1. So I assume you actually can toggle P3.1.
Did you check the pulse width of the signal? this way you'll know the system clock is correct. Make sure the crystal works correctly first.

2. When you test the short TX routine. Are there anything connected to the circuit?

I'll disconnect "EVERYTHING" that is not necessary for this test from all ports. Do not "assume", sometimes we "assume" a connection to an "input" pin of other device should not cause problems...AND you guess what!!! .

3. put RETI for all unused interrupt vector- just in case.

4. Remember, a simulator is a "SIMULATOR" only.

Good luck.

PS While it is unlikely...Are you sure that the CPU and EPROM circuit is connected correcly. Would that be the CPU not reading the right instruction at the right address?