[ Reply | Next | Previous | Up ]
From: Alan G3NOQ
Date: 22 Nov 2002
Time: 10:47:52
Remote Name: 20.138.254.2
Marc - no doubt this is true in principle but the experimental errors in applying thermal methods have been found to be so large that they do not produce useable results, and this approach has been rather discredited. The usual and fairly accurate way of measuring efficiency is to measure the power going into the antenna with an in-line wattmeter, and to measure the field strength, which can be converted to power density, at a set of points in the radiation pattern. Then the integrated power density equals the total radiated power and the efficiency can be calculated according to the official definition "efficiency = total radiated power divided by the power accepted at the antenna terminals". Obviously enough points have to be measured to get an accurate integration. . . For microwave antennas the efficiency can be measured to about 0.5dB accuracy in an anechoic chamber by this method, for HF the accuracy is more like 2dBs. There are other indirect methods, e.g. the Wheeler Cap method which is all right for antennas of reasonably high efficiency but does not work when the efficiency is low, because you have to subtract two nearly equal numbers. . . There are also indirect methods for small antennas, for example using Chu's limit which gives a lower estimate for the Q of the antenna, then the efficiency can be estimated as the ratio of the measured Q to the theoretical Q - the actual efficiency is usually less than predicted by this method because the Chu limit assumes the antenna occupies a sphere in a bandwidth-efficient way, and most practical antennas like dipoles and loops do not do this . . . . Another method for small antennas is to estimate the lossless Q as the ratio of reactance to radiation resistance (both of which can be estimated accurately from formulas) and then the efficiency is the ratio of measured Q to lossless Q - this method seems to work quite well for small dipoles and loops - hope this is helpful, Alan