Scanned the 90MHz patch cables and Heliax cable for faults with the Agilent Vector Network Analyzer (VNA) set to Distance to Fault (DTF) mode. The scan originating at the CER was clean with one suspicions marker near the 20m mark. The scan originating at the PSL did not record this marker, and was much noisier. We suspect this final patch cable between the patch panel, and the I/Q Demod is damaged and needs to be replaced.
After further review, the Balun was causing problems with the VNA. Upon removal of the Balun the plot started to look what it should look like. There is still a suspicious blip at 18.4m from the CER patch panel, but the blip is less of a disturbance than a connector.