There are a number of sites which interface microprocessors (in particulr 8031/8051 type processors) to an IDE drive, but they often use the 8255 PIO device. Since I have got more lazy the older I have got, I want my solution to use as few chips as possible, as this means as little soldering/construction as possible for me! The 8255 also has the strange property that if the control register is programmed to switch the direction of any of the ports the data on any output ports is set to zero. This caught me out on a project many years ago. It was interesting to see that the author of one of the sites had come across this problem and has worked around it with some extra hardware.
When the 2000 Maplin catalogue came out, a few new processor parts interested me. They seemed possibilities for the Atari IDE project.
I some point I stumbled upon an in-circuit programmer for the Atmel part based on the PC parallel port. I intended to use this design. It needed a 74LS126 quad tristate buffer. I didn't have one of these and a search of various suppliers in Everyday Electronics revealed that very few places stocked it. I considered using a 74LS125 (the enable is active low rather than active high), but wasn't happy at the necessity of adding an extra inverter! A further more recent search of the web produced a link to a UK site for another AT89S53 in-circuit programmer, but the page no longer existed. I could access the home page and found the email address of the web site owner. I emailed him asking where the web page had gone and the next day he replied sending me a html page for the construction of the programmer and full source code and Windows binary for the software he had developed. Good chap! This programmer dispensed with the tristate buffers and the only components needed were three 100ohm resistors whose only purpose was to try to match the impedance of the printer port output.