Hi,

I am 55 but a newbie here :)
In the last 3 days, I lived a sort of dilemma with my simple prototype. It is a stand-alone programmer to copy the content of 24C32 into AT89S51. I already did it for AT89C51 and it includes all possible functions (read, write, verify, lock...). By the way, I use an IR remote control unit to send a command because that needs 1 pin for all keys ;)

Here is my story:
I have 2 programmed but unlocked MCUs; AT89C51 and AT89S51.
The first function to test on my new prototype was to 'read' the internal code of the MCU. As you know it is the simplest one. The written bytes in AT89C51 was read as expected but the AT89S51 gave all 1's instead (as if it is locked or blank). On the other hand, my old DOS programmer (EMP-20) has read normally both ICs though the selected device is AT89C51 (not yet updated for the new 89S51!).

So I rewrote my program (of the prototype) to read the MCU code one byte at a time and in the halt state I checked the voltage (using an oscilloscope) of all 40 pins also one by one. All pins were as required for the read procedure in the 89S51 datasheet. And the 8MHz crystal was running fine as well. The dilemma I am facing is that under the same condition only 89C51 outputed its correct byte while 89S51 always gave me 0FFH though in both cases P3.3 is set to low as mentioned in 89S51 table (P3.3 doesn't exist in 89C51 table).
I even tried to momentarely make the reset pin low then high in both cases; with or without oscillation (XTL1 to ground).

After all... perhaps I am getting old :))

Kerim