Measuring TS2000 Noise Figure using NFM and RFD2305

This article describes measurement of Noise Figure of a receiver using a calibrated noise source, a means of measuring noise power output of the receiver, and calculation of Noise Figure using the Y factor method. See Noise Figure Y factor method calculator and Noise Figure Meter for further information on the method.

The RFD2305 is a calibrated noise source from RF Design.

The supplied calibration chart was keyed up in a tab delimited file for import to NFM to allow automatic interpolation of ENR at the frequency of measurement.

Because of the expected Noise Figure of the TS2000, the external 10dB attenuator that is part of the calibrated RFD2305 was removed to obtain a higher Y factor from the receiver. To make allowance for that change, the value of Att1 is set to -10dB.

The abbreviated measurement procedure was:

  1. key input fields in NFM were completed (ENR file imported, frequency, interval, and title);
  2. sound card levels preset;
  3. the equipment was connected up, the receiver powered OFF and SCZ measured;
  4. receiver powered ON, preamp ON, Attenuator OFF, noise source powered OFF, and Stage 1 Noise LO measured;
  5. noise source powered ON, and Stage 1 Noise HI measured; and
  6. NFM automatically calculates the Y factor, and then the receiver Noise Temperature and Noise Figure.

The use of a measurement interval of 5s  with receiver effective noise bandwidth of 1500Hz gives a noise measurement uncertainty of 0.1dB at 95% confidence (see Noise measurement uncertainty calculator), considerably better than would usually be obtained using a multimeter or audio millivoltmeter. Whilst in this case, the sensitivity of calculated Noise Figure to uncertainty in each noise measurement is relatively low, measurement of a LNA with 1dB noise Figure is quite sensitive to uncertainty in each noise measurement.


© Copyright: Owen Duffy 1995, 2016. All rights reserved. Disclaimer.