Notes on upgrading observatory computer to Windows 7

From Telescope
Jump to: navigation, search

This page is to document our upgrade to a new observatory computer with Windows 7. The original machine used to control the telescope (from commissioning in 2009 through summer of 2012) ran Windows XP. So there may be some issues encountered in the upgrade. Thorough documentation here should (hopefully) help with next time we need to upgrade hardware or software.

Initial testing was done with the same computer, but with a fresh Windows 7 install on a separate disk; so not everything in the first section may be completely the same with the new computer. (Eric Jensen, July 12-13 2012)

Software installation

(New notes with new computer, August 16- 2012.) To determine best plan for software installation, I contacted Bob Denny at DC3-Dreams (publisher of ACP) for advice on order of installation. He said "Install the ASCOM Platform 6 first. Then install the TheSky Controlled Telescope driver. Then the TCC Software. Then TheSky. Check TheSky out with your scope. Then check it via the ASCOM driver using the POTH tool that comes with the ASCOM Platform. Then MaxIm as ACP will be looking for it during installation. Then ACP and Hot Fix 4. Make sure at the end of the Hot Fix to use the "Register TCC for ASCOM" option, their software fails to do this and is no longer maintained by them. Next install the Digital Domeworks ASCOM Driver (DDWCP is not needed though can be useful for troubleshooting)."

  • Installed RealVNC free version
  • Installed Norton Ghost and backed up the computer in initial state.
  • Installed ASCOM platform 6 SP1 from here.
  • Installed TheSky Controlled Telescope Driver version 5.2.10 from here. During installation, chose the option to also have it install the source files.
  • Installed RCOS TCC:
  1. Get files from the RCOS support forum.
  2. The last full installer there is 1.5.17, so download and install that first.
  3. Then download and copy over just the 1.5.26 RCOS-ae.exe file into the Program Files -> RCOS TCC folder.
  4. These programs run into problems with Windows User Account Control, so either need to turn that off completely, or do the workaround below:
    • Make shortcuts for both RCOS TCC and RCOS-ae on the desktop.
    • Right-click and then choose "Properties", and in that dialog, select "Run as administrator."
    • Before the TCC starts up, you will then always be asked if it's OK to run a program from an unknown publisher - just click "Yes".
  5. In order for the TCC to run under Windows 7, the executable RCOS-ae has to be run first (just once per reboot). When you run this, nothing appears to happen, but if you don't the TCC won't work - you'll just get "unable to connect to RCOS COM object" or something like that. To make this simple:
    • Move the shortcut for RCOS-ae (created in the step above) into the "Startup" folder in the Start menu (under the "All Programs" listing).
    • RCOS-ae will then run at startup without any user intervention, and so starting the TCC is just as simple as before under Win XP - just double-click the file.
  6. Finally, to get the presets from the old install, copy over the file RCOS.dat from the Program Files -> RCOS TCC folder. This has those values saved.


  • TheSky 6:
  1. Use the DVD we have to install TheSky 6. (This installs 6.0.0.32; need to enter serial number from packaging.)
  2. Apply the updater to update to 6.0.0.65 (downloaded from here.)
  3. Apply a "daily build" update; download from here, then replace only the executable file "TheSky6". This updates to 6.0.0.68.
  4. To copy over previous settings:
    • Copy old Normal.sky file to new installation; it lives in Documents/Software Bisque/TheSky6/Documents. It should contain the Tpoint model as well.
    • Also copy over the folder "Field of View Indicators" in Documents/Software Bisque/TheSky6/.
  5. After telling it that the telescope is a Paramount ME, then also had to configure the server that lets it talk to ASCOM, via the "TheSky-controlled-telescope" plugin. To do this, go to Telescope -> Server Settings and check all the boxes there, as described in more detail here.
  6. To test the ASCOM connection to the telescope, I used the POTH tool that came with the ASCOM distribution. I found that I could not connect to the telescope unless I disabled Windows 7's UAC, so I did that (which makes some other things simpler, anyway).
  7. I also had to choose Telescope -> Options -> PEC -> Enable to get it to apply the periodic error correction (stored in the telescope's flash memory); it wasn't enabled by default.
  8. Note added after an evening test: All seemed to go pretty smoothly, with only the following small glitch:
    • When I first pointed the telescope, the pointing seemed to be off - I couldn't match the starfield I was seeing with what I expected. I wondered about the pointing model, so I double-clicked on the Tpoint model window in TheSky, and at first it said that it couldn't connect to the object. I tested to make sure I could (separately) start Tpoint, then I quit Tpoint and tried again to open the Tpoint model from within TheSky. This time it opened up fine, and showed me the data for the latest pointing model. After I closed that and tried again to point the telescope, the pointing was just fine. So it seems like somewhat TheSky wasn't connected properly to the pointing model? Not sure, but it seems to be working fine now.


