On 11/30/2015 12:18 PM, Terry Stewart wrote:
Speaking of Schrodinger's feline, here are details of my recent Apple II+
repair for those who might be interested:
http://www.classic-computers.org.nz/blog/2015-11-29-more-repairs-to-my-appleII+.htm
Terry (Tez)
Terry,
nice dialog on your repair job.
On the last comment about the ground pin of the defective rom having a signal,
if the apple board is a 4 layer board the ripple from the short to the internal
signals from address current, or other signal current being propagated to the
ground pin, I suspect the resistance in the pin itself may have provided the
needed high resistance to show the signal. Unless you scrap the ground solder
protect off and look at the voltage out in the actual ground conductor, I
suspect the voltage went down to a very low level very close to the pin.
Also where were the decoupling capacitors located with respect to the pin. I
suspect that might have gotten rid of more of the voltage, but they were
probably nearer the Vcc end of the chip.
If you can track down the schematic, it might be that your missing pin doesn't
do much unless you perform some special operation, such as some controller
addressing or memory operation or such that you don't normally do. It may have
also had a fit to the other part of the pin if it was present in the socket to
actually work. I didn't hear if you found that, or maybe it fell off when you
pulled the chip out?
I suspect the short developed from your theory about stress, or perhaps the
chip was programmed by a bad programmer. We had a programmer that we found
developed a tendency to program eproms and like programmable chips and it
destroyed the chips capability to actually reach ground again.
The programmer made chips that verified, but when you ran them in a circuit and
probed the lines with some sync to the system clock, rather than seeing the
signals on the data lines going to zero on the datalines, you could see a
hodgpodge of crap at 1.5 to 3 volts which is TTL la-la land. The chips
programmed in such a programmer as a properly working Data I/O had clean lines
as did reference chips from years earlier.
Due to the fact we didn't program many chips, and I found a cheap programmer to
hook to a PC, we never found out what broke in our programmer (which was a home
design admittedly). But it was build to standard, but had something happen to
start killing eproms. So that sort of fault may have been induced in your chip
and got bad enough to kill off your Apple some years later.
thanks
Jim