[comp.sys.amiga.hardware] Amiga and QMS PS-410

bryan@cs.utexas.edu (Bryan Bayerdorffer @ Wit's End) (02/16/91)

I've just connected a brand-new QMS PS-410 laser printer to my A2500/30 (rev 4.5
motherboard) via both the parallel port and one of the serial ports on a 2232
multiserial card.  While the serial connection works fine, I'm experiencing
problems when sending postscript files over the parallel connection.  The
problems only occur when sending data at high speed, e.g. when copying a file on
disk to the parallel port.  When sending postscript more slowly, e.g. sending
the postscript generated by ProDraw directly to PAR:, correct output is
produced.  This would suggest that there is some sort of synchronization
problem.  According to the pinout diagrams in the rather sparse QMS manual,
everything is wired correctly, but I'm wondering if there isn't something funny
about the Amiga's parallel port, or the printer's or both.  I've verified that
the infamous +5V on pin 14 is definitely not the problem, and it would appear
that at least DATA0 - DATA7 and STROBE are functioning correctly.  The QMS
manual contains the following mysterious quotes:


1. "When wiring the [parallel] interface, be sure to use a twisted-pair cable
for each signal and never fail to complete connection on the return side."

2. "Data transfer must be carried out by ignoring ACKNLG or BUSY signal."


The Amiga parallel port is a DB-25, and the QMS is 36-pin Centronics.  I'm using
a standard PC-clone-to-Centronics cable.  While quote 1 surely has to do with
noise supression, I'm puzzled by quote 2.  Is this telling me that there isn't
any flow control (i.e. assertion of BUSY) on the printer's parallel port?  That
would be really weird.

The QMS tech support people tell me that when the PS-410 is connected to a PC
clone, the parallel port needs to be configured for "infinite retry."
Apparently there is some sort of timeout on the PC side, but the QMS guy
couldn't tell me how the flow control was actually supposed to work.

I'd appreciate hearing from anyone who's gotten a QMS, or indeed any parallel
printer to work with the Amiga and knows what the problem might be.

-----------

I'd be content to use the serial connection, except for the fact that I can't
get things like the AmigaDOS 'copy' command and SKsh's 'cp' to open the serial
port to which the printer is connected with anything other than the default
preferences settings for baudrate and handshaking protocol.  I have a 2400bps
modem connected to port 1 (SER2:) on the 2232 at 2400 baud, RTS/CTS, while the
printer is connected to port 2 (SER3:) and wants 19,200 baud, XON/XOFF.


AdvManyThanksance,
Bryan

gallo@dtrc.dt.navy.mil (Gallo) (02/17/91)

I had a QMS_PS-410 hooked up to a AMAX-II/Amiga 2500
combo and never did get it to print Postscript from the
Mac side. The only way I ever got  print_output was
by sending a postscript file to the printer using a
communication program. In monitoring the printer during
one of the sends done from_the Mac side, I saw_many
error codes given off by the Postscript processor.

One difficulty is  that QMS has all their docs  and
support people_tuned in  to DOS and Macintosh users
and no one I talked with in tech support seemed to know
about Amiga's  at all...Good luck! <Chip>

svante@kberg.se (Svante Gellerstam) (02/17/91)

In article <276@mohawk.cs.utexas.edu> bryan@cs.utexas.edu (Bryan Bayerdorffer @ Wit's End) writes:
>I've just connected a brand-new QMS PS-410 laser printer to my A2500/30 (rev 4.5
>motherboard) via both the parallel port and one of the serial ports on a 2232
>multiserial card.  While the serial connection works fine, I'm experiencing
>problems when sending postscript files over the parallel connection.  The

I am afraid I cannot offer any help, just confirmation of your
problem.  We have the same problems with the 410 connected to an
A3000. The main problem is that the printer hangs sometimes when the
handshake gets out of sync. As the parallel.device has 'infinite
retry' this is equal to a hung machine and a reboot. Very annoying. If
you succeed in solving the problem, I would love to hear about it.

>AdvManyThanksance,
>Bryan

/svante

-- 
Svante Gellerstam		svante@kberg.se, d87sg@efd.lth.se