ditto what chris said.
when i ran into something like this in my "sparky" days (mostly non-audio, but electronics are electronics)... i'd set it up (preferably with an oscilloscope somewhere in the signal path) and then tap, rap, poke, prod, and wiggle anything I could get my hands on, trying to reproduce the problem... and this included the electronic components inside the box(s) using a plastic probe... just don't go too insane 'cause it usually doesn't take much if this is a "mechanical disconnect" problem... and if they contain hi-voltage or microwaves, you're on your own
i know zilch about a 896, but if you can hook this into a computer and somehow be able to see a "real time" output waveform (or
maybe just using any "blinky" VU lights),
along with solid tone inputs into the V3 (you're handle
is cleantone after all
)... that would act like an o'scope sitting on the ass end of things.
like you're suspecting, i'd start by wiggling that AES cable... some would say to just swap it out... but when dealing with intermittents, i'd be carrying this nagging negative in my back brain if it seemed to go away... i.e. "is the problem
really gone?"
but try digging around... we just had a jb3 on another thread that had the same type of issue after someone jerked the optic/line-in cable... found out the jack lifted slightly from the circuit board.