GARTHWILSON wrote:
It seems like there should be an additional digit after the P, indicating the speed. Who made it? The marked speed is the minimum speed where it is guaranteed to not only work but also meet the timing margins in the data sheet. It will probably work a little faster, and it may work a lot faster. I don't think I've ever heard of one specified for 8MHz though. I've heard of 4, 6, 10, and 14MHz
It sounds as though he may be thinking of a part sold in the early 1990s. That number was on the MPU that was in a terminal server for which I wrote a BIOS and other code. The 'P', if I correctly recall, meant a PDIP40 package. It wasn't until the mid-1990s that WDC converted everything to a fully static core, which was represented with an 'S' in the part number (e.g., W65C816S).
In those days, WDC did sell different speed grades (faster = more money, of course), and the speed stamped on the package was the maximum at which the MPU was guaranteed to perform and stay within the timing specs. Wafer yields back then were not as consistent as they are now, so I suspect a certain amount of "cherry-picking" went on to determine just how fast a given part could run before messing up.
With today's MPUs, the F Max vs. VDD curve (page 27 of the data sheet) suggests that all current parts are capable of 20 MHz, assuming reasonable cooling. Of course, as Garth said, a circuit that can support that speed is
de rigueur.
![Smile :)](./images/smilies/icon_smile.gif)