home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!psgrain!qiclab!leonard
- From: leonard@qiclab.scn.rain.com (Leonard Erickson)
- Newsgroups: comp.dcom.modems
- Subject: Re: Serial and Parallel interface ??????
- Message-ID: <1992Jul21.210753.8014@qiclab.scn.rain.com>
- Date: 21 Jul 92 21:07:53 GMT
- Article-I.D.: qiclab.1992Jul21.210753.8014
- References: <mac1.711295729@Ra.MsState.Edu> <5785.2a6709d1@hayes.com> <BrKAoH.4Gp@news.cso.uiuc.edu> <1992Jul20.190425.22188@cc.ic.ac.uk>
- Reply-To: 70465.203@compuserve.com
- Organization: SCN Research/Qic Laboratories of Tigard, Oregon.
- Lines: 48
-
- vulture@carrion.cc.ic.ac.uk (Thomas Sippel - Dau) writes:
-
- >Nothing, really, apart from the mental block of manufacturers which associates
- >modems with serial lines. It would also be a convenient way to get shot
- >of the { tty | clist | namiaf | ... } processing that many computer
- >manufactureres assiciate with a serial line, and which is mostly unwanted
- >these days (i.e. character echo, processing of bs, del, end of anything
- >conditions, ... )
-
- >Any PC I know of puffs out at 115200 bit/sec (or 11520 byte/sec) on the
- >serial ports, mostly because of this unwanted additional processing. On a
- >common notebook parallel port I have seen 60 kbyte/sec sustained over several
- >minutes with a parallel port SCSI or Ethernet adapter. With the usual cable
- >used for parallel ports speeds of between 500 kbyte and 1 Mbyte/second
- >could easily be achieved - the cabling is electrically less demanding
- >than that for a SCSI bus - only two devices, no spurs.
-
- Bzzt! Wrong!
-
- The 115,200 bps limit is a *hardware* limit. At that rate, the UART
- chip is dividing the supplied clock signal by *one* to get the bit
- timing. It can't go any faster than that *period*. The chips aren't designed
- for a much higher bit rate than that anyway. If they, were, you'd need a new
- oscilator chip, *and* you'd have to throw out all your communications
- software (most of which sets the bit rate by *directly* loading ther
- UART's divisor registers.
-
- The other device are using different clock signals *and* substantially
- different chips. Compare the cost of a serial card with the cost of
- an ethernet card...
-
- >The second hurdle is that those who have so far brought out devices for
- >parallel ports do not document what lines and protocols they are using.
- >For example, on the (Epson-) Centronixs port the printer has the signal
- >/paper-out, the computer the signal /auto-feed-xt, inducing the printer
- >to supply a line feed for every carriage return. It would be *nice* to
- >know if the other devices use these and what they use them for.
-
- Funny, every card *I* own that has a parellel port has all the signals
- described. They also describe how to set them for biderectional use
- of the data lines. And they warn that such use is *not* standard.
- (More specificly, that while their port will do it, many *won't*)
-
- --
- Leonard Erickson leonard@qiclab.scn.rain.com
- CIS: [70465,203] 70465.203@compuserve.com
- FIDO: 1:105/56 Leonard.Erickson@f56.n105.z1.fidonet.org
- (The CIS address is checked daily. The others infrequently)
-