CBM 8032 w/ 64K RAM Board... 83 bytes free?

[phreakindee]

I recently was lucky enough to find an 8032 for $79 on eBay, great condition. It arrived in wonderful shape and booted instantly! However, I noticed something odd...

http://twitpic.com/8hq3qy
http://twitpic.com/8hq7h0

Turns out it has what I figure is the 64K memory board. Fab number 8032109, art 8032108. Yet it's showing 83 bytes free on the boot screen.

So... any input as to why this might be? Seems to be that the board is all hooked up properly, at least from my little CBM knowledge. Total PET newb here. I don't currently have any software to test, other than some simplistic type-in BASIC programs which seemed to work. Thanks!

 

[bitfixer]

Sounds like faulty RAM to me. If it displays text and everything then it's mostly working, but probably one or more RAM chips are bad.

 

[phreakindee]

Are there any tests or programs I can use to confirm?

 

[barythrin]

You could take a ram chip from a further down socket and put it in the first ram socket and see if it counts further before it fails, or piggy back one if you have spares.

 

[Drea]

Hi

I recently bought an 8096, so it is the same as yours....it displays 31743 bytes free at boot. From what i could find on the internet, that is the correct amount, so that's what you should aim for.

I can tell you if everything is correctly hooked up, if you take some more pictures of the inside.

 

[Drea]

Here's a few pictures from my system, that shows where the cables go, sorry about the quality, it's from my ipad

 

[dave_m]

Turns out it has what I figure is the 64K memory board. Fab number 8032109, art 8032108. Yet it's showing 83 bytes free on the boot screen.

The power up memory test starts at location 1024, so the first bad location The PET found was around 1107.
Start poking and peeking test patterns there and move up until you find a miscompare. The incorrect data will give us the information on which bit is bad (stuck low or stuck high). That will help determine the bad chip to replace.


poke 1107,0

print peek(1107)

poke 1107,255

print peek(1107)

If no miscompare, go to next memory address:

poke 1108, 0
print peek 1108

poke 1108,255
print peek(1108 )

 

[Eudimorphodon]

So, it could just be a near coincidence, but when I had a stuck address line on my 4032 that caused a page of memory to be repeated the resulting memory size was off by *85* bytes from the 4k boundary where the broken line was. The memory test would quit 85 bytes higher because in reality it was running back into the zero page again and would hit some memory locations which changed as the program ran, thus stopping the test. I sort of wonder if what you're really seeing is the result of an addressing problem that's likewise causing the a page of memory to be repeated.

An "easy" way to see if this is likely is poke over the "end of memory". (1107, like Dave_M said.) Maybe do a very short loop that'll just poke its way upward:

10 FORX=1107TO32767OKEX,255:NEXTX

(typed without spaces to save memory.)

If it is a stuck page the machine will crash hard in pretty short order.

It probably *isn't* this but it might be worth ruling out.