Re: IIGS 115,200 Baud Serial Port - How?

[Hugh Hood <hughhood@earthlink.net>]

in article 
fdd60a38-56d7-4a06-aecd-712ff2b7121d@b38g2000prf.googlegroups.com, David
Wilson at mcs6502@gmail.com wrote on 12/30/08 7:41 PM:

> On Dec 31, 12:17�pm, Hugh Hood <hughh...@earthlink.net> wrote:
>> After taking your advice to heart, it seems the SCC WR11 _does_ appear to be
>> the key to this 115,200 baud stuff (dare I say 230,400 with a 64x clock
>> factor) on the IIGS's SCC built-in serial ports.
> 
> If you look at the description of WR4 you will see that 64x clock
> means that the clock is 64x the data rate so this would be half the
> speed of 32x (not twice).
> 
> With a 3.6864 MHz crystal and the BRG disabled you have the following
> 4 speed options:
> 
> x1  3.6864 Mb/s (not usable for NRZ async comms)
> x16 230,4 kb/s
> x32 115.2 kb/s
> x64 57.6 kb/s

David,

You've made an excellent point. You're right ... I got it wrong. A higher
denominator (clock mode multiplier) will _decrease_ the baud rate.

Extending on the point you made in your prior post, it does seem that _both_
"2's" are removed from the Baud Rate formula when WR11 is set to _not_ use
the SCC Baud Rate Generator, thus rendering the Baud formula as:


          3.6864 MHz [Clock Frequency]
Baud =    ----------------------------
                Clock Mode

Thus, just as you stated:

16x clock:   3.6864 MHz / 16 = 230,400
32x clock:   3.6864 MHz / 32 = 115,200
64x clock:   3.6864 MHz / 64 =  57,600

In the meantime, I've located the code in ProTERM where WR11 is set. ProTERM
enables the BRG (no surprise). So, let me disable it and give it a go.

Thanks for the fine analysis and input.

I suspect Greg Schaefer, when coding ProTERM, really only envisioned dial up
modem transfers (as opposed to high speed null modem stuff), so that speeds
> 57,600 weren't really considered.

Or, as maybe I'll find out, even ProTERM has trouble at that rate, although
I believe at one point Greg modified the program to use the Turbo ASB Super
Serial Card daughterboard at higher speeds. We'll see.

BTW, has anyone modified ProTERM to work with the stock (with a 65C51 for
proper RTS/CTS handshaking) Super Serial Card using the 115,200 mode? I'm
not sure a 1 MHz //e could keep up with a 2-way data transfer, but one with
an 8 MHz Zip Chip sure might.

Hugh Hood...