Technical memory question

Boray · Post by **Boray** » Mon Jul 26, 2004 4:57 am

Thanks! I made it shorter now. It goes just outside the cart + maybe 5 cm. Thanks again!

/Anders

Boray · Post by **Boray** » Fri Jul 30, 2004 4:35 am

Something strange happens when breaking most plug-in games the
way you describe. When switching block 5 off, the vertical
stripes/shades in the display gets much more visible and stays
that way until I reset. Why is that? And does this indicate that
this procedure isn't very healthy for the computer? Maybe it's
better to turn the computer off, switch block 5 off, turn the
computer on and then switch blosk 5 on?

/Anders

Leeeeee · Post by **Leeeeee** » Fri Jul 30, 2004 5:18 pm

The vertical stripes/shades you see are noise which are mostly generated by the CPU. When the CPU is running you don't see much of this noise as it's constantly changing but when the CPU is crashed it becomes more visible. Like the noise on a webcam feed is not very visible but the noise one one single still frame from that webcam would be.

I wouldn't worry.

Lee.

Boray · Post by **Boray** » Sat Jul 31, 2004 3:34 am

Thanks! So the CPU stops completely when the memory dissapears? Like it has no place to go for the next instruction? I thought it would see random instructions or something...

/Anders

Leeeeee · Post by **Leeeeee** » Sat Jul 31, 2004 4:50 pm

It usually continues trying to execute whatever it finds until it runs into the equivalent of ..

Code: Select all

here:
	BRA	here

or some other short loop.

I can't remember if there are any (illegal) instructions which effectively halt the 6502 (unlike HCF on the 6800) but the next interrupt would start it again anyway as the RTI would point to the next effective instruction.

If you really want to know ehere it's gone I can probably hook mine up to my logic analyser and 'crash' it in the same way. Whatever it's doing it won't hurt though.

Lee.

Boray · Post by **Boray** » Sun Aug 01, 2004 5:18 am

I noticed something more strange on one of my vic-20 CRs... (One that has no socketed chips). When running without memory in block 1, the vertical
stripes/shades in the display gets much worse than with expansion
ram in block 1. And as the screen is moved in memory when there is ram in
block 1, I think the difference in these stripes are dependent on
where in the internal memory the vic chip reads??? Could that explain it? Any idea of if
this could be fixed? The stripes are very much visible even on
the default white display. But with expansion ram in block 1, they are
almost not visible at all...

/Anders