Gregory's Server|
25 comments
 this is an indirect jump to the console output routine. it's supposed to beep. but it's jumping to 0000! on initial boot, A600 etc is supposed to alias to 8F80 which is in ROM. the jump table is getting copied into the 6532 RIOT but the write is getting corrupted. it's supposed to be '89' but something else is trying to drive the bus at the same time thanks to the magic of the 16822 logic analyzer trigger output, I can see that the 6532's R/W# never gets driven low.  @tubetime  @tubetime If I had enough $, I'd design a Super-RIIOT IC (SPI Ultra Precision Extra ROM ROM I2C I/O Timer... acronym needs work) sure enough, bad inverter. looks like it was bad back in the day--someone removed the MM to 45 jumper. @tubetime What would've been the purpose of the MM to 45 jumper then? @sergioaguayogr the memory in the 6532 is used as a vector table. it's nice to protect it from accidental modification by a runaway program, so MM to 45 is a write protect function. seems to be a similar issue. when it tries to jump to outvec+1 (previously stored in the 6532's RAM) it goes to 0000 instead of the actual character output routine pin 11 of this inverter is shorted to ground. this is the exact same problem as the last SYM-1! so strange. different manufacturers and exactly the same failure (~20 ohm short from pin 11 to ground). must be damage caused by something the last owner did. woot! |
@tubetime i see the code