Last time I had a problem with one of these, it wound up being either the D/A convertor or the switching setup just ahead of/behind it, I think the D/A. They have the strangest D/A deal I've ever seen. IIRC, and I wasn't quite sure at the time exactly what was going on, they use half the chips they should. They double up the outputs, and data rate, using a diabolical switching setup that struck me as the engineers saying "check this crap out, aren't we the cool kids" rather than being necessary. They switched the output value from every other sample (I think) to the alternate outputs, or somesuch. Somehow, they retained a 20k bandwidth, and a 480 doesn't sound 8 bit, so some juju was uttered to make this all work. I only mention this because you can really get caught up trying to trace down the circuit going WTF! No, it's not a stereo D/A chip. Yes, it's running two channels. No, you're not crazy (well, not because of this, anyway). Yes, it makes me want to damage somebody at Lex too. Knowing this makes it infinitely easier to trace back from the outputs. So check your D/As. PCM 53 or 54 I think, and a cluster of FET(?) switches near them (DIP14 or 16). I want to say DG308 or the ilk, but I might be stuck in V series land.
The other 480 issue I've run into more than once is the little output amp/line driver cards have bypass/coupling(?) caps that fail short (tants, i guess) and drag the beast down. Check your rails. They can usually be seen to be either cracked or seriously discolored with a loupe (ah, surfacemount and aging just don't go together). The little cards themselves are a bear to unsolder without losing any of the legs, which are soldered to both pieces. Solid core Cat5 or busbar can be used to replace any that go astray, but still.