Перейти к содержанию

Lab 10.6 — Digital RF Attenuator: Step, Range and Linearity Check

Goal

Characterize a 0…31.75 dB digital RF attenuator with 0.25 dB steps as a controllable element of the SDR bench. The lab turns manual level setting into reproducible sweep experiments for SNR, BER, ADC overload, level margin and measurement repeatability.

Digital RF attenuator placeholder

Equipment

  • digital RF attenuator with SMA input/output and USB power;
  • NanoVNA, signal generator/spectrum analyzer, or SDR TX/RX pair;
  • 50-ohm cables and loads;
  • fixed protective 10–30 dB attenuator for TX/RX loopback;
  • measurement log or CSV file.

Safety

Do not use the digital attenuator as the only protection element during the first transmitter-to-receiver connection. Start with a fixed attenuator and verify the power budget. The digital attenuator is convenient for sweeps, but its maximum input power and matching must be validated separately.

What to measure

  1. Zero setting: actual insertion loss at ATT = 0 dB.
  2. Attenuation error at 0, 3, 6, 10, 20, 30 and 31.75 dB.
  3. Repeatability when returning to the same setting.
  4. Frequency dependence: at least two frequencies, for example 10 MHz and the SDR operating frequency.
  5. SDR behavior: how dBFS level, SNR and BER change during an attenuation sweep.

NanoVNA procedure

  1. Warm up the instruments for 5–10 minutes.
  2. Perform a Thru calibration for the cables without the attenuator.
  3. Insert the attenuator between the NanoVNA ports.
  4. For each ATT_set, save S21 at the selected frequency.
  5. Compute the error: error_dB = S21_measured - (-ATT_set - insertion_loss_0dB).
  6. Fill the table and decide which settings are safe to use in lab sweeps.

SDR procedure

  1. Configure Zynq/AD936x or a generator for a safe output level.
  2. Connect: TX → fixed ATT → digital ATT → RX.
  3. Freeze gain, frequency, sample rate, filter bandwidth and capture length.
  4. For each ATT_set, record signal level in dBFS, noise estimate, SNR and BER.
  5. Check that BER is not inferred from SNR only: under overload, CFO or sync loss the SNR may look acceptable while BER degrades sharply.

Report table

ATT_set, dB Frequency S21, dB Normalized attenuation, dB Error, dB RX level, dBFS SNR, dB BER Conclusion
0
3
6
10
20
30
31.75

Minimum quality criteria

  • The report includes the connection diagram and the protective fixed attenuator.
  • Instrument settings are saved for every measurement point.
  • 0 dB insertion loss is separated from the requested attenuation.
  • The conclusion states whether the attenuator is suitable for BER/SNR sweeps.
  • SNR and BER are reported together, not treated as interchangeable metrics.

Connection to Block 11

This lab directly supports runtime PL BPSK/QPSK measurements through RTL-SDR and AD936x loopback. A digital attenuator enables a repeatable controlled attenuation sweep and helps show where the link is noise-limited, where it is overloaded, and where synchronization or CFO dominates.