At 08:50 2.9.98 -0400, you wrote:
>Are you sure the pins have not changed to outputs? Test using a 10K
>resistor, try to pull the pin up to +5 and also down to ground. If you
>cannot pull the pin with 10K then it is an output.

Yes, they follow whatever level I apply to them not only with 10K resistor,
but even with wet skin. Some more experiments revealed, that it works wiht
external RC clock (after fixing the CNT0/1 mistake), but when I change the
config word from 0ffb to 0ffa, it has no effect, e.g. the chip still works
with external RC clock, when I disconect the resistor and capacitor, it
stops working - no internal clock. It looks like some problem with the
programming config. I even "sacrified" one plastic package, but it behaves
the same. The programming algorithm was adjusted recently, wasn't it. I'll
download fresh info in the evening and let you know. Thanks so far for
suggestins.

Josef

>I think your problem is above. You probably meant to use CNT1 as the
>second varaible. Using CNT0 for both will cause the loop to never exit.

Thanks