Hallo Oleg,

that was a great reply, and extremely useful for me!



Yes, they even want to have implemented an additional 'on die' voltage regulator for the cpu core with some RD2s, right?

Right now I'm in contact with atmel, because power-on-reset issue with AT89S52 when invoking serial programming session. I have found out, by simply watching the reset line at printer port, that under certain probable circumstances the micro isn't properly resetted at power-up by the their ISP-programmer!! I contacted them, but they meant everything is fine... Well, I cannot agree very much, as I KNOW, that proper power-on-reset AND proper brown-out-reset at power-down IS an issue with flash micros!! Finally it might happen what you also painfully experienced, that you program the micro, make a power-down, and, without noticing anything, some lock bits might have changed...

That would be a total desaster for a company, and it makes me really angry, that atmel doesn't focuse on this problem! Up to this moment I haven't found any publishing, which at least gives a hint how to handle reset line of AT89S52 at power-up and power-down with ISP-programming. Of course, I know that there are people, who tell, that everything is fine, and that they had never problems with it. But this cannot really make me happy, as long as I also hear other stories, like your's.

I think it's a good idea to ALWAYS implement a proper and trustworthy and external power-on-reset and brown-out-reset feature, WHATEVER is planned to do with AT89S52 and similars.

Thank you very much for your detailed reply and for sharing your unlucky experiences with us!

Kai