karl@ddsw1.UUCP (Karl Denninger) (02/05/88)
To those of you who are thinking of buying a high-speed modem: Get a Trailblazer. Even if you have to pay full price for it. After what I have just seen, I am convinced. We've been running the TB+ at 9600 baud, mainly because the 'autobaud' setting wouldn't answer at 19.2. On selected outdial calls we used 19.2 though, with good success. Then I had an idea. Lock interface speed at 19.2k, turn on CTS flow control (Thanks SCO... Xenix is wonderful too)...... and I got a nice surprise -- ANY system I call connects flawlessly, with an interface speed of 19.2k! Those who run slower speed modems just get speed-converted by the modem. It also appears to work for dial-in callers, although I haven't rigorously tested this yet with uploads and the like. (Sure looks good though.) Oh yeah, let's not forget that it's already cut our on-line connect time by more than 70% (!). Our phone bill thanks you....... (AT&T probably wants you guys dead about right now :-) Mandatory disclaimer: We have no connection with Telebit except as a *very* satisfied customer. (Keep up the good work guys - 38.4 is next ;-) ---- Karl Denninger | Data: +1 312 566-8912 Macro Computer Solutions, Inc. | Voice: +1 312 566-8910 ...ihnp4!ddsw1!karl | "Quality solutions for work or play"
grr@cbmvax.UUCP (George Robbins) (02/07/88)
In article <626@ddsw1.UUCP> karl@ddsw1.UUCP (Karl Denninger) writes: > To those of you who are thinking of buying a high-speed modem: > > Get a Trailblazer. > Even if you have to pay full price for it. > > After what I have just seen, I am convinced. We've been running the TB+ at > 9600 baud, mainly because the 'autobaud' setting wouldn't answer at 19.2. > On selected outdial calls we used 19.2 though, with good success. Setting S51 to 254 enables 19200 as the default inteface speed for dial-in. This isn't documented in the manuals they're currently sending out with the modems, but apparently it's official. Also the modems *will* auto-baud to 19200, but you have to send the AT string as A\dT or your local equivalent of sending a delay. If the modem has already "autobauded" to some other rate, you may need to send a break before the AT to force the modem to re-autobaud. Sending a sequence of break, delay, A, delay, T delay, return before your actual ATTD5551212 sequence should be pretty universal, but no wagers accepted. > Then I had an idea. Lock interface speed at 19.2k, turn on CTS flow > control (Thanks SCO... Xenix is wonderful too)...... and I got a nice > surprise -- ANY system I call connects flawlessly, with an interface speed > of 19.2k! Those who run slower speed modems just get speed-converted by the > modem. I'm a little leery of this. It's fine for uucp only connections, but may have undesired side effects for normal human dial-ins. Damn the humans! Full baud ahead! 8-) -- 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)
dbuckler@hpihoah.HP.COM (Dan Buckler) (02/08/88)
karl@ddsw1.UUCP (Karl Denninger) writes: >> To those of you who are thinking of buying a high-speed modem: >> >> Get a Trailblazer. >> Even if you have to pay full price for it. >> >> After what I have just seen, I am convinced. We've been running the TB+ at >> 9600 baud, mainly because the 'autobaud' setting wouldn't answer at 19.2. >> On selected outdial calls we used 19.2 though, with good success. >> >> Then I had an idea. Lock interface speed at 19.2k, turn on CTS flow >> control (Thanks SCO... Xenix is wonderful too)...... and I got a nice >> surprise -- ANY system I call connects flawlessly, with an interface speed >> of 19.2k! Those who run slower speed modems just get speed-converted by the >> modem. >> >> It also appears to work for dial-in callers, although I haven't rigorously >> tested this yet with uploads and the like. (Sure looks good though.) >> ... >> >> ---- >> Karl Denninger | Data: +1 312 566-8912 >> Macro Computer Solutions, Inc. | Voice: +1 312 566-8910 >> ...ihnp4!ddsw1!karl | "Quality solutions for work or play" >> ---------- I feel compelled to point out that the Courier HST also allows connection to the host at 19.2 Kbaud, and the modem will convert differing incoming data rates to 19.2 automatically. To enable this feature one need only set a fixed DTE speed. BTW: I recently got a Courier HST for about $620.00. It supports MNP sevices 1-6, and can make a connection to another HST at 17.6K baud. Dan Buckler Hewlett Packard (408)338-3195
csg@pyramid.pyramid.com (Carl S. Gutekunst) (02/09/88)
