by hand.
(Once everything was shut down, I was able to sit down at the computer and fix the crash that started the whole thing – the first crash I've had in over a year. This one occurred in the focus manager software, which keeps the telescope focused over the course of a night by making occasional focus cycles on bright – non-variable – stars (perhaps once an hour), calculating focus drift over time, and then tweaking the focus every minute or so based on that drift. It has error handling to handle failed focus cycles (like what happens with intermittent clouds), but that night was the first time that the first focus cycle had failed. At the start of the second focus cycle, some parts of the focus manager thought they were in the first cycle (which they were), while other parts of the focus manager thought they were in the second cycle (which they were), and went looking for information from the previous focus measurement. That sloppiness caused the crash, which was fixed with a single line of code.)
So far, I'm quite pleased with the mount. It's precision is great. I can point the telescope, plate-solve to get the actual pointing direction, issue the mount a 30-arcsecond "nudge" command, do a new plate-solve, and find that the mount has actually moved 30 arcseconds. I don't know yet how long I can make my unguided exposures before drift will become an issue. It's a complete non-issue at 3 minutes. I haven't tried anything longer yet. Next week, maybe.

- Mark M
Mark M
Portsmouth, RI
10" Meade SCT
SBIG ST-9E
GM2000 (10Micron)
American Association of Variable Star Observers (AAVSO) observer code: MMU
User avatar
AndyG
Network Admin
Posts: 1160
Joined: Wed Nov 30, 2011 10:18 pm

Re: First light for a new mount

Unread post by AndyG » Fri Feb 24, 2017 9:51 pm

Mark,

Thanks for the excellent write-up. Congratulations on the mount -- it sounds like an amazing work of technology ... both the mechanical aspects with the encoders and the software modeling and control. It must be very gratifying to see the immediate improvements in tracking, pointing, and efficiency (fewer discarded subs). I talked to the Micron guys at NEAF a couple years ago and they seemed very proud of their product, and I can understand why.
Andy
Chuck
Posts: 254
Joined: Tue Mar 31, 2015 1:14 pm

Re: First light for a new mount

Unread post by Chuck » Sat Feb 25, 2017 11:27 am

Congratulations Mark and thanks for the write-up. It's an amazing mount and should serve you well. Thanks for posting.

---- Chuck
Chuck M.

Meade 8" LX200 Classic OTA
iOptron CEM40EC and MiniPier on Meade classic field tripod
Canon EF 200mm f/2.8L II
Canon 7D Mark II
ZWO 60mm f/4.6 GS with ASI120MM-S camera
Starizona SCT Corrector II and Virtual View 2" Rotating Visual Back
Pegasus Astro Pocket Powerbox
User avatar
Apollo XX
MSSF Coordinator
Posts: 2448
Joined: Sun Jul 11, 2010 8:11 pm

Re: First light for a new mount

Unread post by Apollo XX » Sat Feb 25, 2017 2:37 pm

That's a beautiful setup you've got there Mark! Thanks for sharing it.
Mike M.
Post Reply