dgil@pa.reuter.COM (Dave Gillett) (08/28/90)
In <sVRio1w162w@kk4fs.UUCP> genesis@kk4fs.UUCP (Genesis) writes: >Is it possible to push a 2400 baud modem to 4800 without doing any damage?If >so,what must i do to do this?THANKS! There are three areas where you might wish to change the bps rate of your transmissions. Each has its own problems and pitfalls. 1. CPU -> modem This is the digital link from the computer to the modem. (This may not be relevant if you have an internal modem.) If you send characters too fast here, the modem will not be able to handle each one before the next arrives. A Hayes-type modem uses the "AT" at the front of a command to determine the rate at which you are sending to it--in the process, it may not be able to do anything else with them. In order to detect the speed you are using, it must sample the line at some high frequency; the exact frequency used imposes a physical maximum detectable rate. The firmware in the modem imposes a logical maximum configurable rate--which is probably 2400 in this case... 2. modem <-> modem This is the analog link between your modem and the other modem, probably over a telephone line. At higher baud rates, signals are more likely to be lost due to noise. Most 1200, 2400 and above modems actually use 600 baud in each direction, and use the phase of the signal (okay; 9600 gets more complicated...) to encode multiple bits per baud. Since you probably don't want to tinker with the phase encoding/decoding logic in the modems, your alternative is to switch states more often. 2400 bps modems are already at about 80% of the reliable baud rate for voice-grade lines, so switching states twice as often means that you're likely to encounter noise bursts big enough to wipe out entire groups of bits. No good. 3. modem -> CPU This is the digital link back from the modem to your CPU. At some low level, your code that talks to the serial port has to be able to grab characters as fast as the modem can supply them. This is subject to a lot of variables, but it *can* be a non-trivial problem, especially above 2400 bps, which you should be prepared to tackle. Again, the modem components (especially firmware) may impose a maximum here. Conclusion: Assuming that you can get around logical conditions imposed by the modem firmware, you're likely to be running up against constraints imposed by the physical universe. It's not likely to be worth the effort; it's definitely not going to stand comparison against buy 9600 bps modems. Dave
tronix@polari.UUCP (David Daniel) (08/29/90)
[]There are three areas where you might wish to change the bps rate of your
^^^ ^^^^
and...
[] over a telephone line. At higher baud rates, signals are more likely to
^^^^ ^^^^^
are two completely different critters, not interchangeable terms. Baud refers to
how often the analog signal ocillates during transmission, named after the
a French guy - Baudot (or some such french splling). Bps is bits per second.
If memory serves, and it probably doesn't, 1 baud represents roughly 2.xx bits.
[] be lost due to noise. Most 1200, 2400 and above modems actually use
[] 600 baud in each direction, and use the phase of the signal (okay; 9600
^^^ ^^^^
That's actually 1200+ bps. Baud rates BTW, are obsolete. If you see a modern
modem rated at 2400, the speed is measured in bps.
[] Assuming that you can get around logical conditions imposed by the modem
[]firmware, you're likely to be running up against constraints imposed by the
[]physical universe. It's not likely to be worth the effort; it's definitely
[]not going to stand comparison against buy 9600 bps modems.
In other words, the answer is no, unless you use built it transfer protocols -
error correction, data compression or ideally a combination of both to increase
through-put.
--
David Daniel (The man with no disclaimer) tronix@polari.wa.com
"Beware the Truth. If you find a Truth it can demand that you make painful
changes." - Frank Herbertconib@gvgpvd.GVG.TEK.COM (Coni Britten) (08/30/90)
In article <2447@polari.UUCP> tronix@polari.WA.COM (David Daniel) writes: >[]There are three areas where you might wish to change the bps rate of your > ^^^ ^^^^ > and... > >[] over a telephone line. At higher baud rates, signals are more likely to > ^^^^ ^^^^^ > >are two completely different critters, not interchangeable terms. Baud refers to >how often the analog signal ocillates during transmission, named after the >a French guy - Baudot (or some such french splling). Bps is bits per second. >If memory serves, and it probably doesn't, 1 baud represents roughly 2.xx bits. I've found one published definition of "baud"; it's in my Webster's Collegiate: a variable unit of data transmission speed usu. equal to one bit per second So ... they are not necessarily two *completely* different critters. They can in fact be interchangable. And in my experience (13 years in engineering), in common practice this is true! So true, in fact, that even Webster defines it as "usually" true, and gives no other alternative ratio. Of course, Webster is not exactly the last word in scientific definitions, but it does define "common usage". Coni Britten conib@gvgpvd.GVG.TEK.COM -or- ... ucbvax!tektronix!gvgpsa!gvgpvd!conib
kdq@demott.COM (Kevin D. Quitt) (08/30/90)
A baud is an analog unit, representing any number of bits.
