del@Data-IO.COM (Erik Lindberg) (04/22/88)
I am in the position of having to come up with a recommendation for a modem to Taiwan, R.O.C., and would like to use a 9600 baud modem if possible. The line noise to that little country is simply astonishing, and trying to relate specifications and test results to reality is difficult. On voice communications it is necessary to raise your voice to be heard above the background noise given the weak signal strength (faint voice). Currently, they (Taiwan) are using a SmarTeam 2400 baud modem to connect with my system (Kirkland, WA) using a Hayes 2400. Connections at 2400 baud can be made if the moon is full and the conjunction of Mars, Venus, and the Jupiter form an Isosceles triangle. :-) Connections at 1200 can usually be made given two or three attempts. I had previously thought to recommend a Trailblazer, which (theoretically) would fall back in 100 baud increments to get the maximum throughput on the line. I read in the recent PC Rag review, however, that their TB modem gave up entirely at -27db signal/noise ratio. This compares with the SmartTeam 2400 signal/noise ratio performance of about -28db at 2400 baud (doesn't work) and -15db at 1200 baud (works). These figures quoted from the April 26, 1988 and May 12, 1987 PC magazine issues. As you can see, the -27 db figure for the TB is dangerously close to the SmarTeam at 2400, which failes miserably. The USR HST is only marginally better than that (2db). The only modems that had sterling performance down to -19db s/n do not support 1200/2400 baud. That is fine for Taiwan since they can just switch modems, but not for me since I must support a dial up line and cannot justify a dedicated line just for the one caller. Any suggestions, anyone? -- del (Erik Lindberg) uw-beaver!tikal!pilchuck!del
pete@octopus.UUCP (Pete Holzmann) (04/22/88)
In article <872@pilchuck.Data-IO.COM> del@Data-IO.COM (Erik Lindberg) writes: > I had previously thought to recommend a Trailblazer, which (theoretically) >would fall back in 100 baud increments to get the maximum throughput on the >line. I read in the recent PC Rag review, however, that their TB modem gave >up entirely at -27db signal/noise ratio. This compares with the SmartTeam >2400 signal/noise ratio performance of about -28db at 2400 baud (doesn't >work) and -15db at 1200 baud (works). These figures quoted from the April 26, >1988 and May 12, 1987 PC magazine issues. >... Any suggestions, anyone? > >del (Erik Lindberg) >uw-beaver!tikal!pilchuck!del From the looks of it, PC mag used a tester box that doesn't match most noise found in real life. It isn't the S/N ratio that is wrong, it is the KIND of noise they used that is wrong. Telebit modems do just GREAT with most kinds of real-life noise, no matter how bad! I'd go ahead and get the TB. You might try calling Telebit to ask if they have experience with connections to Taiwan. If they don't, try getting a guaranteed refund if the modem simply won't work. Then you won't be stuck if for some reason the experiment fails. It never hurts to ask! Pete -- OOO __| ___ Peter Holzmann, Octopus Enterprises OOOOOOO___/ _______ USPS: 19611 La Mar Court, Cupertino, CA 95014 OOOOO \___/ UUCP: {hpda,pyramid}!octopus!pete ___| \_____ Phone: 408/996-7746
grr@cbmvax.UUCP (George Robbins) (04/22/88)
In article <872@pilchuck.Data-IO.COM> del@Data-IO.COM (Erik Lindberg) writes: > I am in the position of having to come up with a recommendation for a modem > to Taiwan, R.O.C., and would like to use a 9600 baud modem if possible. > I had previously thought to recommend a Trailblazer, which (theoretically) > would fall back in 100 baud increments to get the maximum throughput on the > line. I read in the recent PC Rag review, however, that their TB modem gave > up entirely at -27db signal/noise ratio. It's not clear that the testing done by PC magazine bears any relation to the kind of line conditions experienced in the real world. I'd suggest you call Telebit and tell them that you're intestested in evaluating their modems for use under some "difficult" conditions and see if you can negotiate a loan of a pair of modems or at least "return if they don't do what you need". Telebit users have reported success in just this kind of situation. Beware though of foriegn laws that may make it illegal to attach a piece of non-approved telecomm equipment to that countries phone system. Have the people on the far end evaluate concerns in this respect. > That is fine for Taiwan since they > can just switch modems, but not for me since I must support a dial up line > and cannot justify a dedicated line just for the one caller. This might be a false economy. Around here the monthly charges for a business phone line used only for incoming calls are still quite cheap and would be tiny compared to the international long distance charges this party would accumulate calling you, especially if conditions are sub-optimal, resulting in wasted and prolonged calls. The alternative is, of course, an X.25 packet switched connection. In countries where the PTT pushes packet switched newworks, the cost of a local connection may be considerably cheaper than what you'd expect from US vendors. Use of something like Tymenet's "Asynchronous Outdial" service on this end may be cheaper than the traditional leased line to the PAD or local PAD approach for low volume connections. Throughput might not be so hot, but reliability would probably be much better than a typically lousy international connection. Obviously, much depends on why you are talking to Taiwan, and who is picking up the bills on either end.... -- George Robbins - now working for, uucp: {uunet|ihnp4|rutgers}!cbmvax!grr but no way officially representing arpa: cbmvax!grr@uunet.uu.net Commodore, Engineering Department fone: 215-431-9255 (only by moonlite)
donegan@stanton.TCC.COM (Steven P. Donegan) (04/24/88)
