Jan,

This is the o/p from avrdude -t

avrdude> >>> read lock 
0000  ff                                                |.               |

avrdude> >>> read fuse 
0000  0a                                                |                |

avrdude> >>> read atmelsig 
0000  1e 73 ff ff 24 ff ff ff  ff ff ff 16 17 12 13 ff  |.s..$...........|

avrdude> >>> read usersig 
0000  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
0010  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
0020  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|
0030  ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff  |................|

avrdude> >>> q 

 

I presume that these 'extra' bytes will refer to the revision.
I bought the chip last week from RS. The front markings on the chip say:
AT89S8253 24PU 1024. I presume that means week#24 in 2010.

Yes, I have tried the 4M loading resistor. It makes no difference. Nor does a 1M. Obviously lower values kill everything.

I would guess that the XTAL1,2 inverting amplifier has a massive gain. And there does not seem to be a way to disable the internal RC oscillator. e.g. I think that this is interfering with the crystal.

It is obviously not much of a problem to just reduce the voltage for ISP, but I would like to know exactly what is happening. And it would be nice to have a foolproof fix.

David.