The Tenax, Silica Gel, Charcoal #10 trap tends to hold a lot of water which gets desorbed onto the MS, the water loading of the source on the 5972 can cause some problems like this.
We moved away from this trap many years ago and went to the VoCarb3000 which can use the dry purge to remove some of the water, it does help some.
Is this going through a split injection port? If so you may want to increase the desorb flow/split ratio to also introduce less water and methanol into the system. The higher flow rate will also help pull things off the trap better.
The one thing we did to improve stability for instruments like this was switch to the ETP ribbon style Electron Multiplier, they seem to be less prone to having sensitivity changes as you load the instrument up with organics. One thing that I used to notice the the K&M was when you get hit with something such as gasoline at high levels in a sample was that the sensitivity would increase for several injections afterwards before returning to normal, similar problems with high level standards.
