Well, Doug, you are from Oxford and I am... you know... from Bratislava. It's me who expects that you will know the most scientific answers to your questions... :-)

We have been discussing '51 testing here a couple of weeks (or months was that already?) ago, mainly fuelled by Richard Erlacher's need to keep some ancient '51 alive. You can get inspired there.

You might also go to the manufacturer of your part and ask him for some hints.

Mine favourite is the "differential-monte-carlo-ish": execute a random stream of data on both, dumping from time to time an assortment of registers and memory and comparing.

But, again: if you'd tell us what do you want to achieve, we might perhaps help you more.

JW


PS. I dig out the link for you: http://www.8052.com/forum/thr...ead=130913