In addition to the pressing matter of shipping the hardware out, several tests and small repairs were performed during this trip. The biggest of which was the complete recalibration and reprogramming of the phasing cards. During subsequent builds of Christmas Valley and Adak, a different scheme for how the phasing cards operated was implemented. This change in scheme allowed for better control over the signal level output of each phasing card to try to be uniform across all beams and all cards.
To recalibrate all of the cards, the network analyzer was carted out to the site and hooked up to the local network. Then, code written by Jef Spaleta at the University of Alaska Fairbanks was run which issued commands to and recorded readings from the network analyzer. The code here automated the process for making signal level as well as phase measurements with each possible combination of components switched in to the circuit in each phasing card. As there are about 15 delay line phase sections (some sections were only measured once in the interest of time) and a little more than 5 attenuators, the recalibration for each phasing card took approximately 20 minutes to run. To do each of the 40 phasing cards, this took the better part of 2 days' time.
Also during this recalibration process, any bad delay line sections, attenuators, or amps could be identified as the signal level measurement would present an odd measurement. As well, control of each phasing card was checked by performing the calibrations. The control signals all checked out, however 3 interferometer phasing cards were found to be broken across all combinations of delay line sections. These cards will be shipped to the University of Alaska Fairbanks for repair. Otherwise the phasing cards were reprogrammed and relatively even signal output was observed at the output of some of the transmitters.
The other major task during this trip was upgrading the QNX computer's operating software from the 6.4.2 version to the 6.5.0 version. The QNX computer serves as the main brain to control all of the electronics in the MSI design. In previous tests to this trip, Jef Spaleta found that high network traffic would slow processing times on the 6.4.2 OS. These same test showed that the 6.5.0 OS was much more capable of handling high network traffic while crunching out processes. The OS for the QNX computer was changed on May 2nd with a completely different hard drive being installed in the computer case. It was found that keeping the same radar operating software code, which runs the control programs and hardware drivers, which the new QNX 6.5.0 operating software caused the radar to lock up at random time intervals. In the weeks following the trip, further software upgrades were made such that the Fort Hays radars are now running the latest hardware drivers as well as the latest version of the Radar Software Toolkit. Some small tweaking to the code may be necessary, but the radar is running in a much more stable state than the days following this trip.
In addition to these tasks, the interface box was repaired to restore the 3rd, spare output of the T/R, TX, and Sync signals. Prior to this trip, the 3rd T/R output was found to be not functioning, most likely due to a bad output from a 74LS244 line driver chip. Also, during this trip, some quick diagnostics were performed at a few places along both of the radars where high reflected power was causing the transmitters to turn off. In all, three transmitters on the East and West radars had to be turned off due to high SWR readings looking outside of the building. Otherwise the antenna arrays held up fairly well over the winter with just two places between both radars that will need attention in the future.
