Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!apple!netcom!gandrews
From: gandrews@netcom.COM (Greg Andrews)
Newsgroups: comp.dcom.modems
Subject: Re: USR Courier V.32bis questions etc....
Summary: Answers to Telebit T2500 questions
Message-ID: <1991Apr21.032236.21724@netcom.COM>
Date: 21 Apr 91 03:22:36 GMT
References: <6465@husc6.harvard.edu>
Organization: Netcom - Online Communication Services  UNIX System {408 241-9760 guest}
Lines: 99

In article <6465@husc6.harvard.edu> conrad@popvax.harvard.edu writes:
>
>The V.32 upgrade for the T2500 was so late in becoming available that I
>received a free V.42 and V.42bis upgrade with it 8-(  8-) !
>

No such thing as a "V.32 upgrade" for a T2500.  Every single version of
firmware for the T2500 has supported V.32.  The very first version didn't
support *synchronous* V.32 operation, but the numbers of modems used for
asynchronous links far exceed the numbers used for synchronous links.  I
suspect you use async connections (revealed by your 19200/38400 bps questions
in a later paragraph), so the T2500 had the V.32 support you needed all the
time.

>
>     So finally the pair of modems is relatively happy.  (Actually, I was never
>able to set up my software with DTE rates of 38,400 bps on both ends.  I heard
>something about T2500s not supporting DTE rates of higher than 19,200 bps.  (I
>don't have the manual here -- I might even have read it there...?)  Is this at
>all true?  It is also possibly a software problem....)
>

The T2500 supports RS232 speeds up to 19200 bps.  38400 bps requires a faster
processor than the T2500 has.

>
>     Now the technical questions.  First, I was a bit confused about something
>I read in MacWEEK 03.19.91 V5N11 in an article about V.32bis modems on pages 45
>and 46.  There seems to be an implication that one CAN NOT use data compression
>with the new 14,400 bps link rate.  I haven't seen anything real to explain
>this.  Is it true?  If so why?  I HOPE that one can use V.42bis compression on
>top of the V.32bis 14,400 bps link rate....
>

There's absolutely no reason a modem can't apply an error control protocol
like MNP or V.42 to a V.32bis link.  Given that, there's absolutely no reason
a modem can't apply a data compression algorithm like MNP5 or V.42bis to the
data.  Provided the RS232 speed is faster than the modulation speed and the
data can be compressed by the modem, then you can certainly get throughputs
faster than 14400 bps from the V.32bis link.

>
>If I am right (and the article is either wrong or VERY poorly written),
>what is the highest throughput that could be hoped for?  14,400 X 4 = 57,600.
>I know that this would be difficult to sustain on real data, but is it a good
>number for a top limit?
>     If these modems can theoretically reach throughputs of 57,600 bps, what
>is the highest DTE rate that they support?  I should think 57,600 bps would be
>necessary.  
>

Why should it be necessary if, as you say, you wouldn't be able to get that
much compression out of real data?  How many computers can handle a sustained
data rate of more than 38400 bps?  What kind of processor hardware would be 
necessary for the modem to handle a 14400 bps (full) duplex modulation AND
compression/decompression of the data AND an RS232 interface speed of 57600?
Would you really be willing to give up the amount of money and desk space for
such a monster?  Especially if you know in advance that you won't utilize the
speed capabilities beyond 38400 bps more than perhaps 2% of the time?

I think the reason most modem manufacturers are creating modems with a max
RS232 speed of 38400 are doing it for three reasons:  (a) It doesn't require
exotic, costly processors,  (b) Real life data transfers won't be more than
2.7:1 compressible 98% of the time,  and (c) Few computers can handle data
rates above 38400 bps.  In short, they're designing a modem that will meet
your data transfer needs without exceeding your budget needs.  

>
>(Another dealer told me yesterday that by virtue of the 68000 in the T2500, 
>a V.32bis upgrade would be just a matter of software.  I am especially
>skeptical of this as I had to pull and replace chips to make the thing talk
>V.32 and V.42 and V.42bis....  The truth...?)
>

How do you think a software upgrade is applied to a modem?  The modem functions
are in software stored inside chips - that's why it's called FIRMware.

You saw the innards of your T2500.  Did you notice the little piggyback
board?  That's a Rockwell module.  (I also see that your firmware upgrade
was for V.42 and not V.32, as you mentioned earlier)

No, I'm afraid adding V.32bis to the T2500 is not a matter of just a firmware 
upgrade no matter what any salesperson says.  Telebit originally intended that 
the TrailBlazer Plus have the ability to do everything in DSP so new stuff
would be merely a firmware upgrade, and advertised the modems that way.  Alas, 
adding V.32 proved to be more than the TB+ modem's processors could do, so 
getting V.32 fully in DSP required a new platform of faster DSP processor and 
faster general-purpose processor.  The T1600 is the first modem that uses the 
faster processors, and it does not use a Rockwell module either (take that as 
a sign of how Telebit engineers feel about the Rockwell module).  The T2500 
used the Rockwell module in order to provide V.32 capability at a time when
Telebit customers needed it.


-- 
.------------------------------------------------------------------------.
|  Greg Andrews   |       UUCP: {apple,amdahl,claris}!netcom!gandrews    |
|                 |   Internet: gandrews@netcom.COM                      |
`------------------------------------------------------------------------'