Path: utzoo!attcan!uunet!ogicse!unicorn!blake!milton!whit
From: whit@milton.acs.washington.edu (John Whitmore)
Newsgroups: sci.electronics
Subject: Re: low-level signals and long cables
Summary: Coax or twisted-pair is adequate if you transformer-couple
Message-ID: <2845@milton.acs.washington.edu>
Date: 13 Apr 90 05:38:04 GMT
References: <22845.2623bae8@kuhub.cc.ukans.edu>
Reply-To: whit@milton.acs.washington.edu (John Whitmore)
Organization: University of Washington, Seattle
Lines: 35

In article <22845.2623bae8@kuhub.cc.ukans.edu> rsl09@kuhub.cc.ukans.edu writes:
>
>I have an application that requires sending low level signal (in mv range)
>down a  long cable (about 50 meters). The signal is bandlimited to about
>13 KHz. To avoid potential interference, I thought it would better
>to  digitize the signal ...

	You do not mention the impedance of your signal source; this 
is VITALLY important to determine what, if any, interference you are
protecting against.
	Digitization is certainly not appropriate (unless you have
a large budget, or noisy wiring that you wish to use rather than
making a new wire path). 

	A simpler method: amplify the signal at one end, with 
a high-output-impedance amplifier (i.e. make a current source), and 
detect at the receive end with an appropriate resistor.  There is
a popular industrial standard, the "4-20mA transmitter", using this
trick.  It usually works well with low frequency signals.

	Or amplify to get a higher voltage signal, which you 
attenuate at the receive end.  Or use a transformer at one end
(or both) to break up the "ground loop".  It may suffice to use
coaxial cable, or you may wish to use twisted pair (the second
wire becomes the "sensitive ground" connection for the remote box),
or (in extreme cases) you may use two coaxial cables, one for the
signal, one for the ground reference.   
	The mere presence of digital circuitry in a box practically
guarantees that that box cannot maintain low noise for a small
signal, if only due to modulation of the power supply with the
digital clocking noise.  Avoiding switching circuitry and shielding
the cable can easily keep your 13 mV signal intact.

I am known for my brilliance,                  John Whitmore
 by those who do not know me well.