garvey@cmic.UUCP (Joe Garvey) (11/15/89)
I was told by a vendor that the Centronix interface in a typical (Laserjet II)
interface would only transfer data at about 50 Kbaud (yes I know it's byte
wide, but it's easier to compare to a serial interface this way).
Is this true, if so is it really worth having a centronix interface in a
printer?
--
Joe Garvey UUCP: {apple,backbone}!versatc!mips!cmic!garvey
California Microwave Internet: mips.com!cmic!garvey
990 Almanor Ave HP Desk: xxx ("mips!cmic!garvey")/hp1900/ux
Sunnyvale, Ca, 94086 800-831-3104 (outside CA)
408-720-6439 (let it ring) 800-824-7814 (inside CA)
We recently appeared in the maps. If your site is up to date, we're there.rjn@hpfcso.HP.COM (Bob Niland) (11/16/89)
re: "I was told by a vendor that the Centronix interface in a typical
(Laserjet II) interface would only transfer data at about 50 Kbaud"
> Is this true, if so is it really worth having a centronix interface in a
> printer?
Yes, if only because the serial interface can only hit about 19.2Kbaud.
Incidentally, both interfaces are always present.
As I understand it, the LJ's I/O performance is limited not by the
interface, but by the architecture. The LJ was designed back when RAM was
expensive, so it does not have a full page-image buffer. Each laser scan
line is fabricated in real-time from a tokenized (partially pre-processed)
representation of the input data stream. The processor is an 8MHz MC68000.
This design is also why Error 21 sometimes occurs.
The print engine will hold-off the interface when the parsing and tokenizing
can't keep up with the inbound data. This is what limits I/O rates.
Regards, Hewlett-Packard
Bob Niland rjn%hpfcrjn@hplabs.HP.COM 3404 East Harmony Road
UUCP: [hplabs|hpfcse]!hpfcla!rjn Ft Collins CO 80525-9599
This response does not represent the official position of, or statement by,
the Hewlett-Packard Company. The above data is provided for informational
purposes only. It is supplied without warranty of any kind.chrise@hpnmdla.HP.COM (Chris Eich) (11/16/89)
Our HP LaserJet 2000 with Centronics interface can accept data at up to 28kbytes/sec (x10baud/byte = 280kbaud) when we're printing full-page bitmaps. One data point.... Chris Eich chrise@hpnmd.hp.com
dold@mitisft.Convergent.COM (Clarence Dold) (11/17/89)
in article <218@cmic.UUCP>, garvey@cmic.UUCP (Joe Garvey) says: > > I was told by a vendor that the Centronix interface in a typical (Laserjet II) > interface would only transfer data at about 50 Kbaud (yes I know it's byte > wide, but it's easier to compare to a serial interface this way). We recently drove a parallel interface printer at a full 30PPM, 80 columns, 60 Lines Per Page. That works out to (excuse me while I !bc) 24Kbaud. If you can find a printer faster than 30PPM, maybe it matters. We also found that the 37PPM printer didn't like our UNIX a whole lot. We would blast data to the printer at 50KBytes /second (yes it is Byte wide), then go away for a while, then blast more data... The data link is theoretically capable of 6uSec cycles, or 166KBytes, but the priorities aren't keen on high printer load. A serial link at 19.2K turned out to be much better. -- --- Clarence A Dold - dold@tsmiti.Convergent.COM (408) 434-5293 ...pyramid!ctnews!tsmiti!dold P.O.Box 6685, San Jose, CA 95150-6685 MS#10-007
pipkins@qmsseq.imagen.com (Jeff Pipkins) (11/17/89)
In article <218@cmic.UUCP> garvey@cmic.UUCP (Joe Garvey) writes: >I was told by a vendor that the Centronix interface in a typical (Laserjet II) >interface would only transfer data at about 50 Kbaud (yes I know it's byte >wide, but it's easier to compare to a serial interface this way). Using a typical PC (what does this mean anymore?) running DOS, the Centronics interface is software bound. You can expect to get a maximum of about 1K bytes/sec (equivalent to a 9600 baud serial connection). A program that drives the port directly (*NOT* reccommended) can reach speeds that violate the Centronics timing specs. I wrote a replacement DOS printer driver to speed things up. If you use "copy /b file lpt1" (/b is important), it will drive just about any printer at close to its rated transfer speeds. We've seen bursts of 30K bytes/sec to a PostScript printer and speeds in excees of 90K to a dummy card with a scope attached to it, when using a modest 286 machine. If there is enough interest, I'll ask the boss if I can release the driver. P.S. The /b option of the copy command is equivalent to using MS-DOS function 44h to set the "raw" (uncooked) mode on a character device. Also: You can probably reach speeds of 10K bytes/sec by using the BIOS int 17h function call to print instead of going through DOS.
