karches@utdallas.edu (Tom Karches) (06/06/90)
I am having a problem interfacing a TI 855 dot matrix printer to a Sparcstation. I initially created a cable which successfully interfaces the printer to the Sparc. Other versions of the cable that I have made (identically wired as far as I can tell) cause the system mouse and keyboard activity to freeze when the cable is connected between the computer and printer and the printer is turned on. The cable wiring is as follows : DB-25 at TI-855 printer Sparc 1----------------------- Chassis Ground 2----------------------- RXD - Recieve data to printer 3----------------------- TXD - Transmit data from printer 4--. 5--' 6----------------------- DTR - Data Terminal Ready 7----------------------- Signal Ground 12---------------------- Ready/Busy (Printer busy when low) 15--. 17--| 24--' 20---------------------- DSR - Data set ready (from host) No keyboard or mouse action causes a response until the printer is turned off, at which time the buffered keyboard and mouse movements are executed. I also get the following errors in the console : zs2 : silo overflow zs3 : silo overflow I made no kernel changes on either system where the printer has worked. I also did not modify any of the eeprom values (set to defaults). The printer is set for software (xon,xoff) handshaking. I realize that the cable wiring might be overkill, but my original version does work. Whether or not the printer is on when connected makes no difference. I tried both cables (working and non-working) and printer on another system, and the results are the same. With my original cable, the printer works. With the new cable, it dosen't. The only difference that I can percieve between different cables is the length of the wires that tie pins 4 and 5 together and pins 15,17 and 24 together (wires are shorter on the cable that works). Are there timing problems involved here? Should I be using twisted pair or heavily shielded wire? Has anyone seen symptoms like this before? It appears that I have a "magic" cable and I don't think I like that. All help/comments appreciated. Please e-mail me directly; I don't have news access. Regards, Tom Karches ----------- Disclaimer : --------- / Nothing is foolproof.. \ --. Texas Instruments I speak for myself \ Fools are ingenious... / | karches@utdallas.edu --------------------------------------------------------'