Lab 10.5 — NanoVNA and RF Demo Kit: S11/S21, VSWR and Smith Chart¶

Goal¶

Learn to use a NanoVNA-H4 and an RF Demo Kit as a safe bridge between RF theory and SDR bench measurements. The lab helps the student see not only a spectrum in SDR software, but also real matching, reflection, transmission and passive-network frequency response.

Equipment¶

NanoVNA-H4 or a compatible VNA.
RF Demo Kit for NanoVNA-F or a similar educational board with SMA/IPX ports.
SOLT calibration kit: Short, Open, Load, Thru.
SMA/IPX adapters and short cables.
3–10 dB attenuator for input protection when the connection is uncertain.
Optional: NanoVNA-Saver for plots, CSV/Touchstone export and TDR analysis.

Safety¶

A NanoVNA is a small-signal instrument. Do not connect it to a Zynq/AD936x transmitter output, signal generator or amplifier without a power budget and a protective attenuator. Before connecting an unknown circuit, check that there is no DC voltage and no external RF source on the port.

Short theory¶

A VNA measures complex S-parameters. For this lab two quantities are enough:

S11 — input reflection. It is used to estimate matching, VSWR and the point on the Smith chart.
S21 — forward transmission through a two-port network. It is used to measure the frequency response of a filter, cable, attenuator or thru path.

The SDR connection is direct: poor matching and unexpected S21 notches may look like poor SNR, although the root cause is in the RF path, cable, adapter or filter rather than in DSP.

NanoVNA mini-glossary¶

Term	Meaning in this lab
`DUT`	Device under test: measured circuit or component.
`PORT1` / `CH0`	Source/reflection port; primary port for `S11`.
`PORT2` / `CH1`	Receiver port for the transmitted signal; used for `S21`.
`STIMULUS`	Sweep range setup: START, STOP, CENTER, SPAN.
`LOGMAG`	Magnitude in dB; convenient for filters and attenuators.
`PHASE`	S-parameter phase.
`SMITH`	Smith chart for impedance and matching.
`MARKER`	Frequency marker for reading the value at the operating point.
`TDR`	Time-domain reflectometry: cable length and discontinuity check from reflections.

Mandatory calibration¶

Before measurements, perform SOLT calibration in the exact frequency range and with the same cables that will be used later:

RESET the previous calibration.
OPEN at the end of the PORT1 cable.
SHORT at the end of the PORT1 cable.
LOAD 50 Ω at the end of the PORT1 cable.
ISOLN: usually keep PORT1 terminated with LOAD and leave PORT2 open or terminated if the kit allows it.
THRU: connect PORT1 and PORT2 through the same cables and adapter.
Press DONE and save the calibration to a slot.

Important: if the cable, adapter or frequency range is changed after calibration, the measurement is no longer strict. State this explicitly in the lab report.

Procedure¶

Power on the NanoVNA and set the sweep range, for example 50 kHz to 900 MHz or a narrower range around the SDR operating frequency.
Perform SOLT calibration for the selected range and the actual measurement cable set.
Configure at least four traces: S11 LOGMAG, S11 SMITH, S21 LOGMAG, S21 PHASE or DELAY.
Place markers at the bench operating frequencies, for example 10 MHz, 70 MHz, 144/433 MHz or the current AD936x/RTL-SDR experiment frequency.
Measure the Short, Open, Load and Thru standards on the RF Demo Kit.
Measure the 33 Ohm, 75 Ohm, ATT -3 dB and ATT -10 dB circuits.
Measure the BSF 6.5 MHz, BPF 10.7 MHz, LPF 400 MHz and HPF 500 MHz filters.
Save a screenshot or CSV/Touchstone file for each measurement.
Fill the result table and decide which board elements are useful as educational references for SDR measurements.

Extension A — cable loss and TDR¶

This extension is useful before long Block 11 loopback experiments.

Calibrate the NanoVNA over a wide range up to the useful upper frequency of the instrument.
Connect the cable under test as a two-port network and measure S21 LOGMAG.
Record cable loss at the SDR experiment frequencies.
Connect one cable end to PORT1 and leave the far end OPEN or terminate it with SHORT.
Open TDR in NanoVNA-Saver and estimate cable length/discontinuities.
In the report, state where cable loss can no longer be ignored in the link budget.

Extension B — L/C/resonance in practice¶

This extension connects Block 10 with filters, matching and parasitic effects.

Connect a small capacitor or inductor to PORT1 through a short SMA adapter.
Enable S11 SMITH and S11 Serial L or S11 Serial C if the firmware/software supports it.
Sweep several ranges and check how the L/C estimate depends on frequency.
Build a simple LC tank and find resonance from |Z| or from the S21 peak/notch.
Conclude where the component is still close to ideal and where parasitic effects dominate.

Report table¶

Object	Range	Trace to inspect	Expected behavior	Measured result	Conclusion
Load 50 Ω	near operating frequency	S11, VSWR	low reflection
Open	full range	Smith chart	chart edge
Short	full range	Smith chart	opposite chart edge
Thru	full range	S21	near 0 dB plus cable loss
ATT -3 dB	operating frequency	S21	about -3 dB
ATT -10 dB	operating frequency	S21	about -10 dB
LPF 400 MHz	50 kHz…900 MHz	S21	roll-off after cutoff
HPF 500 MHz	50 kHz…900 MHz	S21	rejection below cutoff
Cable	SDR bench frequencies	S21, TDR	frequency-dependent loss, main reflection at the far end
LC tank	near resonance	S11 \/	Z	or S21	pronounced resonance

Check questions¶

Why should calibration be performed at the ends of the same cables used for the DUT measurement?
Why do S11 and S21 answer different engineering questions?
How can a bad adapter damage BER even when the signal level looks normal?
Why is a Smith-chart trace more informative than a single VSWR number?
What changes when the sweep range is narrowed around the SDR operating frequency?
Why does cable loss increase with frequency?
Why do inductors and capacitors stop being ideal at high frequencies?
How do you distinguish a matching problem from a DSP synchronization problem?

Connection to Zynq-SDR¶

After this lab, the student should be able to validate the passive part of the RF path before BER/SNR experiments: cable, attenuator, filter, adapter and load. This is especially important for Blocks 6, 10 and 11, where an RF-path error can masquerade as a synchronization issue, CFO, ADC overload or FPGA logic defect.

Artifacts¶

NanoVNA screenshots;
CSV/Touchstone measurement files;
NanoVNA-Saver file or screenshot;
filled result table;
short conclusion on which RF Demo Kit elements are suitable for the educational bench;
separate cable conclusion: whether the current cables can be used in the selected range without a noticeable correction.