Precision circuit closely monitors –48V bus
Edited by Bill Travis
Paul Smith and Jim Staley, Analog Devices, Wilmington, MA -- EDN, October 17, 2002
Ever-greater complexity in communications systems has spurred a need for power-supply management. POTS (plain-old telephone systems) obtain power from –48V supplies backed by arrays of batteries in central offices and distributed throughout copper lines. Although nominally –48V, the voltage on the lines can vary from –40 to –80V, and the voltage is subject to surges and fluctuations. Supply regulation at the source has little effect on remote voltage levels. Equipment failures resulting from surges, brownouts, or other line faults may sometimes be undetectable. Capturing power-supply information from remote communications equipment requires precise measurement of the voltages—sometimes in outdoor temperatures. High-common-mode, voltage-difference amplifiers have been useful in monitoring current. However, you can also use these versatile components as voltage dividers, enabling remote monitoring of voltage levels as well.
Figure 1a shows the basic circuit connections when you use a difference amplifier for high-line current sensing. With the addition of a few parts, you can operate this amplifier in a negative-supply system. Figure 1b shows a precision monitor using just two ICs and deriving its power from the –48V supply. All you need to power the circuit are a transistor and a zener diode to reduce the supply voltage for the amplifiers. The AD629 shown in Figures 1a and 1b is a self-contained, high-common-mode, voltage-difference amplifier with unity gain. Connected as shown, however, it reduces the differential-input voltage by approximately 19, thus acting as a precision voltage divider. You need an additional amplifier for loop stability. The circuit features several advantages over alternative approaches. The laser-trimmed divider resistors exhibit essentially perfect matching and tracking over temperature. Linearity errors from –40 to –80V are nearly immeasurable. Figure 2 and Figure 3 show the linearity and temperature-drift curves for the monitoring circuit.
Is this the best Design Idea in this issue? Select at www.edn.com.
