Measurement Results


8-bit Current-Steering DAC

Implementation:
(1) I-V conversion is accomplished using a single-ended TIA.
(2) Waveform generator implemented in ADALM2000.


The pattern generator is used to generate an 8-bit sawtooth signal. Only static parameters are fully characterized for the DAC.



Fig. 1: Digital IO Signals of ADALM2000

DAC produced a stair-step function with voltages. Change from a minimum of 20 mV to a maximum of 5.02 V.



Fig. 2: DAC Output Waveform

The data from the static test for every input was collected and written in an excel file. The DNL and INL of the result are calculated below. The maximum value of DNL and INL is 2 LSB and 2.5 LSB respectively.



Fig. 3: DNL vs. DAC Code



Fig. 4: INL vs. DAC Code

8-bit SAR ADC

Implementation:
(1) Analog Discovery is used to generate the clock signal and differential input signal. It is also used as a logic analyzer.
(2) Only dynamic parameters are fully characterized for the ADC.



Fig. 5: SNDR vs. Frequency


Fig. 6: SFDR vs. Frequency


Fig. 7: ENOB vs. Frequency


Fig. 8: SNDR vs. Amplitude


Fig. 9: SFDR vs. Amplitude


Fig. 10: ENOB vs. Amplitude

Ramp signal in the time domain. 256 steps, each step lasting for 32 us.



Fig. 11: Ramp Test

Radar Frontend

The blue signal is the modulating chirp, and the yellow signal is the output of the radar.



Fig. 12: Radar Frontend Measurement

Sweeping Spectrum Analyzer Signal Chain

The blue signal is the triangular wave used for modulation, the green signal is the radar's output.



Fig. 13: I_Path_DAC

The green signal is the output of the radar and the blue signal is after mixing with the sweeping LO.



Fig. 14: I_Path_Mixer

The green signal is the pulse width modulated sweeping LO. The blue signal is the output signal of the mixer filtered by the anti-aliasing filter. Note that the amplitude of the sweeping LO is not decreasing over time, the droop is due to the limited sampling rate of the oscilloscope. The signal has a deterministic pattern at 250 Hz which is the same as the chirp signal. This signal is here because the radar frontend has a strong TX to RX leakage and the modulating pattern is coupled to the RX directly. This signal needs to be filtered out.



Fig. 15: PWM_LO_I_Filtered

The serialized data will be sent to the FPGA and then sent to Matlab where a 64-tap lowpass FIR filter is used to filter the pattern due to TX to RX leakage. Then the result of I^2 + Q^2 is plotted to find out the power at a certain frequency. As shown above, the result of a 1-meter-away object and a 1.5-meter-away object are shown above.



Fig. 16: Radar Output Spectrum_1m


Fig. 17: Radar Output Spectrum_1.5m


Back to top