Re: [stella] Hardware Problems

Subject: Re: [stella] Hardware Problems
From: "Mark Graybill" <saundby@xxxxxxxxxxxx>
Date: Thu, 5 Dec 2002 00:08:22 -0800
My experience is that the console has more problems with stray capacitance
and inductance than in wire lengths themselves. Of course, longer wire
lengths increase the probabilities of such problems. But I've successfully
used extensions of about 8-10 inches of ribbon cable to reach ROM emulators
in the past. I've even run cables to ROMs on breadboards.

PCB extender boards are a _lot_ cleaner, however, in my experience.

For something like a flash cartridge my inclination would be to use a PCB
from a cart that has a ROM of the right configuration in it, and add an
inverter, as is typically done for a homebrew cart or a "burn and churn"
EPROM cart. To program the flash, my inclination would be to make an adapter
to fit the cartridge right to the programmer without removing the chip, or
to build a custom programming adapter for an I/O port. Another approach I'd
consider is making a new PCB to use small outline package devices for the
flash and the inverter.

An idea I've been toying with myself is that if I get some time away from
the other work I'm late in doing with the AVR microcontrollers is to make a
2600 cart based on one of these. They're flash-based, generally far cheaper
than ROMs and could be programmed to perform all sorts of memory tricks
totally transparently to the 6507. They could feed any program to the 2600
in a completely linear fashion, looping back through the memory map as
necessary, or they could emulate any of the bankswitch modes. For new games,
they could run off their own clock oscillator at a substantial multiple of
the 2600's speed and simply feed the 2600 drawing instructions during screen
drawing times while having time on their own to perform the logic functions
to determine what to draw.

An AVR cart for the 7800 provides even more extensive possibilities...

But first I've got "real" work to do... *grumble*

-Mark G.
----- Original Message -----
From: Mark De Smet <de-smet@xxxxxxxxxxxxxxx>
To: <stella@xxxxxxxxxxx>
Sent: Wednesday, December 04, 2002 7:21 PM
Subject: Re: [stella] Hardware Problems

> I think the most likely cause is the wire length.  When I was making my
> "anygame cart", I found that I could not use a long ribbon cable, but did
> not have a problem when I kept all the wiring(cart connector to all
> circuitry) less than 6in.  When I made my portable atari, I started with a
> 6 switch PCB, and used a ribbon cable to re-locate the cartridge
> connector.  This cable is 3 inches long. with no connectors between.  My
> portable has some problems with a few games, pitfall! being the most
> likely to fail.  I have seen similar problems using a "rom scanner".
> For development, I would recommend just using a "simpler" game.  If your
> final product is enclosed in a cartridge, this shouldn't be a problem.
> I did not spend the time to look at the signals on a scope to see was the
> signal integrity is, but I'm sure that it is not good.
> Mark
> On Wed, 4 Dec 2002, Chris Larkin wrote:
> > As the work on my flash cart continues on I'm still hitting some major
bumps... During a routine test this morning, I programed the cart with
Pitfall, and placed it in the VCS... when I fired it up.... nothing....  So
I checked over everything, every little last detail, still nothing....
> >
> > So I manually manipulated the Address Lines and read the data out of the
EEPROM with a multimeter on the cartridge connector pins, and it coinsided
perfectly with the data in the ROM image...  I even checked the FFD start
address, and it is correct in the EEPROM...
> >
> > This was working fine before, and as usual now it is Caput... so I seek
the help of you all... I removed everything from the breadboard except the
EEPROM and the hex inverter... the only thing I think this can be is the
length of wire coming from the PCB to the EEPROM, it's about 6 inches a
wire... do and of you think this could be the problem???
> >
> > Or the EEPROM is corrupt some how, I cannot check all 4096 addresses by
hand, so I checked about 35 different area's of the ROM, and they all came
out correct.
> >
> > --Chris
> >
> --------------------------------------------------------------------------
> Archives (includes files) at
> Unsub & more at

Archives (includes files) at
Unsub & more at

Current Thread