He says he's verified that no such gratuitous connections exist. While he apparently has no oscilloscope, he does have a multimeter, and claims to have checked for short circuits to Vcc, GND, and adjacent pins. On the 805x, assuming a DIP-40, the adjacent pins are A(1) and Vcc, while, on the SRAM, they're A(1) and D(0). Unless there are other components not yet mentioned, it shouldn't be too hard to verify whether those are involved.

What's the effect of using int data type to preload the buffer and using unsigned char to fetch the data? Well, that could vary, but it wouldn't repeat the 2nd, 4th, 6th, and 8th characters, would it? There aren't many ways in which this set of symptoms could come about, aside from some wiring error, or are there?

I suspect an error in observation or in reporting what is observed. You fellows all have more experience with 'C' on the 805x than I. Perhaps you can explain what will happen if the data types are mixed inappropriately. I can't see how anything other than a short, either between Vcc and A(0) or between A(0) and A(1), would cause the problem in question. What troubles me is the report that using those mixed data types generates seemingly correct output. Can anyon explain that with anything other than the shorted A(0) line?

RE