Traditionally, for slow modems, there is a one-to-one correspondence.
For higher speed modems (2400), this is no longer the case.
To increase your baud rate, you need a corresponding increase in you
bandwidth - normal phone lines will not pass much over 1200 baud. In
order to increase your data thruput, you have to encode more bits in a
given baud. 2400 bit per second modems actually run at 1200 baud, but
use 4 different baud to represent four possible states.
--
_
Kevin D. Quitt demott!kdq kdq@demott.com
DeMott Electronics Co. 14707 Keswick St. Van Nuys, CA 91405-1266
VOICE (818) 988-4975 FAX (818) 997-1190 MODEM (818) 997-4496 PEP last
96.37% of all statistics are made up.hes@ccvr1.ncsu.edu (Henry E. Schaffer) (08/30/90)
In article <1664@gvgpvd.GVG.TEK.COM> conib@gvgpvd.GVG.TEK.COM (Coni Britten) writes: >In article <2447@polari.UUCP> tronix@polari.WA.COM (David Daniel) writes: > ... [bit per second and baud] >>are two completely different critters, not interchangeable terms. ... >>If memory serves, and it probably doesn't, 1 baud represents roughly 2.xx bits. >I've found one published definition of "baud"; it's in my Webster's Collegiate: > > a variable unit of data transmission speed usu. equal to one bit per > second > >So ... they are not necessarily two *completely* different critters. They can >in fact be interchangable. And in my experience (13 years in engineering), in >common practice this is true! So true, in fact, that even Webster defines it >as "usually" true, and gives no other alternative ratio. Dictionary definitions (are supposed to) reflect usage. Back in the days of the popular Bell 103 and compatible modems which used FSK for speeds in the 110 - 300 range - the bps and baud rating were numerically the same (i.e., 1 bps = 1 baud) and people got into the habit of using the two interchangeably, even though their engineering/scientific definitions were not the same. These days most modems are faster than 300 bps and handle multiple bps per baud, but still the terms are often used interchangeably - so let me reword Webster's definition a variable unit of data transmission speed usu. incorrectly considered equal to one bit per second > ... --henry schaffer n c state univ
dgil@pa.reuter.COM (Dave Gillett) (09/01/90)
In <2447@polari.UUCP> tronix@polari.UUCP (David Daniel) writes: {using [] where he quotes me} >[] ... bps ... baud rates >are two completely different critters, not interchangeable terms. Baud refers to >how often the analog signal ocillates during transmission, named after the >a French guy - Baudot (or some such french splling). Bps is bits per second. >If memory serves, and it probably doesn't, 1 baud represents roughly 2.xx bits. You lose, turkey. While you've got the definitions of bps and baud rate correct (boy, what a brain-o!), you (a) missed the point that in fact I was carefully using each correctly where that was what I was talking about, and (b) 1 baud is essentially 1 bps at low speeds; I was briefly recounting *why* 1 baud = 2 bps *at*1200*bps*using*Bell*212*or*v.22 and 1 baud = 4 bps using v.22 bis (2400 bps). If you'd read what I wrote, your memory wouldn't *have* to serve! >[] be lost due to noise. Most 1200, 2400 and above modems actually use >[] 600 baud in each direction, and use the phase of the signal (okay; 9600 > ^^^ ^^^^ >That's actually 1200+ bps. Baud rates BTW, are obsolete. If you see a modern >modem rated at 2400, the speed is measured in bps. That's 1200 bps, 2400 bps, or whatever the modem is rated for. RTFP! Since you know what a baud is, you also *know* that it is no more obsolete than any other unit of measurement: did "Megahertz" become obsolete when "MIPS" were invented? No, they measure related things in different ways, just like baud and bps do. (You'll notice that in the quote I refer to the numerical rating of the modems but *not* to the units; you'll sometimes see them labelled "bps", sometimes "baud", and sometimes just with the number. And since I then specify that these really use 600 baud (again properly using the term as it is defined!), then any reasonably bright reader would say: "Gee, if 2400 bps modems really use 600 baud, then I guess that it's incorrect to label them as 2400 *baud*." The point certainly doesn't need to be laboured--and I don't need to be flamed for it! >[] Assuming that you can get around logical conditions imposed by the modem >[]firmware, you're likely to be running up against constraints imposed by the >[]physical universe. It's not likely to be worth the effort; it's definitely >[]not going to stand comparison against buy 9600 bps modems. >In other words, the answer is no, unless you use built it transfer protocols - >error correction, data compression or ideally a combination of both to increase >through-put. No, is indeed the answer. The question, though, was phrased in terms of increasing the bps rate and not in terms of increasing the throughput. Perhaps Daniel did a better job than I did of reading the original poster's mind on this point. Dave
storm@texas.dk (Kim F. Storm) (09/01/90)
hes@ccvr1.ncsu.edu (Henry E. Schaffer) writes: >These days most modems are faster than >300 bps and handle multiple bps per baud, but still the terms are >often used interchangeably - so let me reword Webster's definition > a variable unit of data transmission speed usu. incorrectly considered > equal to one bit per second Which is illustrated perfectly by the Telebit TrailBlazer running in PEP mode which - as far as I remember - runs at just 7 baud. So here, you have a 1 baud ~ 2000 bit relation. -- Kim F. Storm <storm@texas.dk> No news is good news, Texas Instruments A/S, Denmark but nn is better!
