Dear folks,

Thanks to all of your inputs - Russell, Oleg, and Kai - I am now more clear on the issue. When the setup showed up random error, I was kind off convinced that there was some timing issue - but would never have analysed the way you people did that. And Kai, a special thanks for your effort in working out the time budget.

Just to check out, I switched the crystal to 6MHz and ran the same setup in 6 clock mode - the problem defenitley got worse and started working "normally at times". The suggestion to use the rising edge of the !WR pulse is a good idea and I will go for it in the next PCB revision, that is underway.

And Russell, you had queried as to how the Direction pin of HC245 was handled - well it is connected to the !RD pin of the MCU via an inverter gate. Reason why I did not bring it into picture. Also the other point that I did not mention is that this IO expansion services 2 DACs, one 12 bit ADC, two 8 bit input ports, two 8bit output ports, one printer port, and an LCD display. For simplicity sake I had just showed one LED port.

Now that the root cause is known, I feel so confident at tackling the issue.

Thanks to all of you.

Raghu

PS : Steve, this thread can close at this stage, but Kai has opened a discussion on RC delays - let it continue...please.