| |
K0CQ > IBM 06.05.04 20:31l 31 Lines 1144 Bytes #-8220 (0) @ WW
BID : 16182_W0AK
Read: GUEST DC9BM
Subj: Re: Serial I/O from LPT1 ??
Path: DB0FHN<DB0FOR<DB0MRW<OK0PKL<OK0PCC<OK0NAG<OK0PPL<DB0RES<DK0WUE<HA3PG<
7M3TJZ<SP7MGD<ZL2TZE<LU6DTS<VK3AVE<N8DA<WH6IO<KJ6EO<KQ0I<W0AK
Sent: 040506/1507Z @:W0AK.#CIA.IA.USA.NOAM #:16182 [Des Moines] $:16182_W0AK
From: K0CQ@W0AK.#CIA.IA.USA.NOAM
To : IBM@WW
It could be done but its messy, VERY messy. There's a lot of timing to be done
that demands either using a hardware timer or counting CPU clock cycles which
then depends on the CPU variety and speed. The PC variety UART does all that
timing, clock dividing, and sampling work for you, just one parallel hit per
byte in or out.
Fifteen or so years ago there was a book on interfacing to the microprocessor
that probably had the code for what you want to do. Might have shown up in an
early Byte or Kilobaud magazine also. It was done occasionally but hardware
and software designer's alike praised the UART for its simplicity of use and
superior performance.
Trying to catch incoming characters through the UART in Gee Willikers (I wish
it was a Basic) Basic was difficult, catching bits will be extremely
frustrating, you won't catch many.
The AY3-1015 variety of UART will interface to the parallel port better than
the PC 8250 variety of UART that is designed to interface to the 8088 and
later memory bus.
73, Jerry, K0CQ @ W0AK.#CIA.IA.USA
Read previous mail | Read next mail
| |
