
Remote voltage sensing usually uses a four-wire sensing system, which involves two conductors for high load-current flow and two high-impedance paths for accurately measuring the remote voltage without incurring substantial ohmic losses. Voltage control at the load, therefore, entails measuring the "sense" pair and driving the "output" pair. Over cables in which the impedances in the source and return paths to the load are equal, you can use a three-wire sense system, thereby saving a conductor.
The circuit in Fig 1 takes advantage of equal resistance in the source and return paths. It uses a simple network to sense the remote voltage and drive the output to precisely control the remote voltage. In the output network, RL represents the resistance of a single cable conductor, for example, 4(ohm) for 100 ft of #26 wire. VO is the output voltage to the cable, with reference to the return lead (GND). VS is the voltage of the sense lead, with reference to the return lead (GND).
VL is the desired voltage at the load, and IO is the current flowing through the cable and load. IS is the current flowing through the sense lead, designed such that IS<O. ZL is the load impedance. Assuming all equations have GND as reference, if VS IORL+IOZL, VLIOZL, and VO2IORL+IOZL, then VO2VS-VL. By setting the desired voltage setpoint, VLSETPOINT, using a manual adjustment or control input, you can control VO to make VLVLSETPOINT.
You may have to attenuate the sense voltage and control voltage with added gain in the output control loop to maintain an acceptable common-mode range for the processed signals in the amplifiers (Fig 2). Be careful to maintain high impedance in the sense circuit. Insofar as the regulating supply voltage has enough compliance, the circuit maintains proper operation with changes (such as those caused by heating) in the cable's conductor resistance. (DI #1706)