In article <3276@cbmvax.UUCP> grr@cbmvax.UUCP (George Robbins) writes: >Also the modems *will* auto-baud to 19200, but you have to send the AT >string as A\dT or your local equivalent of sending a delay. The correct way to autobaud a TrailBlazer is to repeatedly send 'A' until the 'A' echos. Then send the 'T'. Alternatively send A\dA\dA\dA\dAT. <csg>
grr@cbmvax.UUCP (George Robbins) (02/09/88)
In article <14591@pyramid.pyramid.com> csg@pyramid.UUCP (Carl S. Gutekunst) writes: > In article <3276@cbmvax.UUCP> grr@cbmvax.UUCP (George Robbins) writes: > >Also the modems *will* auto-baud to 19200, but you have to send the AT > >string as A\dT or your local equivalent of sending a delay. > > The correct way to autobaud a TrailBlazer is to repeatedly send 'A' until the > 'A' echos. Then send the 'T'. Alternatively send A\dA\dA\dA\dAT. Yes, but the key seem to be that there must be some delay between the A's. In my testing just blasting out a stream of A's a 19200 baud didn't work, at least for cases like AAAAAAAAAAAAAAAAAAAT. Close reading of the manual also indicates that if the modem has already auto-bauded to some other speed, you will have to send it a break, or otherwise cause a framing error to put it in a mode where will try to auto-baud again. This is unlike most hayes style modems, which generally try to sense the baud rate any time they are waiting for an AT sequence. -- 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)
wtm@neoucom.UUCP (Bill Mayhew) (02/12/88)
It is true that both the Telebit Trailblazer and the Courier HST provide up 19.2K bps host interfaces. What sets the Trailblazer apart from the Courier HST is that the Trailblazer is able to rcognize the uucp "g" protocol. As such, the Trailblazer is able to locally generate the packet ACKs, and thus the need for a reverse channel is eliminated. The trellis coding used in the Trailblazer is also different, in that the fallback on poor lines is incremental in much smaller bps steps than the HST. Over the next few weeks, I plan to wring-out both the Telebit Trailblazer and the HST. In fact, the I have the following modems on the desk next to me: AT&T 2224 CEO 2400 bps / MNP 2:1 data compression AT&T Dataphone II 2296 2-wire 9600 BPS V.32 Microcom AX/9624c 2400 bps / high speed Racal Vadic 9600-VP 2400 bps / high speed Telebit Trailblazer + 2400 bps / high speed UDS model v.32 2-wire 9600 BPS V.32 US Robotics Courier HST 2400 bps / high speed If I find out anything interesting in this modem war, I'll post the results. My early impression is that the Telebit is the way to go because it integrates so nicely with uucp. If you are using a long block protocal such as hyperaccess, the other highspeed modems ought to be very serious competition, and then you can shop for price. Visual inspection reveals that the quality of construciton is ranked as follows: 1. Trailblazer 2. AT&T 2224 CEO 3. UDS v.32 4. Microcom AX/9624c 5. Courier HST 6. Racal 9600VP 7. AT&T Dataphone II (this appears to be a prototype. It is a very low serial number.) The Racal and Dataphone II had quite a few jumper wires that seemed to be fixes for last minute changes. -Bill
henry@utzoo.uucp (Henry Spencer) (02/16/88)
> This is unlike most hayes style modems, which generally try to sense > the baud rate any time they are waiting for an AT sequence. That is, in fact, the whole reason why everything has to be prefixed with the silly AT sequence! I was disappointed to see that Telebit didn't get the autobauding right. It's not a big deal, but it's annoying. -- Those who do not understand Unix are | Henry Spencer @ U of Toronto Zoology condemned to reinvent it, poorly. | {allegra,ihnp4,decvax,utai}!utzoo!henry
jeff@tc.fluke.COM (Jeff Stearns) (02/18/88)
In article <1004@neoucom.UUCP> wtm@neoucom.UUCP (Bill Mayhew) writes: > It is true that both the Telebit Trailblazer and the Courier HST > provide up 19.2K bps host interfaces. ... The trellis coding used > in the Trailblazer is also different, in that the fallback on poor > lines is incremental in much smaller bps steps than the HST. > > Over the next few weeks, I plan to wring-out both the Telebit > Trailblazer and the HST... I've installed a pair of Trailblazers and studied them fairly closely for several weeks. They are true champions when speaking PEP to one of their kind. HOWEVER, they are tepid performers at 1200 and 2400 baud. This shows up most dramatically when the line levels are low or noisy. Our UUCP connection to an overseas office just plain broke when we replaced our UDS modems with Trailblazers. Uucico would timeout and re-ack so badly that throughput was abysmal. Often the phone connection would break within the first minute. Days went by without a single successful file transfer. Immediately after restoring the UDS modem, we exchanged over a hundred