Re: CBM source code

From: Andre Fachat (
Date: 1998-01-12 13:51:55 wrote:

> Hallo Andre,
> I haven't been able to get your board and papers but they are here in Heerlen. 
> You mentioned not being able to format a diskette: you have to use DS DD 96TPI, 
> no HD-diskettes! DS DD 48TPI will work as well but unreliable.

Yes, but finding the disks is difficult. Anyway, I could read 8050 formatted
disks, so the SFD1001 works. Still didn't have time for the 8296D.
Thanks for the stuff again!
> About the 6522: Shifting PHI2 means a later negative edge. Hasn't it a negative 
> effect on the behaviour of the 6522 in any way?

You should only shift the rising edge. I think there is a schematic on
funet...  (marko? I couldn't find any on a quick browse...) You can AND
the original and the delayed Phi2 to get the right signal, I think.

> Using a 16550: I only have experience with the 8250 but regarding the problems 
> I had with the 6551, I only can support you advise using an Intel device in 
> this case. You only need an extra 74LS139 to generate the RD- and WR-signals.

I'm not sure if I mentioned it already, but if you use a RS flipflop that
is set when RxD goes low and reset when ACIA register 0 is read, and use the
output as rts, you should be able to get higher baud rates too.
I haven't tried it so far. (see at
But, a new byte can then only
be received when the CPU has taken the last one, which - in a decent OS -
involves some context switching (interrupt latency etc) so the overall
performance of the 16550 should be much better still.


Email address may be invalid. Use "fachat AT physik DOT tu-chemnitz DOT de"
------Fight SPAM - join CAUCE, spammers...
Andre Fachat, Institute of physics, Technische Universitšt Chemnitz, FRG

Archive generated by hypermail 2.1.1.