>The noise floor is really decreasing the utility of the new >repeater. The noise source seems to come and go as a quiet signal >on the repeater input can become suddenly noisey, and vice versa - a >noisey signal can become suddenly quiet.
It looks like you have a 10dB degradation. Many times this is from phase noise from other transmitters, but intermod can give you the same problems. I would look for a VHF paging transmitter if phase noise is suspected. This type of troubleshooting requires a good spectrum analyzer in most cases. I typically use the RX part of a duplexer on the front end of an ANRITSU 2721B to keep the first mixer in the spectrum analyzer happy. I hook up a 10dBd Yagi to the duplexer and start looking for the noise floor to rise. Once I have found the site that is causing the increase in noise floor, then the hard part comes, getting the other site to cooperate in further testing. What we ultimately want is for the other site to completely shut down for the time it takes to test the sensitivity of the receiver that is experiencing the degradation. Since you mention that the degradation is intermittent, you may be able to monitor other signals and see a correlation between when the suspect transmitter keys, and an increase in the noise floor in your bandpass. If the problem is because of intermod, then it becomes a little more difficult as you have multiple culprits. The spectrum analyzer you use should have a noise floor of around -120 dBm @10 KHz bandwidth. 73, Mark N5RFX