Frequency conversion group delay test

The parameter commonly used in communication engineering to describe signal phase distortion - group delay (also called envelope delay) is defined as: the slope of the characteristic phase shift of the network versus frequency change curve. When the phase-frequency characteristic of the network is a straight line, the group delay Td is a constant, and signals of various frequencies passing through the network are subject to the same delay, resulting in no phase distortion and envelope distortion of the amplitude modulated wave. When the phase-frequency characteristic is nonlinear, Td is not a constant. To find the group delay at a certain frequency f, just take the frequency and phase data of points f1 and f2 on both sides of the test frequency to find the group delay. In other words, in a physical sense, the group delay of a certain frequency represents the transmission time of a signal in a very narrow frequency band centered on that frequency through a system or network, which is numerically equal to the first-order differential of the phase-frequency characteristic, and its general expression is:

This article introduces two testing methods.

1. Single carrier phase method:

The vector network analyzer can be used to directly measure the phase-frequency characteristics of the system and calculate the group delay.

Generally, when the frequency interval of the vector network analyzer is 500kHz, the group delay test error is less than 1.2ns, and when the frequency interval is 100kHz, the group delay test error is less than 4ns.

2. FM/AM method:

When there is no envelope distortion in the signal transmission process, the envelope delay of the system is equal to the group delay of the system. By using this characteristic and using a smaller modulation frequency Ω, the system can be free of envelope distortion within the ωc±Ω bandwidth, thus achieving the measurement of group delay.

The principle is: the test instrument uses a lower baseband signal to modulate the high frequency carrier frequency/amplitude. After the FM/AM wave passes through the system under test, the standard demodulator in the test instrument demodulates the baseband signal containing phase distortion. The phase difference between the baseband signal at each frequency point and the output signal of the crystal oscillator is detected to obtain the group delay within the frequency range to be tested. In order to avoid the influence of slow drift, the crystal oscillator is phase-locked with the baseband signal, and its phase is locked to the average phase of the input baseband signal.

The advantage of the FM/AM method is that the system under test is closer to the actual situation, its phase detection is at low frequency, and the test accuracy is high. At the same time, since the reference phase in the test is extracted from the demodulated low-frequency baseband signal, no separate reference channel is required. Therefore, it can conveniently test the frequency conversion system.

When the modulation frequency of Anritsu's ME538K microwave system analyzer is 500kHz, the test group delay error is less than 0.5ns.

Anritsu 's 37369C and Agilent's 8720ES and other vector network analyzers. These vector network analyzers are mostly suitable for testing co-frequency systems. For testing variable frequency systems, it requires a reference channel frequency converter with very small group delay. At present, the group delay test of co-frequency and variable frequency bandpass devices can be carried out using Anritsu's scalar network analyzer 54147A. After adding the group delay option, it uses direct frequency response calibration and testing, and calculates the relative group delay through Hilbert transform, with a typical error of <1ns.