Repository navigation
Rev 2 Testing

Given the bandpass filter is on one edge of the board, it seemed like an ideal place to start for assembly and testing. Using the Digilent Analog Discovery 2 (DAD2) Network window, I was able to create a bode plot of the output as the DAD2 swept from 10kHz to 10MHz. Our band that we are aiming for is 5-10MHz, and while initially it looks as if the results are not as intended, peaking around 3MHz, it appears to start to curve back up, which is not terribly surprising, as many of the bandpass filters we tested with in LTspice had a concave top, with the edges slightly higher than the center. While we don’t have a good way of testing higher frequencies, we are at least able to see the low frequencies are highly attenuated.

Our low-pass filter works great. This time I did a sweep from 1kHz to 500kHz. Our intended cutoff frequency was 100kHz, which is shown here to be ~8dB lower than the slower frequencies that make up the roughly-flat line from 1-20kHz. Given that our IF signal is supposed to be 10.7KHz, this should be perfect

Initial tests showed that the voltage smoother did drop an input 5V DC down to ~4.3V as expected. Additionally, the bias point made up from the voltage division of R1 and R4 on the left was as expected around 2.15V. However, Caleb and I found that the analog ground net (the junction of R16 and R17) was at a much lower voltage than intended, reading around 1.3V rather than 2.15V. This GNDA net is used for our instrumentation amps’ REF input to account for the 2.15V bias. Caleb reviewed the data sheet for the instrumentation amps, and it turns out they require a low output impedance from the voltage source connected to the REF pin, as shown below. INSERT IMAGE