peter@sugar.hackercorp.com (Peter da Silva) (09/02/90)
A Trailblazer gets 14000 bps with 6 bauds per channel. Both of these measurements are relevent. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
jon@savant.uucp (Jon Gefaell) (09/04/90)
In article <659@texas.dk> storm@texas.dk (Kim F. Storm) writes: >hes@ccvr1.ncsu.edu (Henry E. Schaffer) writes: > >>These days most modems are faster than >>300 bps and handle multiple bps per baud, but still the terms are >>often used interchangeably - so let me reword Webster's definition > >> a variable unit of data transmission speed usu. incorrectly considered >> equal to one bit per second > >Which is illustrated perfectly by the Telebit TrailBlazer running in PEP >mode which - as far as I remember - runs at just 7 baud. >So here, you have a 1 baud ~ 2000 bit relation. Will misinformation never cease? I'll add to it now: The PEP method uses up to 511 7Hz 'channels' each having one baud. Each channel is 'tested' to determine whether that baud is stable enough to encode 2, 4 or 6 bits (FSK, PSK, QPSK respectively) Have I helped muddy things up? (i.e. set me strayt on where I've goofed) -- +----------- Domain? DOMAIN? We Don't Need No Steeeenkin' Domain! -----------+ |I wish I had something interesting to put in my .signature file, but I don't| +-savant!jon@virginia.edu {...}!uunet!virginia!savant!jon jeg7e@virginia.edu-+
jmaynard@thesis1.hsch.utexas.edu (Jay Maynard) (09/05/90)
In article <1990Sep4.150259.12379@savant.uucp> jon@savant.uucp (Jon Gefaell) writes: >In article <659@texas.dk> storm@texas.dk (Kim F. Storm) writes: >>hes@ccvr1.ncsu.edu (Henry E. Schaffer) writes: ...lots of overly pedantic BS. ARRRRRGGGGGH!!!!!!!!!! Who gives a fuzzy rat's posterior how many bauds in the absolutely correct, theoretical sense a modem runs???? The _only_ important measure - unless you're designing the damned things - is how fast it gets data from point A to point B. Anything else is trivia, and arguing about bauds vs. BPS is sheer pedantry. -- Jay Maynard, EMT-P, K5ZC, PP-ASEL | Never ascribe to malice that which can jmaynard@thesis1.hsch.utexas.edu | adequately be explained by stupidity. "It's a hardware bug!" "It's a +--------------------------------------- software bug!" "It's two...two...two bugs in one!" - _Engineer's Rap_
fsb@stsim.ocs.com (Steve Brailsford) (09/06/90)
I don't know if this is anything like what is being discussed, I just picked up this group, but I am looking into getting a 9600 modem and need some advice on what kind to get and what to look for. I seem to see that the V.32 seems to be a general standard for that rate, is this true? Should I bother with any of the other standards that might also come with that? I also hear about a 19.2K modem called Trailblazer, or is that some kind of protocal? I have seen some articles about how you can get 19.2K from 9600 V.32bis modems. Do these modems require leased lines, or is an ordinary phone line ok? Thanks for any help you can give me here. Steve Brailsford >>> uunet!oscmd!stsim!fsb
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (09/07/90)
There are currently three major players in the 9600 and up modem
protocol game. The standard is V.32, but it's not as widely used as the
others. If you are buying modeems for both ends you don't care.
