This post was last edited by Feihong Haojie on 2019-9-9 12:11
Signal generators are often used as input signals of signal chains in analog circuit debugging. The actual performance of the circuit is verified through local testing to ensure that the design meets the requirements. Circuit debugging is often a long process. This time, TGF4042 and an oscilloscope are used as test tools to observe and verify the characteristics of a 10MHz Butterworth low-pass filter designed before.
At the bottom of the circuit board, you can see a passive filter composed of a string of inductors and capacitors. When I designed it, I just did a simple simulation.
Connect a signal to the filter input, measure the output signal, and observe the amplitude change. First connect a 1MHZ signal
The actual waveform amplitude is close to 1V, and the waveform has not been significantly attenuated
Continue to increase the frequency, 2MHz has no obvious change
Directly add to 10Mhz, the amplitude is reduced to 790mV, the attenuation is obvious
Continue to increase to 11MHz, the amplitude is 712mV, and the attenuation is close to 3dB
At 20MHz, the signal attenuates to 580mV
When the frequency continues to increase to 30Mhz, the signal amplitude is only 272mV
From the actual effect, the 3dB cutoff frequency of the filter is 1MHz higher than the original simulation parameters. This filter is used on the DDS output. The DDS clock is 100MHz and the maximum frequency of the design signal is 10MHz. Therefore, a low-pass filter is made to filter out high-order harmonics. Although there is a gap with the simulation, I think this effect is acceptable.