home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.modems
- Path: sparky!uunet!rde!ksmith!keith
- From: keith@ksmith.uucp (Keith Smith)
- Subject: Serial Interfaces
- Organization: Keith's Computer, Hope Mills, NC
- Date: Fri, 24 Jul 92 03:53:42 GMT
- Message-ID: <1992Jul24.035342.20200@ksmith.uucp>
- In-Reply-To: <1992Jul23.011759.14499@ksmith.uucp>
- Lines: 42
-
- In A Mail message to me Dave Dillon (craywr.cray.com!ddillon) wrote:
- ->In article <1992Jul23.011759.14499@ksmith.uucp> you write:
- -> [....]
- ->>instructions, You get about 41 clock cycles worth of CPU instructions
- ->>per byte you desire to send:
- ->>
- ->>1.8432Mhz / 16 = 115K
- ->>
- ->>therefore 4.77 / 1.8432 * 16 = 41.4 cpu cycles / byte at 115K baud
- ^^^^
- per BIT, not byte
-
- -> [....]
- ->
- ->I believe that bits and bytes have gotten confused here. "115K baud"
- ->is, more fussily, 115200 bits per second -- not bytes per second.
- ->
- ->Assuming that a byte is serialized as 8 data bits, 1 start bit and
- ->1 stop bit, the byte rate (and unbuffered interrupt rate) would be
- ->one-tenth of that you've calculated, or 11520 bytes per second.
- ->
- ->The overworked 8088 @ 4.77 MHz then has 414+ CPU cycles / byte.
- ->
- ->Alas, this error propagates through subsequent calculations, and
- ->detracts from an otherwise-interesting article. But I certainly agree
- ->with your premise, if not the numbers!
- ->
-
- Shit, I apologize, As I was doing this I was wondering how I had ever
- pushed 38.4K thru A 2 Mhz Z-80 .... :-b
-
- It's posted, and I erred by a factor of 10 on the clock cycles. This is
- STILL a killer burden on a CPU because even at * 10 You don't have time
- to get anything ELSE done between interrupts.
-
- Also I shouldn't post after a bottle of Cabernet ...
-
- Sorry if I confused anyone.
- --
- Keith Smith uunet!ksmith!keith 5719 Archer Rd.
- Digital Designs BBS 1-919-423-4216 Hope Mills, NC 28348-2201
- Somewhere in the Styx of North Carolina ...
-