Hi Erik,

Yes, I had the bad idea :) to update my parallel programmer (a copier between AT89C51 and 24C32) to accept also AT89S51. I can anytime forget about it and design instead a circuit using ISP. But what is really bother me now is this is the first time I notice a big difference between what it is said in the IC's datasheet and the experiment.

For instance, and as I said earlier, the programmer ALL-11 has no problem to read/write/lock the new AT89S51. So while waiting a logical answer, I think I had to check by myself the dynamic levels of the pins during the 'read mode' (on ALL-11) using my digital oscillospe because...... I have to know why :))

Kerim