Which he apparently intends to keep secret.

I don't know why he is using an old CMOS ROM-based part, but the only way in which he can use it is in 803x mode, which means external memory, and consumes P0 and P2.

Why he thinks he wants to control the input clock with some sort of externally driven port I/O is not clear, though there could be reasons. However, I don't think he's read about the relationship between the input clock and the serial port baud rate yet.

The reason, if I understand the many dozens of items about this over the past few years, why people use that familiar 11.0592 MHz oscillator is that one can use a simple setup of timer and serial port using auto-reload mode and double-baud-rate bits to generate the "standard" baud rates, notably 9600 and 19200 baud. Timer 2 will do the job with a much wider range of frequencies, but the timer is not available for timing when one does that.

The way to time the operation of a simple EPROM programmer is with a timer or <sigh> a one-shot. Back in the "old days" when 5-volt EPROMs (notably the 2716) became available, it was simply necessary to write the address and data to the EPROM, then fire a 50 ms one-shot, that held the addresses and data steady, with Vpp ON all the while. Later, with the evolution of the 28-pin EPROMs, it became necessary to gate Vpp with that one-shot. As EPROMs got larger, it became desirable to find faster ways of programming these parts, but the old-slow method still works.

The fact that the O/P's goals are a secret, or perhaps a moving target doesn't help us help him.

If Mike Stegmaier would care to tell us precisely what he intends to do and what, exactly, his goals/requirements are, we could be of more help.

RE