Re: PET 2001 fix Part 3 - RAM/ROM board etc.

You're right that it would probably be random, but it looks like it may be
power related since it clears up as the voltage drops, so maybe there's a
recurring spike that affects a weak bit as certain addresses or states come
up, whatever. Wouldn't be the first time that something worked that
shouldn't have and I thought it'd be a really easy way to rule out the RAM.

My first approach to a video problem is usually to pull the CG and look for
a clean blank screen, and then replace it, pull the video RAMs and look for
a clean checkerboard; often that points me in the right direction.

Looks like the problem is mostly in the fifth character column; wonder if
that's where the horizontal line was originally? That'd tend to point the 
finger at the '93 counter (A1) or the '165 SR (B2).

m

----- Original Message ----- 
From: "Ethan Dicks" <ethan.dicks@gmail.com>
To: <cbm-hackers@musoftware.de>
Sent: Thursday, May 12, 2011 2:50 PM
Subject: Re: PET 2001 fix Part 3 - RAM/ROM board etc.


> On Thu, May 12, 2011 at 2:35 PM, MikeS <dm561@torfree.net> wrote:
>> AFAIK the old 2001 had the video RAM connected directly to the CG ROM
>> (unlike the newer ones which had a latch in between), so if one or more
>> bits
>> of the video RAM were unstable (or you had an intermittent socket) which
>> caused the data to change part way through a character line, wouldn't you
>> then get corrupted characters as the CG switched back and forth?
>
> You could, but in order for that to be stable, it would need to be
> synchronized with the pixel clock.  If the symptom was random
> pixels/no visible character shapes, I'd be willing to suspect the
> SRAMs.
>
>> I just thought it'd be an easy and worthwhile thing to try, especially
>> since
>> someone on the VCF PET forum just fixed a 'flickering snow' problem by
>> replacing one of the video RAMs.
>
> I can see how snow could be an SRAM problem because it's not tied to
> the same pixels changing in the same characters in the same way frame
> after frame - that is one way I'd expect random addressing of the CG
> ROM to manifest itself.
>
> -ethan
>
>       Message was sent through the cbm-hackers mailing list


       Message was sent through the cbm-hackers mailing list