files without a single hiccup. This isn't an isolated event. We use a cut-rate long-distance carrier to carry our calls to the Bay Area; the UDS modems could usually handle them adequately at both 1200 and 2400 baud. The Trailblazers simply cannot. Transmissions are so garbled that one can barely recognize that the destination computer is running at the proper baud! (I've experimented with other carriers; it's true that ATT offers superior service. The Trailblazers work well when the long-distance call is carried by ATT. Unfortunately, it's not a viable option for us right now.) In summary, I find that Trailblazers are superb modems when speaking PEP or communicating at low speeds over clean, quiet lines. However, they are poor performers at 1200 and 2400 baud over suboptimal connections. Final word: My Trailblazers have firmware rev "BA3.01". Telebit tells me that release 4.0 may improve performance at 2400 baud. -- Jeff Stearns Domain: jeff@tc.fluke.COM Voice: +1 206 356 5064 UUCP: {uw-beaver,decvax!microsof,ucbvax!lbl-csam,allegra,sun}!fluke!jeff Snail: John Fluke Mfg. Co. / P.O. Box C9090 / Everett WA 98206
karl@ddsw1.UUCP (Karl Denninger) (02/25/88)
In article <2905@fluke.COM> jeff@tc.fluke.COM (Jeff Stearns) writes: >HOWEVER, they are tepid performers at 1200 and 2400 baud. This shows up >most dramatically when the line levels are low or noisy. > >This isn't an isolated event. We use a cut-rate long-distance carrier to carry >our calls to the Bay Area; the UDS modems could usually handle them adequately >at both 1200 and 2400 baud. The Trailblazers simply cannot. Transmissions are >so garbled that one can barely recognize that the destination computer is >running at the proper baud! I can't agree here. We have the BA4.00 units, brand new, and they are flawless -- on noisy lines or quiet, at any baud rate. Our lines here are *horrid* -- loud audible noise on voice calls, crosstalk, etc. Before the Telebit outgoing calls at *any* baud rate were a 50-50 proposition as far as getting a usable connection goes. Not any more -- I *never* have had a "bad" line with the Trailblazer. When it's not in use and I have to dial out guess which modem I request..... In fact, they make our 2400 baud MultiTech 224AH look like a toy -- something I didn't believe myself. The MT224 has proven itself to work on both bad and good lines -- lines which make our USR Courier 2400's barf, spit, and disconnect. Our Telebit runs fine over lines that even the MultiTech can't handle, and at low baud rates too (300-2400). Perhaps you *do* need that PROM upgrade. ----- Karl Denninger | Data: +1 312 566-8912 Macro Computer Solutions, Inc. | Voice: +1 312 566-8910 ...ihnp4!ddsw1!karl | "Quality solutions for work or play"
jerry@oliveb.olivetti.com (Jerry Aguirre) (02/27/88)
In article <824@ddsw1.UUCP> karl@ddsw1.UUCP (Karl Denninger) writes: >In article <2905@fluke.COM> jeff@tc.fluke.COM (Jeff Stearns) writes: >>HOWEVER, they are tepid performers at 1200 and 2400 baud. This shows up >>most dramatically when the line levels are low or noisy. > >I can't agree here. > >We have the BA4.00 units, brand new, and they are flawless -- on noisy lines >or quiet, at any baud rate. Our lines here are *horrid* -- loud audible I suspect you guys are talking apples and oranges. There was a review (<7164@pur-ee.UUCP> in comp.dcom.modems by George Goble <ghg@pur-ee.UUCP> on the net a while back that made note of the differences between the TrailBlazer and the TrailBlazer Plus. Apparently there was extensive redesign of the hardware. Two features mentioned were EMI and a "new analog front end". The EMI was improved so that the modem didn't drive every nearby radio crazy. The old TB required a choke in the phone cord to filter out noise before it got into the phone wiring. The old TB also had a fan. The TB+ doesn't require that filter and has a standard modular cord. It also doesn't have a fan. The reviewer also noted that the new analog front end improved the performance especially on bad lines. So, when you complain you should give your model and firmware revision so people can make some sense of what you are saying. Jerry Aguirre
pjh@mccc.UUCP (Peter J. Holsberg) (02/28/88)
The phone lines between my home and office are so bad that it's
frequently impossible to login on the office machine, what with all the
little ~~~~~~ characters and their brethren. However, a Trailblazer on
the same line never shows a single line noise character! And that's
with the V3 ROM!
--
Peter Holsberg UUCP: {rutgers!}princeton!mccc!pjh
Technology Division CompuServe: 70240,334
Mercer College GEnie: PJHOLSBERG
Trenton, NJ 08690 Voice: 1-609-586-4800