home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.modems
- Path: sparky!uunet!decwrl!sgi!rhyolite!vjs
- From: vjs@rhyolite.wpd.sgi.com (Vernon Schryver)
- Subject: Re: What's a good 14400 modem?? (Intel?)
- Message-ID: <p9nuclo@rhyolite.wpd.sgi.com>
- Organization: Silicon Graphics, Inc. Mountain View, CA
- References: <1992Aug24.143532.8721@lowton.rn.com> <PZR7PB2w165w@infopls.chi.il.us> <1992Sep1.034335.22089@psg.com>
- Date: Tue, 1 Sep 1992 05:01:12 GMT
- Lines: 68
-
- In article <1992Sep1.034335.22089@psg.com>, randy@psg.com (Randy Bush) writes:
- > vjs@rhyolite.wpd.sgi.com (Vernon Schryver) writes:
- > > I tested a pair of Intel 14.4EX modems with SLIP. I was unable to move
- > > more than 1800 Bytes/sec through them, compared to > 3000 for all other
- > > v.32bis/v.42/v.42bis modems I tested, some of which did > 3600 B/s.
- > > Judging from the LED's, the Intel 14.EX cannot transmit at the same
- > > time as it it receives. That is implausible, but it would explain the
- > > less than amazing speed.
- >
- > Gak! Vernon, you ain't no dummy, so something subtle was going on. We
- > regularly get 3.7kcps+ on SLIP with ka9q and PC-Route with these puppies.
- > They're the best we've tried, and we've tried and tried.
- >
- > Our configs are simple.
- >
- > AT&F"H3\N3E0Q1&D0&W
-
-
- Well, maybe I'm a dummy about this modem. I was using
-
- &F L2 M4 X4 &C1 &D3 &Y0 S2=128 S25=1
-
- L2 and M4 should be irrelevent.
- So should X4.
-
- &C1 is the default, so I don't know why I set it.
-
- &D3 is a fetish of mine that saves a lot of grief
-
- &Y0 is something I learned the hardware from Telebits
-
- S2=128 just turns off +++
-
- S25=1 just makes the modem respond to DTR (why would you want the
- default which ignores 50 ms DTR turn-off's? I'd expect that to
- make it some times fail to honor the computer's request to hang
- up. Oh, well. It's not relevant.)
-
- "H3 is the default, so I don't know why you set it
-
- \N3 is the default, so I don't know why you set it
-
- E0 and Q1 should have no effect on speed
-
- &D0 is a difference that should not affect speed.
-
- None of those appear to me likely to affect the speed at which the
- modems move bytes.
-
- The tests I was running were the same as with the other modems. I use
- a pair of Silicon Graphics IRIS's with "6-port boards", with new
- firmware which I guarantee can do about 35KBytes/sec FDX as measured by
- 4 to 5 `cat foo > /dev/ttyd#` processes, loop-back connectors, and
- either a 'scope or `sar -y 1 10000`.
-
- So, what's the deal?
-
- Could it be that your 3.7KB/s is half-duplex but my 1.8K is half of
- full-duplex number?
-
- My "throughput test" consist of `ping -c ZZZZ remote`, increasing the
- value of ZZZZ until the line stops keeping up, as observed for 20 to 60
- seconds. That forces the modems to process ZZZZ*2 bytes. In other
- words, the other modems I've tested were really shuffling up to 7.2KB/s
- of data, while the Intel 14.4EX would only do 3.6KB/s.
-
-
- Vernon Schryver, vjs@sgi.com
-