  • Tpoint: Download the installer from here. Version 1.00.512; need to enter serial number from inside the manual.


  • Apogee camera and filter wheel. Before installing MaximDL, I wanted to make sure that the camera and filter wheel were talking to the USB bus for the computer.
  1. I downloaded the driver package from the Apogee website, [1], and unzipped and installed the drivers. On installing, I initially got the error "setupEventLogging has stopped working" and couldn't continue. I realized that I hadn't rebooted since disabling Windows UAC (a pretty major system change), so I rebooted and tried the install again, and it went smoothly the second time.
  2. I powered up the CCD camera and plugged it in. Windows recognized it, installed the necessary drivers, and then reported that it was working.
  3. Same for the filter wheel - powered it up, plugged it in, and Windows found it without incident.


  • Meade DSI camera: going to the Meade download site, [2], I wasn't sure whether to install the full Autostar Suite or just Envisage. The latter was smaller, so I went with that to see if it would work. It installed smoothly (version 7.09), and then when I plugged in the DSI camera (our guide camera), Windows recognized it and reported that it was working.


  • MaxImDL:
  1. Download latest version from Diffraction Limited website. Currently version 5.21.
  2. Installed just fine - I entered our license info, which is currently good through November 2012.
  3. Following instructions from Doug George in the Yahoo MaximDL group here, I copied over the Users\telescope\My Documents\MaxIm DL 5\Settings folder to the new installation to preserve our old settings.
  4. Testing this out, the connection to the telescope worked just fine. It didn't initially find the camera and filter wheel on the USB bus, so I went into the configuration and choose the Advanced option where it can search for a device. In this way, it easily found both the camera and filter wheel. I gave it the model of the filter wheel, AFW50-10S.
  5. For the filter wheel, after discovery it could talk to the filter wheel, but didn't have the correct filter names - it had some default set instead. To fix this, I quit MaximDL and then searched through the new and old settings files. It turns out that the filter names are in Settings/MaxImCCD/SetupFilterWheel.txt. That file had the old filter names in a set of entries with names like Camera0_ApogeeUSBFilterF1, Camera0_ApogeeUSBFilterF2, etc., but it also had a new, default set of filter names with entries like Camera0_ApogeeFilterF1, Camera0_ApogeeFilterF2, etc. (without the "USB" in the name). I copied the old set of filter names and used them to replace the new set, deleting the "USB" from all the names. Then when I restarted MaximDL, the correct set of filter names was there.
  6. Other than that, the MaximDL settings all seem to be there and to be working fine, though I haven't yet given it a full night-time test.
  7. Note added after an evening test: All seemed to go pretty smoothly, with only the following small glitch:
    • I needed to change/set the path to the Guide Star Catalog so that Pinpoint Astrometry could plate-solve images correctly.


  • Digital Dome Works version 5.2, revision date 12/02/2008, downloaded from here.
  1. When I tried to install it, I got an error that the file "VB6STKIT.dll" was not installed, and it couldn't proceed with the installation. Searching the web indicated that some other people have had this problem, too. I checked with Jerry Smith at TI, who offered to send me the file. In the end, I just copied the file from our old Windows XP installation onto the new machine - it goes in directory C:/Windows/System32/ . Putting this file in place then allowed the installation to run to completion.
  2. During the installation, the installer asked about overwriting certain already-installed system files; in each case, I chose to keep the (newer) installed file rather than overwrite it with the older version from the installer.
  3. To recover our old settings, I copied the file ddw.ini from our old installation into the folder Program Files/TIDigitalDomeWorks.
  4. Tested it out - seems to work fine.


