Date: Wed, 20 Mar 96 10:58:10 MET From: Greg Lehey <lehey.pad@sni.de> To: phk@critter.tfs.com (Poul-Henning Kamp) Cc: hackers@freebsd.org (FreeBSD hackers), isdn@muc.ditec.de (Distribution List; FreeBSD ISDN) Subject: Re: ISDN: "modem" or board? (Was: Microsoft "Get ISDN"?) Message-ID: <199603201000.LAA14668@nixpbe.pdb.sni.de> In-Reply-To: <11362.827304930@critter.tfs.com>; from "Poul-Henning Kamp" at Mar 20, 96 6:55 am
next in thread | previous in thread | raw e-mail | index | archive | help
>>> Watch my lips. A B channel gives you 64,000 bits per second. When you
>>> add one start bit and one stop bit for every 8 bits, you end up with
>>> 80,000 bits. Bind two such channels together and you have 160,000
>>
>> Your lips don't make any sense.. :-) You SUBTRACT the start bit and
>> stop bit, you don't ADD it dude!
>
> Actually, you can subtract them in some cases.
I don't understand why people have so much difficulty understanding
this point. Let's try a picture:
DTE "modem" "modem" DTE
------ ------ ----- ------
| | async | | ISDN | | async | |
| |--------->| |--------->| |--------->| |
| | 10 bits | | 8 bits | | 10 bits | |
------ 80 kb/s ------ 64 kb/s ----- 80 kb/s ------
The B channels run at a speed of 64000 bits/second. Nothing changes
that. The 2 bits overhead on the async line are purely timing
information generated and swallowed by the UART. From left to right,
units of 10 bits are processed by the UART in the first "modem",
producing 8 bits of output. These are transmitted with no further
encoding (except the TDM implicit in ISDN BRI) to the other "modem",
which puts them into a UART, which then surrounds them with a start
and a stop bit and puts them out on the async line.
Admittedly, this view is simplistic to a degree: without some kind of
protocol around the data, you can't tell what's data and what's empty
noise. Obviously, though, the lower the overhead, the better.
Typically, HDLC is used.
> My Zyxel 2864I ISDN-gadgets can speak async HDLC on the serial port and
> sync hdlc on the ISDN.
>
> I've tried it out and can confirm that "64000 < one B-chan < 80000".
> My figures indicated 75000 bits/sec for FTP.
I assume your 75 kb/s are on the async lines.
> The principle is not entirely unlike the various UUCP/XMODEM/KERMIT spoofing
> done by other brands of modems.
I don't understand this statement.
Greg
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199603201000.LAA14668>