neil@cpd.com (Neil Gorsuch) (11/21/89)
In article <8390002@hpfcso.HP.COM> rjn@hpfcso.HP.COM (Bob Niland) writes: >re: "I was told by a vendor that the Centronix interface in a typical > (Laserjet II) interface would only transfer data at about 50 Kbaud" >> Is this true, if so is it really worth having a centronix interface in a >> printer? >Yes, if only because the serial interface can only hit about 19.2Kbaud. >Incidentally, both interfaces are always present. Not in most workstations, especialy desktop workstations (until now, that is, with our new product 8-). And ours can do over 40 Kbytes per second throughput in our low performance model 8-). -- Neil Gorsuch INTERNET: neil@cpd.com UUCP: uunet!zardoz!neil MAIL: 1209 E. Warner, Santa Ana, CA, USA, 92705 PHONE: +1 714 546 1100 Uninet, a division of Custom Product Design, Inc. FAX: +1 714 546 3726 AKA: root, security-request, uuasc-request, postmaster, usenet, news
mrm@sceard.Sceard.COM (M.R.Murphy) (11/22/89)
In article <48@qmsseq.imagen.com> pipkins@qmsseq.UUCP (Jeff Pipkins) writes: >In article <218@cmic.UUCP> garvey@cmic.UUCP (Joe Garvey) writes: >>I was told by a vendor that the Centronix interface in a typical (Laserjet II) >>interface would only transfer data at about 50 Kbaud (yes I know it's byte >>wide, but it's easier to compare to a serial interface this way). > >Using a typical PC (what does this mean anymore?) running DOS, the >Centronics interface is software bound. You can expect to get a maximum >of about 1K bytes/sec (equivalent to a 9600 baud serial connection). >A program that drives the port directly (*NOT* reccommended) can reach >speeds that violate the Centronics timing specs. [description of DOS driver replacement deleted...] Seems that I remember that the original Centronics interface spec was 4us/byte (250Kbytes/sec). For 132 characters, then the paper had to move, and 60lpm was real good. 60x132 characters/min = 132characters/sec. Not too big a load and not likely software bound :-). Tempting for the printer manufacturer to use one little Z80 chip instead of a whole big board full of disrete logic to do the job. And while they're at it, might as well do graphics and better fonts than the 101A, and wouldn't laser engines be nice, and how about full page graphics and page description languages, and the thread runs on like this sentence. The Z80 became bogged, 2.5Kbytes/second typical rate for accepting data, but that's fast compared to the printers (up to 300lpm or so) that were using it. Use more Z80's in one printer, maybe better use one bigger CPU. The 68000 can get bogged, too. Especially when called upon to manage the print engine, the font generation, the page description language, the host interface, the protocol, the coffepot, and whatever else is in printers these days. The work expands to bog whatever the CPU is. For vector graphics, 1000 vectors/second converted to 300x300dpi bitimage isn't too atypical now. In fact, maybe a bit fast. If a vector takes about 5 bytes to describe (13 for QUIC(tm), less for 4014(tm), less still for others), then 5000 bytes/second is what the printer will accept when doing vector graphics. Font management and generation can be a real sponge of CPU power. The process gets even more complicated when something like POSTSCRIPT(tm) can take 2MB of data to generate one lousy dot on the page (if you're clever :-) or a dozen or so lines can take a day or so to produce one page. I like bytes/hour as a unit of data transfer. -- Mike Murphy Sceard Systems, Inc. 544 South Pacific St. San Marcos, CA 92069 mrm@Sceard.COM {hp-sdd,nosc,ucsd,uunet}!sceard!mrm +1 619 471 0655