  • ACP (version 5)
  1. Installed the latest version from the CD-ROM, then ran the installer to download and install the latest hot-fix from their website.
  2. GSC installer didn't work - confirmed with Bob Denny that this is a bug in the latest installer. Instead, I copied the GSC files directly off the CD-ROM onto the disk. I put them in Libraries\Documents\ACP Web Data\ (since that was already an ACP folder) and then ran the installer to configure ACP to use that location.
  3. Note after an evening test: When MaximDL is connected to the telescope through ACP (which is itself talking to TheSky), it seemed that using the "Nudge" panel in MaximDL did not actually move the telescope. I need to test this out more thoroughly, though. Entering coordinates to slew from MaximDL did work, however.


  • FocusMax.
  1. Got the latest FocusMax installer from here. Version 3.6.0.77. Haven't yet tried to transfer over our old settings and Vcurves.


  • Dimension4. Downloaded and installed Dimension4 (for keeping the clock accurately set) from here.


  • AAG CloudWatcher: Downloaded from here (version 7.10) and installed with no problems. Plugged the cloud sensor into the (only) hardwired serial port on the PC itself, so it is using COM1.
    • Currently it is logging its data to the PC, but we might want to set it up so that it copies its data and/or graphs to a directory available to the Linux machines so that it is web-accessible.
    • The default parameters for cloud sensing seem to work pretty well - on the evening of 08-23-2012 is was reporting "Cloudy" but not "Overcast", and looking at the moon, I could see that there was definitely significant haze/cirrus, but that the moon and some stars were still visible. Daytime test indicate that the boundaries between Clear/Cloudy and Cloudy/Overcast seem to be about right. Haven't had a chance to test the rain sensor yet.


  • PEMPro:
    • Downloaded and installed from here and entered the old license key we have.
    • Copied all the settings over from the "My Documents/CCDWare" folder from the old installation - chose not to replace the handful of files that the installer had created in that folder, but just let it copy the files that weren't already present there.
    • Tested it out on the night of 08-23-2012. The PEC table we had uploaded seemed off, so I measured a new one. I fit it with just a few frequencies (I think just 0.5, 1, and 2 times the worm period); the amplitude was roughly +/- 1.5 arcsec, maybe a little less. I chose "Create Paramount PEC table", said "Yes" when asked to invert the curve, and pasted it into TheSky. With this table uploaded, I re-measured the PEC. After 10 worm gear cycles, it appeared to be only noise, with an RMS of about 0.5 arcsec. Success! And no evidence of that weird phase shift we saw when we tried to do this a few years ago. Need to update our page on PEC.


Ports / connectivity

  • The Icron Ranger 2104 (4-port USB hub over ethernet) just worked - no installation of drivers needed. Plugged in a flash drive in the USB hub to test.
  • The Edgeport/4 (USB-serial converter), plugged into the Icron hub, did not work instantly. But without prompting, Windows found and installed the necessary drivers without incident, and then it started working. COM ports showed up in Device Manager. There is a standalone application that can configure the device to map its ports to particular COM port numbers if it's necessary to change these. For some reason, the order in which the ports became visible to the computer causes it to report that the serial ports 1,2,3,4 on the device are mapped to COM ports 4,5,7,6. (Note the swapped 7 and 6.) I used the tool to remap these to 4,5,6,7, but on reboot, they revert to 4,5,7,6. However: when I was setting up devices later, it turns out that the third port from the left on the hub really is COM 6, not COM 7, although the OS reports port 3 as being mapped to COM 7. So despite the odd reporting by the OS, it appears that the serial ports on the device really are, left to right, COM 4,5,6,7.
  • Current serial port usage:
  1. Port 1 / COM 4: TCC
  2. Port 2 / COM 5: Digital Dome Works
  3. Port 3 / COM 6: Telescope
  4. Port 4 / COM 7: empty