Question about modems

allison allisonportable at
Wed Nov 13 18:02:13 CST 2019

On 11/13/19 4:16 PM, Grant Taylor via cctalk wrote:
> On 11/13/19 1:31 PM, Fred Cisin via cctalk wrote:
>> But, stuff like commands to the modem didn't need much of that, and
>> needed to be able to communicate in spite of wrong parameters.  It
>> made sense for a modem to recognize a command, even with wrong parity,
>> etc.
> Okay....
> Now I'm thinking that there are really two phases / modes of
> communications:  1) computer to modem commands, and 2) computer to
> computer via modem connection data.
> I think my previous statement applies to #2.  I can see the value in #1
> being more liberal in what it recognizes and accepts.
> But, I'd still be surprised if the following would work for #2.
> [A]---(7E2)---{modem}==={modem}---(8N1)---[B]
> Would A and B be able to transfer data between each other with different
> (local) settings?

The case presented is an true corner case.  Usually modems can be set
for rate, bits, and parity that matches the target system.  AS modem
became more sophisticated (26/33/56k) they were easily set and I just a
few I have.  Older 300/1200/2400 baud modems like DEC DF02/03 had
switch settings likely others did too.  The 9600 were generally smart.

Those that used AT command set were very setable as there was usually
large menu of options.

There were exceptions "winmodems" that were only hardware port
and the system did all the work in software.  Generally flaky
to the max.  DEC DFxx series had their own non AT command set.

What is forgotten is that many of the modem programs could swich
parameters on the fly after carrier detect went true.

Like RS232/432 signaling the art and science is getting lost
to time and age.  I'd bet more than half here never used a
modem or have not in more than 15-20 years.


More information about the cctech mailing list