The purpose of this experiment is to verify how much switching noise the switch-mode power stage introduces to the output of the excitation amplifier. Figure 5-14 shows the test setup circuit diagram for the output spectrum measurement. Similarly, to the transfer function measurement the arbitrary waveform generator is isolated from the input using the isolation transformer. This allows connecting the spectrum analyzer across the amplifier output. A passive 1:10 probe protects the spectrum analyzer input from overload.

Figure 5-14 Test Setup for the Output Spectrum Measurement

Figure 5-15 shows the output spectrum from 0 to 10MHz and selected zoom window from 0 to 1MHz.

Figure 5-15 Output Spectrum of the Excitation Amplifier (Harmonic Input)

Further detailed analysis in Figure 5-16 shows first 8 harmonics. The first unwanted harmonic at 30kHz has 40dB (100x) lower amplitude to the carrier. This is a very good result.

Figure 5-16 Detailed Analysis of the Output Spectrum 0-100kHz

Figure 5-17 shows comparison of the output spectrum when the signal generator drives the input with the harmonic signal (A) and modulated PWM (B). The PWM signal introduces new harmonics.

Figure 5-17 Output Spectrum Comparison for Harmonic (A) and PWM (B) Signal Driving the Input

Figure 5-18 shows detailed analysis of the output signal with the PWM signal as an input. The detail also shows aliased harmonics wrapped around 400kHz. Note that the switching frequency of the TSD5402-Q1 device is 400kHz. However, due to the BD modulation scheme the frequency practically doubles and increase the Nyquist criterion to 400kHz.

Figure 5-18 Output Spectrum of the Excitation Amplifier (PWM Input)