Re: [stella] 2600 on-a-Chip new site

Subject: Re: [stella] 2600 on-a-Chip new site
From: "Eckhard Stolberg" <Eckhard_Stolberg@xxxxxx>
Date: Thu, 22 Aug 2002 15:02:41 +0200
> Thanks!! I was very careful about designing a 100% compatible TIA
> I haven't tested Cosmic Ark yet. I'm writing a script that will allow a 4K
> bin to be inserted in a VHDL code to initialize the FPGA internal
> This will let me test more carts than I have available. I will post my
> results as soon as I finish something.

This is a great project. I really hope you can finish it.

Some other games you should try:

The first 18-sprites demo that John Seager posted to this list. This
version would show 18 sprites in a row on some VCSs, while only showing
17 sprites on others. Later demos were changed to work on all 2600s,
so this first version might be a good compatibility test.

Kool Aide Man has problems on some 2600 Jrs and on some 7800s. The
numbers for the score would get positioned on top of each other, so
that the collition detection would constandly be triggered. This
makes the game unplayable. The wierd positioning happens due to
the game messing around with the positioning registers too much
during a HMOVE.

Also you should ask Thomas Jentzsch for a copy of the first unpublished
version of Robot City. In this version he also accessed HMCLR too early
after a HMOVE, but he gets totally wierd position errors, probably
because he changes a NUSIZ register inbetween.

And if your TIA passes all of these tests and shows the same positioning
errors as the original TIA, it would be nice, if you could explain us
how exactly object positioning works on the TIA, especially during HMOVE.

Ciao, Eckhard Stolberg

Archives (includes files) at
Unsub & more at

Current Thread