Problem(s) Encountered:

After observing for about an hour, some dozens or hundreds of bright columns started to appear on the CCD, but not the same ones every time.  There was some low-level horizontal banding as well.  They are present in bias frames as well as sky exposures.  It was very windy from the NW, and there is some suspicion that this change was related to a move to the SW. I couldn''t reload the Owl software using the setup function, as an error message appeared. I eventually tried every combination of restarting Owl, powering down the CCD controller, rebooting the mdmarc2 computer, and reseating the fiber optics at both the controller end and the computer end, all about 6-7 times.  Waiting for Eric to call back, I decided to try one more reboot, and also reseated the bnc connector from the filter wheel to the CCD controller, and Owl finally loaded the software.  However, when starting to observe again, the bright columns came back!  This time, with some patience, I continued to take bias frames and other exposures until the bad columns gradually went away.

Solutions:

Found that the issue would crop up initially whenever the CCD was cycled or reset.  After numerous frames were taken, the lines would either improve or disappear.  One thing I did notice were that the fibers demonstrated far better loss numbers after [softly] cleaning them.  Also, the spare fiber running from the computer room to the dome seems to be a fully functional fiber as well, with good loss numbers.  In fact, it is now swapped in and in use.  

I was able to get the lines to appear with two of the CCD control boxes.  When I set things back up with the silver box, the lines never recurred.  After another night on sky, they still have not come back.  Perhaps the controller just was acting out.  Perhaps it is a sign of some component in the CCD head electronics starting to go.  Regardless, it looks better and we’ll have to wait for a more persistent failure before we can stress test it any further.