NanoVNA Deep Dive

Hardware revisions (V2, SAA-2N, Hugen V3, LiteVNA, NanoVNA-F), calibration (open / short / load / through), S11 / S21 measurement, Smith-chart workflow, time-domain reflectometry, NanoVNA-Saver host software

Section	Topic
1	About this volume
2	What a VNA does — the 2-port S-parameter primer
3	NanoVNA hardware history — V2, SAA-2N, Hugen V3, LiteVNA, NanoVNA-F
4	Comparing models — frequency range, dynamic range, screen, cost
5	Calibration — OSL (open / short / load) and through
6	Calibration kits — quality matters
7	S11 measurement — return loss, SWR, impedance
8	S21 measurement — gain, loss, filter response
9	Smith-chart workflow on the device
10	Time-domain reflectometry — finding faults
11	NanoVNA-Saver and other host software
12	Measuring an antenna in the field — the deployment-day workflow
13	Measuring a BALUN / UNUN / tuner
14	Buying — current good NanoVNA models in mid-2026
15	Limitations — what the NanoVNA isn’t good at
16	Common gotchas and myths
17	Resources

1. About this volume

The NanoVNA is the most important measurement tool in this entire series. Every antenna in Vols 6-15, every BALUN/UNUN/tuner in Vols 16-19, every dummy load and matching network gets sweep-tested with a NanoVNA. The 1-port S11 plot is the antenna’s report card; the 2-port S21 plot is how you characterize filters, BALUNs, and the rest of the matching-network family.

This volume covers the NanoVNA family at engineer-grade depth: the hardware revisions, the calibration discipline, the S-parameter measurement primer, the Smith-chart workflow on the device’s tiny screen, time-domain reflectometry for fault-finding, NanoVNA-Saver and host-side software, and the field-workflow for measuring a deployed antenna.

The NanoVNA democratized VNA access. Before 2018, a 2-port VNA covering DC – 3 GHz cost $5,000 minimum (used HP 8753) or $25,000+ new (Keysight FieldFox). The NanoVNA delivers comparable amateur-level measurement quality for $50-300 — a 50-100× cost reduction that put real S-parameter measurement on every amateur bench. The trade is precision (the NanoVNA’s dynamic range is 50-80 dB vs 100+ dB for premium VNAs) and absolute accuracy (NanoVNA calibration kits are imprecise compared to traceable lab kits). For amateur antenna work, the NanoVNA’s precision is more than adequate.

This volume is cross-linked into from every per-antenna chapter (Vols 6-15) for “use a NanoVNA to verify the build” guidance and from every matching-network chapter (Vols 16-19) for “use the S21 mode to characterize the network.” Cross-link to Vol 25 (Other VNAs) for what to do when the NanoVNA isn’t enough.

2. What a VNA does — the 2-port S-parameter primer

A Vector Network Analyzer (VNA) generates an RF stimulus signal and measures the complex (magnitude and phase) reflected and transmitted signals.

2.1 The S-parameters

For a 2-port device under test (DUT):

S-parameter	Meaning	Typical use
S11	Reflection at port 1 (with port 2 terminated)	Antenna SWR / impedance
S21	Transmission from port 1 to port 2	Filter / BALUN loss
S12	Transmission from port 2 to port 1	Reverse isolation
S22	Reflection at port 2	Output impedance

For amateur antenna work, S11 is the dominant measurement. You connect the antenna to port 1, sweep the VNA across the operating band, and read return loss, SWR, and complex impedance. The Smith chart’s polar plot of S11 visualizes the antenna’s impedance behavior across frequency.

For filter/BALUN/tuner work, S21 adds the second dimension: insertion loss across frequency. A good 1:1 BALUN shows S21 < -0.3 dB across 1.8-30 MHz; a good filter has a sharp cutoff visible as S21 dropping by 20+ dB above the passband.

2.2 Why “complex” matters

A traditional SWR meter measures magnitude only — it tells you “1.5:1 at the carrier frequency” but not whether the reactance is inductive or capacitive. The VNA’s complex measurement tells you “Z = 73 + j42 Ω” (resistive + reactive parts), which:

Lets you visualize the impedance on a Smith chart
Lets you compute reactance compensation (the trim direction for the antenna)
Lets you design matching networks (L-network or T-network values)
Lets you spot common-mode currents (visible as anomalous loops on the Smith chart)

The Smith chart’s geometric properties (constant-resistance circles, constant-conductance arcs) translate complex impedance into visual matching moves — see Vol 17 §6.

2.3 The NanoVNA as a budget 2-port VNA

The NanoVNA family delivers:

DC – 1.5 GHz (V2) or DC – 6.3 GHz (LiteVNA-64) frequency range
50-80 dB dynamic range
Touchscreen UI for standalone operation
USB connectivity for PC-side software (NanoVNA-Saver, NanoVNA-App)
$50-300 USD price range

The NanoVNA-Saver host application provides large-screen plots, file export (S2P format for cross-software interchange), and additional analysis features (TDR, calibration management).

