I apologize for taking so long to post this summary. I just wanted
to exhaust every possibility before I submitted my findings to management.
My original question was:
========================
Now I have an even bigger problem with this laser postscript
Level 1 printer (TracJet III by Advanced Matrix Technology,
Inc.). I set it up for parallel printing on a SUN4/600 host
running SunOS 4.1.3 through the CoAlm card using its Centronics
port.
The printer was set up for "parallel interface" and "postscript
emulation." Each time when I sent a simple "cat file > /dev/ttyZ8"
command to talk to it, the feedback on the console window would be
"css: printer out of paper
css: no printer."
We checked the cable with our test equipment--both ends were sending
and receiving the right signals. But the "ready light" on the printer
never blinks a wink, like it has not received any signals at all.
Every hardware person here tells me it's a handshaking problem but
no one could come up with the correct suggestion.
I had hooked this printer up to an IBM PC and was successful to
print a postscript file through the parallel interface--that means the
parallel interface on the printer is good.
I also had hooked up a TI MicroLaser II parallel printer on the same
SUN node through the CoAlm Centronics port and had NO problem printing
on it--that means the SUN setup is good.
The only remaining conclusion is: The AMT printer, though parallel,
does not respond to the SUN's parallel signals.
=======================================================================
Response:
========
I had exactly the same problem with a cheap Apple
laserwriter. I ended up sending the Apple back and
getting a cheap (well, the cheapest) QMS.
Some printers are set up and tested with PC's only.
A SUN can sometimes overrun them. I'm not familiar with
the CoAlm, but the Magma sbus add-in board can be set
with different handshaking protocols on the parallel
port that might work with slower printers.
If you have no money, maybe you can print serially.
Not much help huh?
Steve Raffensberger
AT&T
[Thanks Steve!]
=======================================================================
I had also received info and specs from Magma which had been submitted
to management for evaluation approval. The following is taken from
portion of my report:
Printer: TracJet III laser formfeed printer by Advanced Matrix
Technology Inc., with 4MB of RAM (upgradable to 8MB)
Serial Interface testing:
We were able to use this printer as a serial postscript printer
hooked up to a SUN 4/600 running OS 4.1.3. The major problem we
encountered with this interface was the length of the serial cable.
We originally used a 50-ft serial cable that met the maximum AMT
cable specification. However, data dropped from the queue and
80 percent of our print jobs, which varied from 100,000 bytes
to 2,600,000 bytes, failed to print. When we substituted the
50-ft cable with a 10-ft one, the problem was solved.
We printed gray-scale grams (a little over 1MB per page) on this
printer using serial interface and postscript emulation. The
approximate time that took to print each page of a multiple-page
gram file was 21 minutes.
Ethernet Interface testing with Milan Fast Port:
We tested this printer as an Ethernet printer connecting a
Milan cable from the Milan Fast Port to the AMT parallel port.
A Milan filter program was used. The same gray-scale grams
used for the serial interface test took 10 minutes to print
(each page) on the AMT printer with 4MB of RAM. These files,
when printed on a Texas Instrument MicroLaser XL Turbo printer
with 6MB of RAM (not a formfeed laser printer) through the same
Milan Fast Port, took 2 minutes to print each page.
Parallel Interface testing:
We were unable to successfully interface this printer with a
SUN 4/600 through the CoAlm parallel port. This interface
works with the Texas Instrument MicroLaser printer, but does
require a filter program. This AMT printer may require a
different filter program to work with the CoAlm parallel port.
The following is a summary of the efforts we made to test the
printer with the parallel interface:
1. Four different parallel cables were tested with the same
negative result. All cables were tested with the breakout box
as known good parallel cables.
2. AMT printer parallel outlet was tested with the breakout
box-control line status matched with given Centronix parallel
printer interface specification.
3. The printer was tested on 5 different SUN 4/600s with
identical negative result.
4. Used the following UNIX command to send output directly to
the parallel port without going through any filter:
% cat file > /dev/ttyZ8
Received CoAlm driver error messages at the console:
css: printer out of paper
css: no printer
5. When the printer was connected to an IBM computer through the
parallel interface, postscript and ASCII files were successfully
printed through the postscript and HP laserjet emulation respectively.
6. Connected the AMT printer to a SUN 4/600 which currently uses
a TI MicroLaser parallel printer. Operating system, software, and
hardware remain the same except for the change in printer. The AMT
printer did not indicate any data coming through and print jobs
stayed in the queue. Once we reattached the TI printer to the
4/600 computer, print jobs in the queue started printing.
7. Tested the printer on a node with the upgraded CoAlm driver
v2.3.4 with same negative result.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:21 CDT