Re: I killed it.. :-(

[Doppler Dog <doppler_dog@my-deja.knickers.com>]

Linards Ticmanis <ticmanis@coli.uni-sb.de> writes:

> On Mon, 30 Jul 2001, Mark Cummings wrote:
> 
> > 
> > You could try another processor, even a 65C02 just to see if it's that. By
> > the description of the garbage on the screen, I would be more inclined to
> > suspect the RAM. If it's socketed like the II+ you could try swapping a bank
> > of 8 around and see if the screen garbage clears up or changes. 
> 
> Thanks. Unfortunately that does not seem to be it. Since the computer
> currently behaves exactly the same no matter if there is a CPU in it or
> not, I still assume that the CPU is the problem. 

There's faulty logic here (no pun intended).  If RAM has gone bad,
then the CPU cannot do it's job properly, and will not be able to
clear the screen area RAM and put up a bootup message.  Pulling
out the CPU will exhibit the same symptoms: gargbage on the screen.

You can't conclude that the CPU is the problem given the information
you have.


> By "flashing" characters
> I meant the kind of things that you can get by typing FLASH from BASIC,
> not characters that hop around randomly, so I doubt the RAM is really the
> problem. 

That does not follow, see above.

> I exchanged the lower bank with one of the others anyway, apart
> from the fact that now closing square brackets occur more often among the
> garbage nothing has changed. 

It's very likely that more than one RAM chip has gone.  If you spiked
an address or data bus line, then pretty much all RAM chips connected
to the bus have been exposed to risk.  RAM is usually much more
sensitive than a CPU.  In most broken machines I've seen it's usually
the RAM that's gone bad, not the CPU.  And the symptoms are exactly as
you describe.  Alternatively, it could be some control logic, but as
this is usually TTL logic, it also tends to be less sensitive than
CMOS RAM.

In fact, if you see a difference when swapping RAM around, then it's
most likely NOT the CPU, because if it was, then swapping RAM around
wouldn't make any difference (i.e. not even seeing the closing square
bracket more often).  This is not 100% correct, though, because on
power up RAM contents are undefined, even if usually consistent.

On the other hand, it could still be the CPU, but I'm willing to bet
otherwise.  All I'm saying is that you can't conclude it's the CPU at
fault given the symptoms.

If you're brave, switch the board on, let it sit for a while, and put
your finger on the ceramic (or plastic) part of the CPU... avoid
touching the contacts.  If it's very hot, then you can safely say it's
bad.  If it's not, it's probably ok.  You could try doing the same on
some of the other chips.  When chips go bad they usually burn through
to a closed circuit, dissipating a lot of heat.  But that's not always
true.

But if you find that you are electrocuted by the dangerous 5v on the
power line, or if you suffer 3rd degree burns from the experience
leading to a horrific infection and consequent amputation at the
shoulder, don't sue me, I'm clinically insane for all you know.  I
won't assume any responsibility.

> 
> I have already run ads looking for broken VIC-20s and CBM 1541 drives in
> our local free adspapers, those have socketed 6502s and also the 6522s I
> need for the mockingboard, unfortunatly I had no success so far. Ordering
> them via mail order seems to cost about $13 per chip which I think a bit
> excessive.. oh well. 
> 
> Thanks anyway,
> 
> Linards Ticmanis
> 

-- 
Remove the knickers from my email to reply.