Notes on IF and AF Response |
||||||
![]() |
||||||
| This test measures the overall selectivity of a receiver. It assesses the cumulative impact of the receiver's selective circuits in the RF, IF, and AF systems. This is an important test. Receivers usually have enough sensitivity (at least on 14 MHz and lower), and often have enough dynamic range for North American operators. But some receivers suffer from dramatic deficiencies in their selectivity. Textbook discussions of selectivity always concentrate on a receiver's IF and AF filters. There is, however, another way to approach selectivity, based on audio "presence" (which is one of the main virtues of direct conversion receivers). If a receiver's distortion levels are low enough, the brain can separate signals remarkably well. And the audio is a joy to listen to. The ARS Lab will investigate this strategy in the future. For the time being, we will follow the traditional path of focusing on filters. Filter-based selectivity divides into two issues: are the receiver's filters the right shape, and are they functioning in the appropriate ways in the receiver's RF, IF, and AF circuits (the issue of "distributed selectivity"). For now, we'll concentrate on shape, but we hope to include more material on distributed selectivity in the future. The ARRL tests a receiver's IF and AF response by measuring the 6 dB bandwidth, and so do we. The test is very simple, using a single signal generator and a true RMS AC voltmeter. We also test the cumulative "shape" of the receiver's selective circuits with a computer-based AF spectrum analyzer. We connect a noise source to the receiver's antenna port and feed the AF output into the sound card of our computer. We use SpectraPLUS, version 2.32, to analyze the AF spectrum. Keep in mind a caveat. A spectrum analyzer responds differently to noise than sine-wave signals. So noise will not generate accurate amplitude measurements, unless the spectrum analyzer is calibrated for noise. In addition, the noise bandwidth of a filter is different from its resolution bandwidth. Nevertheless, an uncalibrated AF spectrum analyzer will show relative amplitudes well, and its overall information about the cumulative RF, IF, and AF response of a receiver is still quite illuminating. |
||||||