tcamp@UNIXPC.UUCP (Ted Campbell) (07/13/89)
I seem to be having trouble with my newly upgraded Unix PC motherboard and/or printer drivers. Any suggestions will be appreciated. SYMPTOMS: The printer doesn't work. When I 'cat' a file to /dev/rawlp, the first line of the file is printed, and a line feed (no CR) is given. When I 'cat' a file to /dev/lp, six or seven random characters from the first line are printed, then a formfeed is issued. WHAT I'VE TRIED SO FAR: I have tried this only with the parallel port, but with two printers, to parallel cables, and two different versions of the system software (2.0 and 3.5). I have repeatedly tried reinstalling the printer drovers with various settings (I'm using a Star NX 10 printer, so I have tried Epson, Star SG10, and a number of other options). Same results every time. I even sent the board back to San Leandro, they sent me another one, and I have precisely the same results. CLUE: The failure has only occurred after I myself have installed a motherboard. ANy idea what I might be doing wrong in installation that would account for this nonsense? WEIRD: The parallel printer test on the diagnostics disk appears to work perfectly. I am, baffled also, Ted A. Campbell Duke Divinity School ...!dukeac!numen!tcamp
root@ycomp.UUCP (The Super User) (07/14/89)
I have what appears to be the exact problem with a board just back from San Leandro. If I 'cat' to /dev/rawlp 1 line at a time it works OK. More than 1 line at a time results in mostly garbage. Diagnostics work fine. My guess is that the BUSY line is not working correctly. The board they sent back (I don't think its the same one I sent them) has 'REV H' on it. I first thought they had sent me a bad board, but... Time for a closer look at the board. AH HA!! Someone has soldered a diode across a capacitor that comes off of pin 11 (the BUSY line) on the parallel port. I wish I had another board handy to see if it has this mod. I really don't feel like disecting another machine. I wonder what will happen if I remove it. It's worth a try. Works fine now!!! Either they put the diode in backwards or they put in the wrong place. It can't possibly work the way they had it (with the cathode to ground) as it would short the BUSY signal to ground. If anyone has a board they can look at, I would be interested to know if they have this diode and where & how it is connected. There are 4 small caps between 2 resister packs just above the printer port as shown below. Mine had the diode across cap 1 with the cathode (banded end) up. | 1 2 | | 3 4 | | | __________________ | parallel port | | | I probably would have sent the board back to San Leandro until I saw Ted's request for help. I hope this helps someone else also. Bob Campbell
wieland@ea.ecn.purdue.edu (Jeffrey J Wieland) (07/18/89)
In article <124@UNIXPC.UUCP> tcamp@UNIXPC.UUCP (Ted Campbell) writes: > >I seem to be having trouble with my newly upgraded Unix PC motherboard >and/or printer drivers. Any suggestions will be appreciated. > >The printer doesn't work. When I 'cat' a file to /dev/rawlp, >the first line of the file is printed, and a line feed (no CR) >is given. When I 'cat' a file to /dev/lp, six or seven random >characters from the first line are printed, then a formfeed >is issued. > >The failure has only occurred after I myself have installed a >motherboard. ANy idea what I might be doing wrong in installation >that would account for this nonsense? > >The parallel printer test on the diagnostics disk appears to work >perfectly. We have seen three 3b1's with this problem. One of them had it from the day it was delivered by AT&T, the others "developed" it after the motherboards were replaced for other reasons. I seems that AT&T is not checking the motherboards out very well after they replace them. All of these machines would pass diagnostics; none of them could print properly. We have also had problems with "new" motherboards having bad serial ports. In both instances, the port could send data but would not receive any data. Don's assume that the replacement motherboard is any better than the one you took out. -- Jeff Wieland wieland@ecn.purdue.edu