Design Note 70: A Broadband Random Noise Generator

Filter, audio and RF communication testing often requires a random noise source. The circuit in Figure 1 provides an RMS amplitude regulated noise source with selectable bandwidth. The RMS output is 300mV with a 1kHz to 5MHz bandwidth selected in decade ranges.

Figure 1. Random Noise Generator with Selectable Bandwidth and RMS Voltage Regulation.

Figure 1. Random Noise Generator with Selectable Bandwidth and RMS Voltage Regulation.

The A1 amplifier, biased from the LT1004 reference, provides optimum drive for D1, the noise source. AC coupled A2 takes a broadband gain of 100. The A2 output feeds a gain control stage via a simple selectable lowpass filter. The filter’s output is applied to LT1228 A3, an operational transconductance amplifier. A3’s output feeds LT1228 A4, a current feedback amplifier. A4’s output, the circuit’s output, is sampled by the A5 based gain control configuration. This closes a gain control loop back at A3. A3’s ISET input current controls its gain, allowing overall output level control.

To adjust this circuit, place the filter in the 1kHz position and trim the 5k potentiometer for maximum negative bias at A3, Pin 5.

Figure 2 shows noise at a 1MHz bandpass while Figure 3 plots amplitude vs RMS noise in the same bandpass. Figure 4 plots similar information at full bandwidth. RMS output is essentially flat to 1.5MHz with about ±2dB control to 5MHz before sagging badly.

Figure 2. Figure 1's Output in the 1MHz Filter Position.

Figure 3. RMS Amplitude vs Frequency for the Random Noise Generator Is Essentially Flat to 1MHz.

Figure 4. RMS Amplitude Holds Within ±2dB Before Sagging Beyond 5MHz.

Author

Jim-Williams

Jim Williams

James M. Williams (April 14, 1948 – June 12, 2011) was an analog circuit designer and technical author who worked for the Massachusetts Institute of Technology (1968–1979), Philbrick, National Semiconductor (1979–1982) and Linear Technology Corporation (LTC) (1982–2011).[1] He wrote over 350 publications[2] relating to analog circuit design, including 5 books, 21 application notes for National Semiconductor, 62 application notes for Linear Technology, and over 125 articles for EDN Magazine. Williams suffered a stroke on June 10 and died on June 12, 2011.