home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!sgigate!odin!news.csd.sgi.com!oblivion.csd.sgi.com!kevinb
- From: kevinb@oblivion.csd.sgi.com (Kevin Brokaw)
- Newsgroups: comp.dcom.modems
- Subject: Re: IBM 7855-10 Disconnect at Phone Pick-Up
- Message-ID: <1992Sep9.163125.14059@news.csd.sgi.com>
- Date: 9 Sep 92 16:31:25 GMT
- References: <9209090702.AA17611@ucbvax.Berkeley.EDU>
- Sender: news@news.csd.sgi.com (Net News)
- Organization: Silicon Graphics Computer Systems
- Lines: 53
-
- In article <9209090702.AA17611@ucbvax.Berkeley.EDU>, ST7021@SIUCVMB.SIU.EDU writes:
-
- |> A problem that I experience is that it lose connection when someone in my
- |> house picks up a phone. It seems to do this commonly when running at
- |> 9600/12000bps.
- |>
- |>
- |> This device will accept line noise with no problems. I've (for fun...) brought
- |> the line noise upto "F" (on a scale from 0-F). It still hung in there.
- |>
- |> o Is this some V.32(bis) thing (my old Hayes VS9600 doesn't do this at
- |> 9600)?
- |> o Could it be interpreting the drop in volume (or whatever) as a Break
- |> signal? Would that be relevant?
- |> o Is there a register that I can alter that will make it more tolerant
- |> to this?
- |>
- |> I can understand how rough a local line pick-up can be on the carrier, but I
- |> was hoping the error correction would be a bit more fault tolerant.
-
- Thomas,
- Based upon your mention of error correction, I'd imagine you already have ECL
- enabled for your session (must be, actually, as the 7855 won't fall forward to
- 12k unless an ECLC link is established at 9600 first). My guess is that the type
- of interruption thrown on the line by picking up the other receiver is so severe
- that the modem can't recover... perhaps the signal level decreases so drastically
- that the carrier can't be maintained. On the 7855 I had, it would survive another
- extension phone being briefly picked up and replaced. It's not suprising that the
- V-series modem doesn't suffer from this same condition, as I don't believe its'
- modulation technique is as demanding upon signal levels as the V32 specification
- is... Hayes' modulation technique is not as well known to me as, for example,
- USR's HST technique, which is MUCH less demanding on the line (only maintains
- full speed in one direction, with a greatly reduced (~750bps) backchannel. V32
- spec calls for full 9600bps bi-directional channels). I don't believe the 7855
- can be configured to be more tolerant of signal level drops... since it's designed
- to be certifiable in virtually every country IBM does business in, it's got a
- rather narrow set of tolerances... my favorite "idiosyncracy" of the modem was
- the inability to keep it off-hook with an ATH1 command for more than 60 seconds
- before it hangs itself up... apparently because some countries do not permit
- working telephone sets to be left off-hook. Made running my nightly maintanance
- routines on my BBS a little interesting, as I ran a watchdog program to boot the
- PC if a pre-specified number of rings were detected by the modem without the system
- issuing an answer command. I finally found that setting the modem in loopback test
- kept it offhook until the test was terminated.
- Cheers,
- Kevin
-
-
- --
- Kevin Brokaw - Product Support Engineer |
- Silicon Graphics Computer Systems | All wight, rho sritched
- Internet: kevinb@sgi.com | my kegcaps awounf?
- |
-