PEP protocol (Trailblazer) offers higher throughput than the line
speed for some protocols, notably uucp and (I believe) xmodem and
kermit. The packet envelope is stripped at the sending end and
reconstructed at the other end. Only the data flows, and throughput can
be 20-40% faster.
The standard used by many ham radio and CP/M systems is (the HST US
Robotics). I'm not sure what this buys you other than compatibility.
That's not saying something bad, I just don't have a lot of time on
these modems.
V.42 is a data compression scheme, and it may be effected by the
recent UNISYS patent on data compression. The last word I heard was that
software compression would be ignored, hardware compression (like a
modem) would be licensed.
If I were buying modems I'd go Telebit T2500, with PEP and V.32.
--
bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen)
VMS is a text-only adventure game. If you win you can use unix.esmith@goofy.apple.com (Eric Smith) (09/07/90)
In article <1990Sep4.150259.12379@savant.uucp> jon@savant.uucp (Jon Gefaell) writes: > Will misinformation never cease? I'll add to it now: > > The PEP method uses up to 511 7Hz 'channels' each having one baud. Each > channel is 'tested' to determine whether that baud is stable enough to > encode 2, 4 or 6 bits (FSK, PSK, QPSK respectively) Yes, misinformation will never cease. The 511 channels are each approximately 7 baud. If they were 1 baud, the highest bit rate thoretically possible with the modem would be 3066 bps. At 7 baud each, the maximum is 21462, of which some is lost since not all carriers will actually be able to encode 6 bits/baud, and some is lost due to packetizing and error handling. -- Eric L. Smith Opinions expressed herein do not necessarily reflect those esmith@apple.com of my employer, friends, family, computer, or even me! :-)
tnixon@hayes.uucp (Toby Nixon) (09/07/90)
In article <2495@crdos1.crd.ge.COM>, davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) writes: > There are currently three major players in the 9600 and up modem > protocol game. The standard is V.32, but it's not as widely used as the > others. If you are buying modeems for both ends you don't care. V.32 is much more widely used that either Trailblazer or HST, if you consider the entire marketplace and not certain limited applications. V.32 modems are available from virtually all major modem manufacturers, and have the distinct advantage of running all unix-related protocols with no need for strange settings or protocol "spoofing". > PEP protocol (Trailblazer) offers higher throughput than the line > speed for some protocols, notably uucp and (I believe) xmodem and > kermit. The packet envelope is stripped at the sending end and > reconstructed at the other end. Only the data flows, and throughput can > be 20-40% faster. All error-control modems strip the start and stop bits from async characters, so they'll all experience a 25% speed increase. The other "increase" you mention is due to protocol "spoofing" in the Trailblazer. But V.32 modems don't need to do this, because they're full duplex to start with; the Trailblazer (and other multicarrier and assymetrical modems) need to do it to make up for the negative impact on protocol performance caused by the basically half-duplex nature of their modulation schemes. > The standard used by many ham radio and CP/M systems is (the HST US > Robotics). I'm not sure what this buys you other than compatibility. > That's not saying something bad, I just don't have a lot of time on > these modems. The HST is "popular" primarily because USRobotics practically gave them away ("sold" them for less than manufacturing costs) to BBS sysops all over the country. Telebit did basically the same thing with Trailblazer modems a few years ago, which accounts for their "acceptance" in the unix community. The HST, like the Trailblazer, is OK for some applications like file transfer, but due to the extremely slow turnaround time is unacceptable for other applications. > V.42 is a data compression scheme, and it may be effected by the > recent UNISYS patent on data compression. The last word I heard was that > software compression would be ignored, hardware compression (like a > modem) would be licensed. V.42 is the international standard for error control in modems, defining two protocols: the primary protocol, LAPM, and an "alternative protocol" for backward compatibility with MNP2-4 modems. You may be thinking of V.42bis, which is the international standard for data compression in modems. It is based on Lempel-Ziv compression, and two companies, Unisys and IBM, claim that patents issued to them apply to the LZ compression used in modem applications; British Telecom also claims that patents for which they have filed for but that have not yet been issued will also apply to V.42bis modems. > If I were buying modems I'd go Telebit T2500, with PEP and V.32. One should select a modem based on, among other things, what the people use that one intends to call. But you're quite right in preferring a modem with V.32 over one with only proprietary high-speed protocols, since V.32 is the international standard and very quickly gaining in worldwide acceptance. -- Toby ----------------------------------------------------------------------------- Toby Nixon, Principal Engineer Fax: +1-404-441-1213 Telex: 6502670805 Hayes Microcomputer Products Inc. Voice: +1-404-449-8791 CIS: 70271,404 Norcross, Georgia, USA BBS: +1-404-446-6336 MCI: TNIXON Telemail: T.NIXON/HAYES AT&T: !tnixon UUCP: ...!uunet!hayes!tnixon Internet: hayes!tnixon@uunet.uu.net MHS: C=US / AD=ATTMAIL / PN=TOBY_L_NIXON / DD=TNIXON -----------------------------------------------------------------------------
