Installation of prototype 1

The system was installed in the late afternoon / evening of Dec 30, 2010, with the help of Paul Stewart, Josh Walawender and Ben Berkey.

Video of prototype 1 installation at Mauna Loa Observatory (Dec 30, 2010).

Jan 2011: Snow storms

A few days after installation, the first hard test came for the system: I left Hawaii for a few weeks, and, the day after I boarded the plane, a snow storm came through. I was remotely monitoring system health, and did not run observations, due to bad weather and fear something would go wrong while I'm too far to quickly get to the system. Overall, things went fairly well, with only two failures:

Water got into a USB connector on the outside of the main electronics enclosure. The connector was for one of the webcam, which stopped working
Water also got into the connector for the AC power monitoring: the software thought that AC power was off upstream of the UPS. This was easily fixed by changing the voltage limit in the software while waiting for a permanent fix

Both issues were fixed by drying and cleaning the connectors. The origin of the problem was aluminum tape around the connector, originally thought to be able to keep water from getting inside the connection. The aluminum tape was removed and the connectors were sealed with silicone.

Snow-covered Mauna Loa observatory, with Mauna Kea in the background, Jan 8, 2011.
Image by Mauna Loa Observatory North camera.

Snow-covered Mauna Loa observatory, looking East, Jan 8, 2011. The camera system is just outside the image, to the right.
Image by Mauna Loa Observatory tower camera.

Snow-covered camera and mount, Jan 19, 2011.
Image by Ben Berkey, Mauna Loa Observatory

Snow-covered camera and mount, wide view, with VYSOS dome in the background, Jan 2011.
Image by Josh Walawender, VYSOS / University of Hawaii.

Feb 2011: Final tests and start of robotic operation

By mid-feb 2011, hardware was stable, and software was ready for robotic operation. The weather was bad for most of this time, so little data was taken.

Image of a field centered near M104 (Sombrero galaxy), Feb, 15 2011 (UT).
Image acquired with partial moonlight and through thin clouds. Raw image, no processing (no flat field or bias subtraction), binned 2x2. M104 is visible below and slightly to the right of the center. 125 sec exposure at ISO 200, F1.2.

The system started robotic observations on Feb 16 UT.

March 1, 2011: first clear night with fully working system

Image of a field centered near the Rosette Nebula, March 1, 2011 (UT).
Single 315 sec exposure time at ISO 100.

Image of a field centered near Antares, March 1, 2011 (UT).
Single 362 sec exposure time at ISO 100.

March 30, 2011

Image of a field centered near Z Cma, March 30, 2011 (UT).
Single 303 sec exposure time at ISO 100.

Apr 2011 - Apr 2012: System works smoothly

Continuous robotic operation for one year, with no failure.

Apr 11, 2012 - Aug 2012: System failure due to power cable, and its ripple effects

During the night of Apr 11 2012 (UT), the cable carrying the power to the mount's motors failed. I had used a standard coax cable, which does not have a braided core, and part of the cable was bent with a small radius of curvature. After more than one year of bending the cable back and forth, the core snapped during the night, and the mount was left stuck pointing up. Thanks to Ben Berkey (MLO) for noticing quickly that something was wrong, and for covering up the system the next day. The main risk for the system is that the Sun enters the FOV during the day and damages the camera/lens. The system was stuck in a position such that the Sun did not enter the FOV, but passed fairly close to it (more on this later on).

I took the system down to Hilo. Finding and fixing the problem was fairly easy, and I took advantage of this interruption to perform various minor upgrades on the system. However, after the system was re-installed at MLO, bright stars were surrounded by a faint large halo of bluish light, as can be seen in the example picture below.

Image showing the halo around bright stars, July 27, 2012 (UT).

These annoying halos were seen in both cameras, and did not go away. After a few weeks, I decided to investigate the problem on-site by removing the lenses from the camera bodies. I was horrified to find that the sunlight had melted plastic inside the camera body and that the fumes had coated the back of the lens assembly for both lenses.

Backside of the 85mmF1.2 lenses, with a haze clearly visible. The haze is due to sunlight burning plastic inside the camera after the power cable failure.

Cleaning this haze was difficult, as (1) conventional lens cleaning solutions did not do anything and (2) the fumes had moved inside the lens assembly and coated more that one surface. I used acetone to clean the haze, and partial disassembly of the lenses was necessary to reach all surfaces affected. The cleanup was very successful.

Aug 16, 2012: System back online, with no halo around bright stars

The system is back online since Aug 16, 2012, with no evidence of the halo around bright stars. The image below shows that even around a bright source (Jupiter), the bluish halo is gone.

Jupiter in the Hyades, Aug 16, 2012 (UT).