Thanks for your patience, I can't seem to upload any more pictures (I can email them to you if you like?), so I'll try to describe the encoder and application.
It is TTL as you say, there are only four terminals used on the plug: A, B, and supply + & -. Input shaft is I guess 12mm, mounting is 3 x M6 holes around the input shaft, approx 40mm pcd. There is a "window" on the back end plate through which a row of LEDs can be seen (two light up on power up and stay on all the time, none of the others ever light)
The application is on the carriage head of a band mill, used to measure the position of the carriage (and therefore the tree trunk) relative to the saw blade. The control system which the signal goes into is a system designed specifically for this application, I guess 1980s vintage retrofitted to this machine, replacing a valve system. The "computer" is a self contained operator interface, mains and encoder in, with relay outputs to drive carriage backwards and forwards.
Current situation is that carriage (and encoder) been stored outside over winter, worked previously (though I didn't see it work). Customer has spare computer, swapped this, same outcome (i.e. measured value display doesn't change when encoder shaft turned). Shaft turns freely, LEDs on back of encoder don't change. Continuity of all four cores between encoder plug and computer input terminals checked and fine.
I don't have access to a scope (and as I haven't used one for the best part of 30 years, I'm not sure one would be much use!) but from the above I'm 90% sure it's the encoder - but I'm always open to other ideas.