Re: MOS 8565 VIC-II

From: Laze Ristoski (
Date: 2003-04-29 22:13:46

> AFAIK there are two major Versions of the VIC-II.
> The old one has no vertical lines at all.

Well, I can notice them on my 6569 R5, although they
are almost invisible.

> It has a
> quite unreadable red color on blue background.

If you mean dark red (2), and dark blue (6), you
are right. But it's even worse on the 8565. New
VIC doesn't give that sharp picture. The only thing
I find better on 8565 is light blue text on dark blue
background. (like when you switch the C64 on).
Old VIC makes the text look a bit weird. One
letter is mixed with red, the next one with green.
While 8565 gives sharp lt. blue color.

> According to their periodicity, I guess it has something to do with
> the phi/2-clock. Maybe some interaction within parts of the chip, since it
> manages
> the clock itself?!?

Do you think there's something else apart of the VIC that produces
these lines? (eg. the RF modulator, etc...)

> The grey dot's always appear on a new VIC if you access a VIC-register.
> It even works if you write f.e. to a register that does not change
> anything
> on the screen, you'll get grey-dot-snow.

That's not the way it seems to me. Let's say that you write 0 at $D020,
in an infinite loop. The dots (or sparkles as Marko calls them) will only
appear at the border. If the VIC plots the paper area during the write
access, you won't see any dots.

> There are at least three major versions of the 6567.  The 6567R56A
> uses 64 bus clock cycles per line and draws 262 lines per frame.  The
> 6567R8 and later consume 65 cycles per line and produce 263 lines.
> Then there's the HMOS version of the NTSC VIC-II.

Seems I have left out the revision numbers when I said there
were 4 types. :)

> If I remember correctly, none of the chips listed in the above document
> suffer from the "sparkles" or produce strong vertical bars on the
> screen.

Right. 656x based VIC does not produce sparkles, nor do they
produce lines. Well, they do, but they are almost invisible (although
many people claim they don't produce lines at all).

> I don't have any C64 with a 85xx video chip, but I know the
> phenomenon from the C128.

Aren't there any sparkle-free VICs for C128?

> Well, it looks like the register is shortly disconnected from the rest
> of the chip.  During that period, the rest of the chip sees a $ff value.

Strange thing is that they occasionally disappear. And the lines are
(as I already mentioned) sometimes more, sometimes less visible.

> On the TED (the video chip of the C16, C116 and plus/4), the sparkles
> are light green (also $ff).

I don't have any of these commies, but seems 8565 is not the only chip
with this kinda problems.
Ok, I do have a C116 (with burned RAMs), and a plus/4 (with no
keyboard, and whoknows whats wrong with it).

Btw, except for changing the name (from MOS to CSG), did Commodore
change something in the manufacturing process?
To my eyes, MOS 8565 produces a _bit_ better picture than the CSG 8565.
It might be just my brain, but...
I'm talking about the same revision (8565 R2). I've never seen a plain

I have just tried to put an old 6569 VIC on a new board. The result
was no picture at all. I don't dare to do the other way (new VIC on
old board). You think I can fry it?


       Message was sent through the cbm-hackers mailing list

Archive generated by hypermail pre-2.1.8.