Some of this
http://digilander.iol.it/rar2k/TRS80/index.html
may be of interest.
Phil W7OX
On 11/6/16 6:00 PM, Josh Malone wrote:
I would guess a bad RAM chip, too. The shared
address/data bus lines lead to weird behavior
when an address of RAM is not responding
properly. I'm currently working on fixing my 102
that also has symptoms of RAM failure.
https://www.vintageboot.net/2016/10/tandy-102-ram-exam/
On Nov 6, 2016 7:53 PM, "Gary Snethen"
<gls...@snethen.com <mailto:gls...@snethen.com>>
wrote:
Hello Everyone!
Last week, I picked up a Model 100 (8K) at
an amateur radio club meeting.
I'm fixing it up and could use some advice
on a problem I've encountered.
When I turn it on and enter a BASIC program,
specific bits of memory are consistently
corrupted.
Some examples (these lines are from
different experiments -- not a single
program as shown):
10 PRINT "Hello!" ==> 10 PRINT "Hellg!"
20 REM blah blah ==> 20 REM bdah blah
30 FOR B=0 TO E STEP A ==> 30 FOR BAND0 TO E
STEP A
The first two examples suggest that bit 3
(0x08) is getting set when it shouldn't be.
Maybe someone can confirm whether that's the
case in line 30 above as well. I don't have
a list of the Model 100 BASIC op codes.
To check whether it was happening at a
consistent stride in memory, I tried this:
clear
Ok
10 rem
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
list 10
10 REM
aaaaaaaaaiaaaaaaaaaaaaaaaiaaaaaaaaaaaaaaaiaaaaaaaaaaaaaaaiaaaaaaaaaaaaaaaiaaaaaa
Ok
Note that every 16th letter 'a' had bit 3
set and turned into a letter 'i'. This
leads me to think the most likely culprit is
a bad memory chip. I don't know much about
the memory used in the Model 100. Does a 16
byte stride between bad bits make sense
based on the RAM chips and layout used?
Can anyone with more experience think of
other potential causes? What experiments
should I try to rule in or out those other
causes?
I could just start poking around with my
scope, but thought it best to ask first in
case this is a familiar problem.
Thanks!
---Gary
