Thinking of keyboard interfaces (HP9915)

I've not seen oen either. For digital signals (which keyhorad scan lines tend to be), and if you know which are the driven lines and which are the inputs (pretty much esenetial if you're goign to use the diodes anyway), then a digital cicuit solution is possible. Store the simulated key states in D-type flip-flops. The output of each D-type feeds on input of an AND gate, the other input is driven by the appropriate scan line, the outptus of all the AND gates sorresponding to a given input line from the keyboard are ORed together and fed to that input. (OK, you may need to invert the scan lines and inputs, but I mean 'logically' ANDing and ORing here). This works with machines that allow multiple scan lines to be driven simultaneously, like the TRS-80 M1/3/4. The downside is that htis logic is really only easy to implement usign an FPGA or simiular. Building it -tony