home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!sun-barr!olivea!hal.com!decwrl!adobe!andrich.mv.us.adobe.com!user
- From: rfall@mv.us.adobe.com (Richard Fall)
- Newsgroups: comp.dcom.modems
- Subject: Re: Why is this modem cable's DTR mapped to RTS?
- Message-ID: <rfall-080992145008@andrich.mv.us.adobe.com>
- Date: 8 Sep 92 22:12:08 GMT
- References: <1992Sep6.050516.4767@news.columbia.edu>
- Sender: usenet@adobe.com (USENET NEWS)
- Followup-To: comp.dcom.modems
- Organization: Adobe
- Lines: 121
-
- In article <1992Sep6.050516.4767@news.columbia.edu>,
- ia4@cunixa.cc.columbia.edu (Imran Anwar) wrote:
- >
-
- [....]
-
- > But , despite attempts, we could not figure out what was happening. I did
- > remember that the program ONLY works with a modem cable the company gave.
- > And one of the things I noticed between this and other modem cables I had
- > was that with it the Modem Ready light and CTS come on but NOT the DTR light.
- >
- > Interestingly, when this cable is hooked to the Worldblazer the RTS light
- > comes on (the Trailblazer Plus does not have an RTS light or I would have
- > noticed this sooner). instead of the DTR light. When I make any other
- > communications program eventhen the RTS light comes on instead of DTR. And
- > "inverting DTR" in the application makes the RTS light go on and off.
-
- I'm a little confused by what you've described, but it sounds to me very
- much like you've got a problem with the setup of your WorldBlazer and/or
- you serial cable.
-
- Let me see if I can help by providing an explanation first:
-
- The Mac serial port has a limited number of pins available for its use.
- For this reason, there is only _one_ signal available for _any_ handshaking
- use--it's a signal called "Handshaking Out (HO)".
-
- Depending on the communications program you use, this signal may, in
- general, be used for one of two purposes: either as a means to signal
- the modem to reset/go into command mode, or as a means to provide hardware
- flow control from the modem to the Mac.
-
- In the former case, this is done by connection HO to the DTR pin of the
- modem, and configuring the modem to treat a deassertion of this signal
- in the appropriate manner (see the documentation for your WorldBlazer
- and refer to the &D register settings--you probably will want to set
- &D1, &D2,or &D3 for this mode). In this case, however, since there is
- nothing connected to RTS, you must tell the WorldBlazer to ignore the
- signal level on the RTS pin, by setting S58 equal to 0 or 3 (no flow
- control, or XON/XOFF flow control, respectively). You must be sure,
- however, that the communcation program is using HO in the fashion, since
- if it uses this signal for flow control, the modem will drop the
- connection or drop into command mode the first time it sees DTR go
- away; probably not what the program meant to do.
-
- In the case where the comm package uses HO for hardware flow control,
- you must connect HO to RTS, and tell the modem to use RTS/CTS flow control
- (S58 = 2 and S68 = 2 or 255), and you must also tell it to assume that
- DTR is asserted all the time, since there will be no signal driving it
- (&D0).
-
- In general, you're better off using RTS/CTS flow control, since many
- file transfer protocols assume a clean 8-bit data path, end to end, and
- won't "like" seeing XON/XOFF characters disappearing from the data
- stream, as they're interpreted as flow control characters by the other
- modem.
-
- There is an equivalent situation, BTW, with regard to input handshaking
- to the Mac--a signal called "Handshaking In (HI)" is available for use
- either as a signal from the modem for certain events (usually used as
- a DCD indication) or as flow control in the other direction (CTS).
-
- So, bottom line, you need to know what your comm program wants to use
- HO/HI for, and wire and set S-registers appropriately.
-
- > That really suprises me. Could this have any implications for fast data
- > transmission, because when I use ordinary hardware cables with the
- > Trailblazer or with the WorldBlazer and send data to the same numbers in
- > Pakistan with any other standard program like White Knight, Microphon etc
- > using X-, Y- or Z-modem in all the cases the file transfers fail, or
- > achieve 80 cps or even lower :-(
-
- You bet. If hardware flow control is not set up correctly, but the
- comm package expects it to be, you will be overrunning the receiving
- modem, and retranmitting packets, lowering throughput.
-
- > I am totallyt confused if it is just some super programming that they
- > employed and is it related to DTR/RTS switching on the cable or is that
- > kjust a trick to sell expensiuve cables. The company is in England
- > and they want about $5000 yes that is $5000 for an UPGRADE!
-
- Well, several companies seem to use this as an excuse to tell you that you
- need to buy an expensive serial cable to work with high-speed modems, but
- you could just as easily wire one yourself.
-
- BTW, I try to solve this problem in a slightly different way, since I have
- comm packages that work both ways I've described above. I have created a
- pair of WorldBlazer settings, once for each, and have the comm program
- pick the right one when it initializes the modem. The cable I use has
- HO wired to _both_ DTR and RTS. When I want to use hardware flow control,
- I set S58=2, S68=255, and &D0, and the modem ignores the transitions on
- DTR. When I want to use XON/XOFF flow control, and have the comm program
- be able to drop the connection by lowering DTR, I set S58=3, S68=255, and
- &D2. The transitions on RTS are then ignored.
-
- In fact, several commercial cables are wired this way (DTR and RTS wired
- together) causing no end of confusion for some.
-
- Unfortunately, it's not possible to do the same for the input direction to
- the Mac, so I simply have CTS on the modem connected to HI. I can't
- then "see" DCD at the Mac, but that's a minor annoyance.
-
- > The program shows that it is based on the Think C libraries so I am not
- > evenm sure if the Think C code has some such standard high speed data
- > transmission routines.
- >
- > Hope people on the net can please help me out with this.
- >
- > Thanks
- >
- > Imran
-
- I hope this explanation helps. Good luck.
-
- +--------------------------------+-----------------------------------+
- | Richard Fall | Voice: (415) 962-4889 |
- | Adobe Systems, Incorporated | Fax : (415) 903-9652 |
- | P.O. Box 7900 |-----------------------------------|
- | 1585 Charleston Road | Internet: rfall@mv.us.adobe.com |
- | Mountain View, CA 94039-7900 | UUCP: ..!decwrl!adobe!rfall |
- +--------------------------------+-----------------------------------+
-