LSI@UMass.BITNET (Peter Lawall, Logical Solutions, (413) 256-6800) (12/09/86)
Rick Flashman (Flash%UMass.BITNET@WISCVM.WISC.EDU) writes: > BTW, Atari's Printer port on the ST is NOT all that STANDARD. If you > are trying to hook a Okidata dot-matrix printer to an ST, talk first to > your local Atari dealer. There are cable modifications that must be made > cause Atari did not use a standard connection....some voltages are > messed up or something. Need a tech for that one. The only technically "non standard" thing about the parallel printer port is that it only supports BUSY handshake as opposed to both BUSY and ACKNOWLEDGE. This isn't usually a problem, as every printer that I know of provides a BUSY output in addition to the ACK output. The problem with OKI printers (at least the 192, as this is the one that I diagnosed and fixed said problem on) is that they provide fairly low value resistive termination of all the input signals. Unfortunately, in an attempt to save every last penny for Tramiel's pocket, Atari decided not to buffer the STROBE signal, which originates at one of the MOS devices (probably the 68901). The resistive termination excessively loads this MOS output, which causes the signal in its active state to fall outside of the recommended spec for a logic "0". Anyway, the printer worked, but duplicated characters once every few lines due to the marginality of this signal. To cure the problem one could probably merely buffer the STROBE signal with a simple buffer gate like one section of a 74LS08, but as I had a 74LS14 in the top drawer of my desk, I used two sections of this back to back for added noise immunity. Anyway, this cured the duplication of characters 100%, and the printer now worked fine... Keep it in mind if you come across a printer that exhibits this problem. Incidentally, the above above speculation of the STROBE signal coming from an unbuffered MOS output is actually just that--speculation; however, the fact that the added buffer cured the problem seems to point strongly in this direction (as does the penny-pinching nature of Atari). I didn't bother opening the Atari at the time to verify this for sure, and don't have another one handy right now to check it on. But I'm sure that was it... Peter
cabbie@chinet.UUCP (12/10/86)
Actually the STROBE output for the printer originates from the music chip. There have been a few failures of this output pin but no real major percentage have failed. Mine did in my development cpu but that is probably my fault. I have been known to blow an occasional chip now and then. -- ******************************************************************************* Any opinions expressed above are my own. Rich Andrews They can be yours too. Please send $19.95 to.....ihnp4!chinet!cabbie *******************************************************************************
grr@cbmvax.UUCP (12/10/86)
In article <861208225517.00000698.ATCV.MA@UMass> LSI@UMass.BITNET (Peter Lawall, Logical Solutions, (413) 256-6800) writes: >Rick Flashman (Flash%UMass.BITNET@WISCVM.WISC.EDU) writes: > >> BTW, Atari's Printer port on the ST is NOT all that STANDARD. If you >> are trying to hook a Okidata dot-matrix printer to an ST, talk first to >> your local Atari dealer. There are cable modifications that must be made >> cause Atari did not use a standard connection....some voltages are >> messed up or something. Need a tech for that one. > > The only technically "non standard" thing about the parallel printer port >is that it only supports BUSY handshake as opposed to both BUSY and >ACKNOWLEDGE. This isn't usually a problem, as every printer that I know of >provides a BUSY output in addition to the ACK output. > > Peter It should perhaps be pointed out that not many micros, nor many micro printers really conform to the real Centronics 'standard'. This however, doesn't keep most of them from inter-operating, although there are these occasional problems. Specifically, the Centronics 101A manual required ~15mA drive on the data and control lines. Not too many NMOS chips can do this reliably. There are also a bunch of relatively useless or partially redundant signals that most people don't bother with. Also, as Peter mentioned you can do handshaking on either BUSY or ACK, but on some printers you may find some performance degredation if you choose the wrong one, due to buffering techniques and the relation of BUSY to the actual printing action. -- George Robbins - now working for, uucp: {ihnp4|seismo|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@seismo.css.GOV Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)