[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Extended 80 column card issue
On 05/23/2011 08:54 AM, Jeff Shanholtz wrote:
> On first thought it would seem to me that a ram chip that isn't used
> by the 80 column feature is faulty,
No such thing, they're all used (one chip holds bit 0 of every address,
one holds bit 1, etc.). But of course it's possible that inside one of
the chips, one or more bits outside the 1 kilobyte area $400 to $7FF
that's used for the 80-column text screen are bad.
You can try a larger area by booting some game that uses double hi-res.
Pirates!, Prince of Persia, Legend of Blacksilver, Zork Zero come to my
mind, but there are probably lots of others. If that works, you'll know
that (depending on what screen a game uses) $2000-$3fff or $4000-$5fff
are working.
And what about the build-in self test (hold CloseApple during power up),
does that pass?
Then there is that "Call-A.P.P.L.E. test disk" that when booted on a //e
should give you an option to test aux memory. You can try that. The ROM
test on that disk is expected to fail on an enhanced //e by the way,
since the disk came out before the enhanced ROMs and doesn't know about
them. And DON'T use the drive alignment test on that disk to actually
modify drive alignment, it does a lot more harm than good.
> but on second thought I wouldn't
> expect that to cause a hang
It can certainly cause a hang if the computer tries to run code from aux
memory, and some opcode is changed by the bad bit(s) into a jump to
nowhere or something that causes an endless loop, for example, or into a
CPU-crashing code on an unenhanced //e (I don't remember whether you
said you got an enhanced machine or not, the CPU in an enhanced //e
can't be crashed as such).
> so it seems more likely one of the logic
> chips is bad. Opinions?
A bad logic chip would be more likely to cause complete failure of the
card, IMHO. All data bits from and to the card go through these buffers,
one of them for video data (such as characters being displayed on an 80
column screen), the other for CPU data (such as characters being written
to an 80 column screen); so you wouldn't see a working 80 column screen
if one of them had failures.
--
Linards Ticmanis