In article <872@pilchuck.Data-IO.COM>, del@Data-IO.COM (Erik Lindberg) writes: > I am in the position of having to come up with a recommendation for a modem > to Taiwan, R.O.C., and would like to use a 9600 baud modem if possible. > The line noise to that little country is simply astonishing, and trying to > relate specifications and test results to reality is difficult. On voice > communications it is necessary to raise your voice to be heard above the > background noise given the weak signal strength (faint voice). I recommend that you use the Trailblazer. We do to Singapore and even with the satelite delays, local loop noise, total carrier drop-outs and generally really sh%^^y line quality the Trailblazer works flawlessly. I don't see any real comparison between a Trailblazer and a non-PEP modem in the sig-to-noise arena - the Trailblazer can be set so that almost no line screwups will cause it to drop the line - it will just work around the problems or wait until things get good enough to send data again. No other dial-up modems we have tried in the Singapore to US dial-up environment have been able to hold the line for more than a few minutes; the Trailblazer has held the same connection without a single failure yet (about 3 weeks of 8hrs/day error free on-line time) -- Steven P. Donegan Sr. Telecommunications Analyst Western Digital Corp. donegan@stanton.TCC.COM
wtm@neoucom.UUCP (Bill Mayhew) (04/26/88)
Rick Adams at seismo reported connections of up to around 11 kbaud on really awful lines to sites in the Andes mountains. The circuits there probably aren't any better than what you'd get going to Taiwan. I've done things like shouting into the phone or dailing DTMF tones while the T'blazer is talking and it just keeps right on going. If you really make a lot of racket, it will stop and retrain. I've never seen a T'blazer break a PEP mode connection due to poor lines, and there are some lines that belong up in the Andes here in lovely Rootstown! --Bill
wtm@neoucom.UUCP (Bill Mayhew) (04/26/88)
I was just wondering if anyone out there has tested T'blazer modems with mulitple satelite hops, such as might be experienced going to Taiwan. Chances are that it would take at least two hops, for close to 100,000 miles traveled, meaning a turnaround delay of 0.75 to 1.0 seconds. From my understaning, the TelCo usually routes overseas calles so that 1/2 goes by cable, while the other half of the duplex goes by satelite. This keeps the delay from getting too out of hand. Of course, one just might get cable both ways, or even satelite both ways depending on luck of the draw. --Bill
phread@ih1ap.ATT.COM (J.F. Shumway) (04/29/88)
Taiwan's noise problems are in their local wire plant, that final mile from the switch to the wall jack. In general, local loop maintenance is nowhere near the quality one is used to getting in the US of A, unless of course you can bring some clout to bear, you're a high official in the telephony bureaucracy, or, your 2nd cousin works in the installation and maintenance group at your local phone office. On the other hand Taiwan's toll network is well maintained, and they've been investing heavily in state of the art local CO switching equipment. Their noise problems are in the local wire plant. If you ever get the chance to visit, which I heartily recommend, talk a stroll through the older sections of Taipei and notice the power and phone distribution technology (or lack thereof). It will remind you of a Phyllis Diller hairdo. I've heard, that the telephone authorities will sometimes just keep assigning you a new local loop, from the pole to the CO, until they find one that works! I've seen uucp throughput figures ranging from 700 bytes/second using a TB+ with an EIA speed of 9600 baud (not bad for a satellite link) to only 30-150 bytes/second using Microcom 9624Cs through a local circuit that was so noisy that voice com was often impossible. Some in the Taiwan telephone authority suggested increasing Microcom's transmit power from the US default of -12 dBm to 0 dBm. This, not suprisingly, can't be done with this model Microcom and most other reasonably priced modems built to FCC (USA) specs. And in any case this strategy only helps get half of your data above the noise floor. The Taiwan telephone authority (DGT) offers DCE or async PAD access to their island wide X.25 packet network which can feed to Tymnet and Telenet here in the US, among others. This service is not cheap and I can't vouch for its quality. The only available description of this service is written in Mandarin, which I don't even pretend to understand. Although I can't tell you more it may be worth investigating. Good Luck, dcom in Taiwan is a matter of luck. -- Jesse Fred Shumway ihnp4!ih1ap!phread
donegan@stanton.TCC.COM (Steven P. Donegan) (04/29/88)
In article <1140@neoucom.UUCP>, wtm@neoucom.UUCP (Bill Mayhew) writes: > I was just wondering if anyone out there has tested T'blazer modems > with mulitple satelite hops, such as might be experienced going to > Taiwan. Chances are that it would take at least two hops, for > --Bill It's a winner...the Trailblazer works flawlessly over incredibly sh&**y lines both sattelite (definite delay and echo) and the other luck of the draw connections one gets over the dial-up network (re-route from Ca thru NJ to Singapore etc.). I can't recommend the Trailblazer enough (although I can't seem to get it to work with my brain damaged uucp reliably - software prob). Another happy customer... -- Steven P. Donegan Sr. Telecommunications Analyst Western Digital Corp. donegan@stanton.TCC.COM
wtm@neoucom.UUCP (Bill Mayhew) (05/02/88)
One nice feature about the TB+ is that it has a programming resistor to set the output level for leased line applications. I suppose this could be done in Taiwan eventhough the modem is attached to a dial-up line. I don't think the programming resistor is accessible on the older "standard" model Telebits. --Bill