Once apon a time Danny Keogh said, 'Re: Baud and bps' to Kris Roebuck...

 KR> Could someone explain to me the difference between baud rates and
 KR> bps.  I'm told that they have different meanings, but no-one will tell
 KR> me why.

 DK> There really isn't a difference.  Baud is actually the name of
 DK> some guy who was in the datacoms field a while ago.  BPS stands
 DK> for bits per second.  Baud rate is just another way of saying BPS
 DK> rate.

Oh dear, Baud vs BPS rears its ugly head again :)

Let me state emphatically BPS (bits per second) and Baud are *NOT* the
same thing. They may under some circumstances have the same value, but
under other circumstances they are related by a factor, eg 2, 4 etc...

BPS or bits per second, is a data-related term that means exactly what it
says - 14400 bps means that 14400 bits of data will be sent in one second.

Baud (named after some guy whos last name was Baud :) is a modulation-related
term that describes how many signal transitions per second are being sent
down some communication path.

The factor that connets the two is dependant on how many states there are
in the modulation scheme. If there are two possible states that the signal
can be in at one instant in time, then you can describe a 0 or a 1 for
each instant in time. Here bps=baud. However, if you devise a system where
there are 4 valid states that the signal can be in at any instant, then
each state describes either 00, 01, 10, or 11. (Since 2^2 =4) Here bps=2*baud.

Since Baud is the number of _signal_ transitions per second, it is related
to the bandwidth of the signal, and like all transmition media, phone lines
have a limited bandwith, something like 3000hz +/- 500hz.

The neat trick is that by comming up with cleverer and cleverer modulation
techniques, you can increase that bps vs baud factor considerably, and
get faster data rates over the same bandwidth - which is what has made
14.4K and 28.8K modems possible.

The pay off is that as you get more and mode possible signal states, it
gets harder and harder to tell adjacent states apart reliably when random
noise is taken into account, which translates into needing a better signal to
noise ratio, for the more advanced techniques, eg V32bis, V34 etc...in
simple terms you need a nice quiet phone line without hiss and crackle :)

Hows that?

Regards,
Simon

... HAL 9000: Dave. Put down those Windows disks. Dave. DAVE!
--- FMail/386 1.02