This is a universal truism that can also be expressed as “Murphy’s Law.” I don’t rightly know how Murphy received credit for this, however, I chalk it up to either the luck of the Irish or the gift of self-promotion. Either way, that principle was demonstrated again with a 950 MHz STL link between Mt. Beacon and Peekskill, NY for WHUD.
I noticed, while doing some transmitter maintenance, the receive signal strength of the STL had dropped from 300 µV to 30 µV. That is an alarming development. Therefore, we scheduled a tower crew for the next day, not wanting to go off the air over the coming holiday, which would be a sure bet otherwise. Upon arrival, the tower crew noticed a strange thing in the STL transmission line at the base of the tower, which looked like some type of a splice. Truth be told, I have been associated with this station since 1999 and had never noticed the splice before. This STL system was installed in 1998 when the station’s studio moved from Peekskill to Beacon. I can say, of all the things that have gone wrong over the years, this STL system was always very reliable. Regardless of that, I quick check with a spectrum analyzer showed a 3 dB return loss at 137 feet (41.75 m), exactly the distance from the transmitter room to the base of the tower.

A 3 dB return loss coincides exactly with the drop in received signal strength at the other end of the path. Thus, the tower crew took apart the splice and water poured out of it. I would estimate at least 4-6 ounces of water (180 ml), perhaps more.

We then began to take in the details:
- The 7/8 coax coming out of the building was Cablewave FLC78-50J
- The 7/8 coax going up the tower was Andrew LDF4-50A
- The splice connector was Andrew L45Z
- The center conductor threaded connector did not fit properly into the Cablewave cable, it was too loose.
- The cable was chaffing on a tower leg, about 50 feet above the splice because it was not properly secured to the tower
- The 7/8 splice connector was missing an O ring on the backnut of the Cablewave cable
Thus, water ingress causes the high return loss. Problems with this system began immediately after Hurricane Irene, at the end of last August. We were able to make a temporary fix using two type N connectors of the proper manufacturer for each type of cable. The radio station returned to air just before noon, about 45 minutes after turn off. After the repair, the return loss dropped to about 20 dB, which is good.
The permanent fix is for the entire run of cable from the transmitter room to the STL antenna to be replaced. That type of line splice should have never been used on a 950 MHz STL, and it was certainly wrong to mix cable types with an Andrew connector. Those little details will always manifest themselves eventually.