Because 75dB is just not adequate isolation @600kHz separation for a 50w+ repeater. On VHF, in amateur service, or LMR service with less than 1.5MHz separation between receive and transmit, I would use no less than a 6 cavity duplexer. I've experienced desense, as a user, with only 4 cavity duplexers in ~500kHz frequency splits(I had seen the way the company I worked for, set up their repeaters.
I've tuned lots of BpBr Sinclair ResLoc duplexers, and I have never experienced the issue you are seeing. They are usually absolutely rock solid onced tuned with the proper gear. My only question is if this site is temperature controlled(I don't recall if thay was asked)
Yes, the site is climate controlled. Duplexer problem has been addressed and fixed due to two faulty trimmer caps. It's been running rock solid as you say for about a month now.
It has been said on this thread that 75 dB is "not adequate isolation at 600 KHz separation for a 50 watt repeater." What exactly dictates this? The purpose of the duplexer is to allow the transmitter and receiver to operate on the same antenna with optimally zero effect with each other. If I can run the power I intend to run and get,
zero transmitter-to-receiver desense or degradation in performance when the transmitter kicks on, then it seems as if I certainly DO have enough isolation. What else could I possibly need? I would totally understand how I would need more isolation if I kicked the transmitter on and it desensed the receiver by 10 dB... but I don't have that issue, running the test into a good dummy load -AND- running the test into the site antenna.. all the same.
Doing some math here.. the duplexer has 75 dB of notch per side, giving 150 dB of
total isolation between the two frequencies. The transmit power is 50 watts, which is +47 dBm. +47 dBm minus the 150 dB of isolation gives a -103 dBm result, meaning the receiver at 147.600 MHz for example will "see" the 147.000 MHz transmitter with a signal level of -103, or about 1.6 microvolts. Any decent receiver with a clean front end amplifier stage and especially a clean I.F. chain will certainly tolerate this amount of signal 600 KHz away without any degradation. This has been my experience anyway. Now, if you use two crappy mobile radios back-to-back (like a lot of so-called "repeaters" are made of these days) then no way..
I totally agree that a six-cavity, >100 dB rejection duplexer would certainly be needed if you're running a 250W power amp on the TX and a GaAsfeT preamp on the RX and a 600 KHz split.. yes, I agree that 75 dB would definitely fall short. No argument here. I'm just saying it's working for me, maybe my repeater TX is just super clean and I got lucky.