30 Dec
2007
30 Dec
'07
2:08 p.m.
"Michael B. Brutman" wrote:
This question is bugging me from another forum.
Is it possible to connect two modems (eg: Hayes 2400 to Hayes 2400)
using a 'dead' or isolated pair of copper wire and have them be able to
communicate?
I always thought that this can not work because the line that the modems
are connected to has to have some current. The phone system works
because an action on one end of the phone (talking into the carbon
microphone) causes a reaction on the other end. Without some sort of
current on the line, how can this work? Hence the need for 'line
simulator' circuits
Some people are claiming that it works without the line simulator. I'd
like to understand why.
(I could get out the multimeter if things get really desperate, but
maybe somebody can tell me that the line current thing only applies to
acoustically connected modems, not direct connect. Or the direct
connect modems put enough juice on the line to make it work. Or
something else that might make sense.)
Yes, it will work (or can, acknowledging the idiosyncrasies of some modems) (I
have used it on occasion to transfer between machines that had no other common
communication/IO facility).
You're on the right track about the modems putting juice on the line. The
original design of the phone system, as you suggest, was such that the carbon
mike modulated the DC line current to produce an AC component (the audio) on
the line, the carbon mike acting as a variable resistance and the DC line
current being the only source of energy on the line.
However, many directly-connected modems are transformer coupled onto the line
and actually inject their own AC energy onto the line, with no inherent need
for the DC line current.
On the other hand, some modems (generally later, higher-speed ones, I believe)
use opto-isolator-coupling onto the line and so work more like the carbon mike,
and will require the DC line current.
It's been awhile since I've done it, but the general idea is to simply connect
up the modems (one pair/two wires), set one into answer mode (ATA), and set
the other into originate mode (ATD or ATH1(?)), forcing one/both off-hook.