Re: Why do old IIc and IIe have poor comm skills?

[david@wraith.cc.uow.edu.au (David E A Wilson)]

Robert Jensen <bob_jensen@jhuapl.edu> writes:
>Why is it that a II+ can run Kermit at 2400 bps (or more) and yet
>the old IIe and IIc computers have trouble and miss characters?
>What did the II+ have that the next generation lost?

Only half as many characters on the screen to scroll.

If you mean in 80 column mode then many 80 column cards
for the ][+ used the 6845 CRTC chip and could be programmed
to scroll the screen with about 4 writes rather than needing
to read and write 1920 bytes.

>A while back, Apple would upgrade the motherboard of a IIc if
>the owner had a demonstrated need, like having a 2400 bps modem.
>What is the difference between the old and new boards that solves
>the problem? I witnessed before and after performance and it was
>significantly different.

The original //c motherboard divided the master clock (14.31818MHz) down
to get the baud rate clock for the ACIA chips. Unfortunately it was only
approximately the required frequency (about 3% out?). This did not bother
eary modems but was not good enough for 2400 and up modems. The revised
motherboard added the correct baud rate generator oscillator.

>I recently slipped a IIe into the place long occupied by a II+
>and am sad that something (Kermit) performs worse than before.

The old ][e did not enable interrupts while scrolling so would loose
characters. The enhanced //e fixed this - I ran a simple terminal program
I wrote myself with a 256 byte input buffer and it only ran into problems
above 4800bps (a screen refresh at 9600 would corrupt the last two lines)
becuase I did not implement hardware flow control (when I wrote it I had
a 2400 bps modem).
-- 
David Wilson	Dept CompSci Uni Wollongong Australia	david@cs.uow.edu.au