chris@alderan.uucp (Christoph Splittgerber) (09/08/90)
In article <887@stsim.ocs.com> fsb@stsim.ocs.com (Steve Brailsford) writes: > [Text deleted] >need some advice on what kind to get and what to look for. I seem to >see that the V.32 seems to be a general standard for that rate, is this >true? Should I bother with any of the other standards that might also >come with that? I also hear about a 19.2K modem called Trailblazer, [Text deleted] While the V.32 is the more general standard for that rates, the Trailblazers seem to become (to be) the de facto standard in the uucp-world. The Trailblazers are quite a bit faster, especially when you transmit compressed data (e.g. netnews) because the MNP-5 compression (used with V.32) is quite useless with precompressed data or might even expand it. If you worry about standards, I think the "Telebit Trailblazer T2500" is a good solution Steve, because it supports them "all": PEP, V.32, V.25, V.22bis, V.22, V.21, V.23, Bell-103, Bell-212. Christoph
m1comm01@spacm1.uucp (Bob Straschewski) (09/17/90)
In article <887@stsim.ocs.com>, fsb@stsim.ocs.com (Steve Brailsford) writes: > I don't know if this is anything like what is being discussed, I just > picked up this group, but I am looking into getting a 9600 modem and > need some advice on what kind to get and what to look for. I seem to > see that the V.32 seems to be a general standard for that rate, is this > true? Should I bother with any of the other standards that might also > come with that? I also hear about a 19.2K modem called Trailblazer, > or is that some kind of protocol? I have seen some articles about how > you can get 19.2K from 9600 V.32bis modems. Do these modems require > leased lines, or is an ordinary phone line ok? Thanks for any help > you can give me here. > > Steve Brailsford >>> uunet!oscmd!stsim!fsb For dialup support from home for our rpogramming staff, we are using a bunch of MICROCOM 9624Cs. These babys do "fake" 9600, and work really well. Very tough too. Bob Ski..................
rah@public.BTR.COM (Richard A Hyde rah@btr.com) (09/18/90)
>In article <887@stsim.ocs.com>, fsb@stsim.ocs.com (Steve Brailsford) writes: >> I don't know if this is anything like what is being discussed, I just >> picked up this group, but I am looking into getting a 9600 modem and >> need some advice on what kind to get and what to look for. I seem to >> see that the V.32 seems to be a general standard for that rate, is this >> true? Should I bother with any of the other standards that might also >> come with that? I also hear about a 19.2K modem called Trailblazer, >> or is that some kind of protocol? I have seen some articles about how >> you can get 19.2K from 9600 V.32bis modems. I feel that it is very important to select a V.32 COMPLIANT modem if you are considering purchasing a 9.6 modem. COMPLIANT means IS, compatible means is LIKE. There is a world if difference. V.32 compliant modems will talk to other manufacturers V.32 compliant modems. Compatible modems will only talk to modems made by the same manufacturer. The major networks (commercial ones, anyway) are standardizing on V.32 with various levels of MNP. Accunet (whom I don't work for), is installing V.32 MNP 9 modems at all of their 9.6 nodes. Modems like the Telebit T2500 (which I like) will communicate with other V.32 compliant modems at 9600 bps. They will achieve 19.2 only with other Telebit modems. V.42 is supposed to create >9.6 speeds (using compression) among different manufacturers modems, but it doesn't appear to be reality quite yet. I use a Racal-Vadic 9632VP modem which also seems fairly robust, but I under- stand it has problems in MNP mode. Multi-Tech Multi modems also seem clean and easy to operate. Since I manage a datacom tech support team, this translates to "few complaints". You WILL have to read the manual of whatever modem you decide to buy, and get some help in configuring it to your needs. Factory configuration never quite seems to work... When you buy a modem, drop me a line if you have any questions... -Rick- -- ------------------------------------------------------------------------------ | Richard Hyde | RaH@btr.com | I don't need a disclaimer | ------------------------------------------------------------------------------