Posted By
JamesC
on 2014-09-05
19:35:14
|  Re: New to Plus4 scene - Introduction and SD2IEC and 1541 loading issues
@Siz - only TI$ is held at the top of memory in my ram-test. Numeric variables are held immediately after the end of BASIC program code. 
TI$ is re-calculated from the jiffy clock on each request, it's not updated from the string pointer in RAM. Proof: run the ram-test for a few seconds, then pop into MONITOR, set 07F8 to 80 (so we can see RAM under ROM), and look at FCF8-FCFF. Change them all to FF, exit the monitor, and then PRINT TI$.
@Paradroid - yes, you have a bad memory location. Probably several. Here's how to tell:
a. Power down your Plus/4, leave it a minute or so (or put the original Kernal in!), then turn it on.
b. Type MONITOR and press Return.
c. Type M 07F8 07F8 and press Return.
d. Cursor up to the first 00 after 07F8 and change it to 80. Press Return.
e. Type M F9D0 and press Return.
f. Look for non-00 and non-FF values. Keep typing M and pressing Return until you reach FD00 -- at this point you're in the ACIA, not RAM, so we can ignore anything above FCFF.
We know F9D1 is bad, as it crashed the ram-test (the last number displayed was the last good byte). Odds are you'll see either 8-16 bad bytes in a row, or a quickly noticable pattern.
If you see no non-00, non-FF values, then power down, put JiffyDOS back in, and repeat this procedure. There's a chance that JiffyDOS itself is corrupting these values. (As I said before, JiffyDOS is buggy....)
|
