| ??? 02/22/07 06:52 Modified: 02/22/07 06:57 Read: times |
#133483 - I don\\\'t see what\\\'s not straight about this Responding to: ???'s previous message |
Caslav Nedeljkovic said:
To make life easier, please, tell me with which of the following assertion you don't concur? For your reference I've added numbers before each of them.
1) PCs rs232 port can (note, it is can not should or must) be used for communication using only 3 wires. This is opposed to your statement PC's require special wiring, i.e. pins 4 and 5, and pins 6, 8, and 20 (based on the DB25 model) have to be connected together, else the PC won't talk. 2) By using tiebacks you can disrupt intended functionality of rs232 communication if by hardware/software design this communication requires additional signaling. This is opposed to your statement The ones that don't need the signals generally ignore 'em. 3) Tiebacks are most common way of wiring rs232 cable, but that doesn't mean that it is the only way to wire it. This is to answer your question I wonder why so many EVB's have those tiebacks etched into the board? 4) Which part of the following table you don't understand or object to (apart from using DB9 nomenclature): http://www.lammertbies.nl/comm/info....html#conc This is opposed to your argument for over 30 years, and haven't yet encountered one that wouldn't tolerate this tieback scheme. I hope you'll give me some straight answers this time. Best luck, CN Caslav, What I've told you is that (a) unmodified PC's, running, say, a COPY command from the DOS prompt under DOS, not WINDOWS, expect to have the hardware connected in some way. (b) If you provide the tieback I suggested, which I've been using with PC's since 1981, when I encountered my first PC/XT, and which has, so far, not let me down, it seems to work, and then allows the use of X-on/Xoff, if that's a valid alternative, e.g. in the case of a program like PROCOMM (not the Windows version.) With Windows, you have, and need, more options. I don't suggest that situations can't exist wherein this tieback might not work, but, so far, despite the many, many hundreds of instances I've hooked up, I've never encountered one that didn't work satisfactorily. The null-modem box I built back in the early '80's has worked fine. The cables I've wired have worked fine and trouble free, in some cases for decades. In the event I want to disconnect one or another of the wires, I simply open a switch. It's really not rocket science. DTR is a signal to tell the DCE that the terminal is present, in which case it's generally ready. This is witnessed by the many terminals that assert it as soon as power is turned on. RTS says the terminal wants to send, and CTS gives permission. Why IBM chose to use DSR and DCD in the way in which they did is a mystery. It's not far off what most people did, however. There hasn't been a convention as to how these circuits are to be used, so everyone is allowed to do whatever they see fit. The matter of tying back some signals in order to satisfy a handshaking requirement is not new. I've always preferred the 3-wire comm's that are sufficient if one uses X-on/Xoff flow control. PC's, however, don't make that easy, because it's up to the application to control the serial port control signals. Hyperterminal allows you to manage the handshaking requirement. Many applications do not. All I'm after is a way to get the job done. If I send a file from point A to point B using a serial port, I don't want to have to worry about whether the handshake works. For a one-off task, I can manage it with my little switch box. If needed over the long term, I can wire a cable. If that weren't the case, nearly no equipment would talk to a PC, since most older serial printers don't adhere to the IBM-PC handshake. RE |
