whitcomb@aurs01.uucp (Jonathan Whitcomb) (02/01/91)
I am trying to understand the differences between HST and the V.32/V.32bis standards. I understand that HST does not have echo cancellation, and instead uses a low speed channel for full duplex operation. I also gather that HST can change the high speed channel from the caller to receiver by retraining. I am still fuzzy on the following: 1. Does HST high speed channel use V.32 modulation specs for constellation mapping and Trellis decode algorithms? 2. What are the specifics of the low speed channel? I assume that it is an FSK channel, and if so, what are the mark and space frequencies? Is it correct that the 9600 uses a 300 baud channel and the 12000 uses a 450 baud channel? Do both have the same mark and space frequencies? 3. Are there any other modem manufacturers that are implementing pseudo-CCITT implementations of V.32/V.32bis? It seems like folks are jumping through all kinds of hoops to avoid echo cancellation and Trellis code implementations. 4. Is the HST modem popular enough to worry about, or will it vanish once V.32/V.32bis catch on? Thanks. ********************************************************************** Jonathan Whitcomb UUCP: <...!mcnc!aurgate!whitcomb> (919) 850-6231 I'm not a software engineer, Raleigh, NC but I play one on TV.
roberson@aurs01.UUCP (Charles "Chip" Roberson) (02/01/91)
In article <59490@aurs01.UUCP> whitcomb@aurs01.uucp (Jonathan Whitcomb) writes: >4. Is the HST modem popular enough to worry about, or will it vanish >once V.32/V.32bis catch on? My understanding that the HST is *very* popular with the BBS types, particularly because of USRobotics sysop program. However, I've noticed that many sysops are speed freaks and as such buy the latest and greatest technology ASAP. I believe many are/will be moving to USR Dual Standard modems to get the higher transfer rates but retain the HST compatability. But this is only semi-knowledgable speculation on my part. cheers, -chip * Work: 2912 Wake Forest Road, Raleigh, NC 27609 (919) 850-5011 * (...!mcnc!aurgate!roberson) || (roberson%aurgate@mcnc.org) || * (71500.2056@CompuServe.com) || (Chip.Roberson@f112.n151.z1.fidonet.org) #include <disclaimer.h>
wallach@motcid.UUCP (Cliff H. Wallach) (02/02/91)
In article <59490@aurs01.UUCP> whitcomb@aurs01.uucp (Jonathan Whitcomb) writes:
-I am trying to understand the differences between HST and the
-V.32/V.32bis standards. I understand that HST does not have
-echo cancellation, and instead uses a low speed channel for
-full duplex operation. I also gather that HST can change the
-high speed channel from the caller to receiver by retraining.
-
-I am still fuzzy on the following:
-
-1. Does HST high speed channel use V.32 modulation specs for
-constellation mapping and Trellis decode algorithms?
Yes.
-
-2. What are the specifics of the low speed channel? I assume that
-it is an FSK channel, and if so, what are the mark and space frequencies?
-Is it correct that the 9600 uses a 300 baud channel and the 12000 uses
-a 450 baud channel? Do both have the same mark and space frequencies?
The low speed channel uses phase shift keying, with 3 bits/baud.
The earliest HST had a 300 bps backchannel. After a year the
450 bps backchannel was added to the 9600 HST.
backchannel 300 450
----------- --- ---
baud rate 100 150
carrier freq 350 375 (I may be off here)
-
-3. Are there any other modem manufacturers that are implementing
-pseudo-CCITT implementations of V.32/V.32bis?
Yes.
- It seems like folks
-are jumping through all kinds of hoops to avoid echo cancellation and
-Trellis code implementations.
The echo cancellation requires a lot of DSP time. This
translates into an extra DSP chip.
-
-4. Is the HST modem popular enough to worry about, or will it vanish
-once V.32/V.32bis catch on?
Over 100k are out there. Sales have steadily increased.
Only recently have V22bis modems surpassed Bell 212a modems
in volume.
-
-Thanks.
You are welcome.
Cliff Wallach ...uunet!motcid!wallachtnixon@hayes.uucp (02/03/91)
In article <59490@aurs01.UUCP>, whitcomb@aurs01.uucp (Jonathan Whitcomb) writes: > 1. Does HST high speed channel use V.32 modulation specs for > constellation mapping and Trellis decode algorithms? Yes. The HST modulation, like Hayes' fast-turnaround ping-pong modulation in the V-series Smartmodem 9600, uses the constellation and trellis coding of V.32 at 4800 and 9600; the HST uses the same constellation and trellis coding as V.32bis at 7200, 12000, and 14400 (in fact, the text describing these modulations that was used in the V.32bis standard was taken from the old "V.asym" draft asymmetrical modem standard, which had been submitted by US Robotics). > 2. What are the specifics of the low speed channel? I assume that > it is an FSK channel, and if so, what are the mark and space frequencies? > Is it correct that the 9600 uses a 300 baud channel and the 12000 uses > a 450 baud channel? Do both have the same mark and space frequencies? No, it is not FSK, but PSK. The 450bps reverse channel is 150 baud using 8 phase states to send 3 bits per baud. The carrier frequency is 375 +/- 1 Hz. The 9600 did use a 300bps reverse channel; I think it was 100 baud, 8 phase. > 3. Are there any other modem manufacturers that are implementing > pseudo-CCITT implementations of V.32/V.32bis? It seems like folks > are jumping through all kinds of hoops to avoid echo cancellation and > Trellis code implementations. As I mentioned above, the Hayes V-series Smartmodem 9600 uses a fast-turnaround ping-pong half-duplex modulation scheme that uses the same constellation and trellis coding as V.32. These two are the only non-standard high-speed modems I'm aware of that have even this in common with V.32. When these modems were designed, real V.32 modems cost $2,500+, due to the extreme high cost of implementing an echo canceller with technology that existed then. Since then, the cost of implementing echo cancellation has plummeted, and I doubt that you'll see many new modems being developed that are non-standard. The HST and V-series will continue to be sold for at least a while, particularly in "dual standard" implementations such as the Ultra 96, for customers who require backward compatibility with their installed base of non-standard modems. > 4. Is the HST modem popular enough to worry about, or will it vanish > once V.32/V.32bis catch on? "Vanish" is a strong word, but I suspect you'll see fewer and fewer non-standard high-speed modems of all types being sold in the future. It just makes so much more sense to stick with a real international standard. -- Toby Nixon, Principal Engineer | Voice +1-404-449-8791 Telex 151243420 Hayes Microcomputer Products Inc. | Fax +1-404-447-0178 CIS 70271,404 P.O. Box 105203 | UUCP uunet!hayes!tnixon AT&T !tnixon Atlanta, Georgia 30348 USA | Internet hayes!tnixon@uunet.uu.net