Georgia Institute of Tec hnologySc hool of Electrical and Computer EngineeringECE 4435 Op Amp Design Laboratory Fall 2004Design Project 1, PreliminariesA White Noise and Pink Noise GeneratorIn troductionLoudspeaker systems in professional audio applications are normally set up by equalizing theirfrequency response with graphic and/or parametric equalizers. The equalization requires a testsignal that is called pink noise. Pink noise is generated from white noise by filtering it with a filterthat has a frequency response that varies inversely with the square root of the frequency over theaudio frequency band. White noise is noise which has a flat spectral density. That is, it has equalvoltage per unit frequency. In contrast, pink noise has equal vo ltage per octave or per any fractionof an octa v e. The object of this design project is to design a white noise source and a filter whichwill con vert the white noise output into pink noise.Preliminary ProcedureThe first step in the lab is to assemble a white noise source. The circuit shown in Fig. 1 issuggested. There is no theoretical way to design this circuit. It must be optimized experimentally.The transistor can be either a 2N4401 or 2N3904. The circuit sho ws the base connected to thecollector. The circuit ma y generate more noise if the base is grounded and the collector is opencircuited. Alternately, more noise may be generated with the collector grounded and the base ope ncircuited. You are to decide the optim um connection in the lab. For the power supply voltages, use+15 V and −15 V. The transistor is operated in the reverse breakdown region of its base-emitterjunction, so that it resembles a zener diode. The noise it generates will vary with the dc currentthrough it. This curren t is set b y the resistor R1. This resistor must be experimentally determinedto obtain the maximum noise and noise bandwidth. A suggested starting value is R1= 100 kΩ.For C1, a suggested value is 10 µF.ForR2and R3, respectively, suggested values are R2=10kΩand R3=100kΩ.Figure 1: Suggested white noise generator.After the optimum value of R1is determined, R3should be adjusted so that the observed noisewaveform on the oscilloscope has peak values that are approximately in the range of 120 mV to180 mV. For a noise signal, the rms value is approximately 1/4 to 1/3 of the observed peak values.Thusthermsvalueofvoshould be in the range from 30 mV to about 60 mV. You should turn on1the 20 MHz bandwidth limit on the oscilloscope to measure the noise. With Dr. Brewer’s reluctantassistance, the noise spectrum of your noise generator can be measured and documented.This is only the first part of this experiment. Further details will be posted. In particular asuggested circuit for con verting the white noise in to pink noise will be given with its design