3. NanoVNA hardware history — V2, SAA-2N, Hugen V3, LiteVNA, NanoVNA-F

The NanoVNA family evolved through several generations:

3.1 NanoVNA V2 / V2.2 (the original)

The original NanoVNA (edy555 / hugen fork, 2018-2020):

Frequency range: 50 kHz – 900 MHz (extended to ~3 GHz with multi-harmonic operation)
Dynamic range: 50-60 dB
Screen: 2.8” color touchscreen
CPU: ARM Cortex-M0
Cost: $30-80
Limitations: limited dynamic range above 1 GHz; small screen for plot reading

This is the entry-level NanoVNA — adequate for HF amateur work, marginal for VHF/UHF detailed measurements.

3.2 SAA-2N (the upgrade)

The SAA-2N is a higher-performance NanoVNA variant:

Frequency range: 50 kHz – 3 GHz
Dynamic range: 60-70 dB
Screen: 2.8” color touchscreen
CPU: improved ARM Cortex-M3
Cost: $100-150
Improvements: better dynamic range, more stable calibration, better DSP

The SAA-2N is the mid-tier choice — adequate for amateur HF/VHF/UHF work, with better measurement quality than the V2.

3.3 Hugen NanoVNA-H4 / NanoVNA-V2 Plus4

The Hugen H4 series adds a larger touchscreen:

Frequency range: 50 kHz – 4 GHz
Dynamic range: 60 dB
Screen: 4” color touchscreen — major UX improvement
CPU: ARM Cortex-M4
Cost: $130-200

The H4 is the choice when the small V2 screen is uncomfortable for extended use.

3.4 LiteVNA-64 (the modern premium)

The LiteVNA-64 is the top-of-consumer-line NanoVNA:

Frequency range: 50 kHz – 6.3 GHz
Dynamic range: 70-80 dB
Screen: 4” color touchscreen
CPU: dual-ARM
Cost: $200-300

The LiteVNA-64 is the practical premium amateur NanoVNA. The 6.3 GHz upper range covers Wi-Fi (2.4 GHz, 5 GHz), 23 cm amateur band (1.3 GHz), and most microwave amateur work.

3.5 NanoVNA-F V3 (the alternative architecture)

The NanoVNA-F V3 uses a different architecture (5” screen, no touch, dedicated frequency-domain display):

Frequency range: 1 MHz – 3 GHz
Dynamic range: 70 dB
Screen: 5” color, no touch (uses buttons/encoder)
Cost: $200

The F V3 is the “tabletop” NanoVNA — the larger screen and physical controls make it more suitable for bench work than the smaller touchscreen models.

4. Comparing models — frequency range, dynamic range, screen, cost

Model	Freq range	DR	Screen	Approx $ (mid-2026)
V2 (gen3)	50 kHz – 3 GHz	50-60 dB	2.8”	$50-80
SAA-2N	50 kHz – 3 GHz	60-70 dB	2.8”	$100-150
Hugen H4	50 kHz – 4 GHz	60 dB	4”	$130-200
LiteVNA-64	50 kHz – 6.3 GHz	70-80 dB	4”	$200-300
NanoVNA-F V3	1 MHz – 3 GHz	70 dB	5” (no touch)	$200

4.1 Selection guidance

Tightest budget / HF-only work: V2 ($50-80) — gets the job done
Mid-tier general amateur: SAA-2N ($100-150) — better DR, same form factor
Larger-screen preference: Hugen H4 ($130-200) — touchscreen + 4” display
VHF/UHF/Wi-Fi work: LiteVNA-64 ($200-300) — 6.3 GHz coverage
Tabletop bench preference: NanoVNA-F V3 ($200) — 5” non-touch screen

The “right” NanoVNA depends on the operator’s primary use case. For most amateur HF work, the SAA-2N is the sweet spot. For multi-band work including VHF/UHF, the LiteVNA-64 is the premium choice.

5. Calibration — OSL (open / short / load) and through

Calibration is the most-important step in NanoVNA use. A NanoVNA without calibration is essentially useless — the measurement is dominated by the connector parasitics, the cable impedance, and the internal directivity.

5.1 The OSL procedure

For port 1 calibration:

Connect Open standard to port 1 → save (the “open” reference)
Connect Short standard → save (the “short” reference)
Connect 50 Ω Load → save (the “load” reference)
Repeat on port 2 (for 2-port calibrations)
Through calibration (optional for 2-port): connect port 1 to port 2 with a known-good calibrated cable → save

The NanoVNA’s firmware uses these reference measurements to correct for connector and cable parasitics, producing accurate S-parameter measurements.

5.2 The calibration discipline

Re-calibrate whenever:

Sweep range changes: a new frequency span requires new calibration data
IF bandwidth changes: faster scans use less averaging, requiring re-cal
Port temperature changes meaningfully: ~10°C change is the threshold for serious work
The calibration kit is connected/disconnected: physical movement of the kit changes the parasitic

For amateur work, re-calibrate at least once per measurement session. For serious work, re-cal every 10°C temperature change.

5.3 The calibration data save

The NanoVNA’s firmware can save calibration to memory slots:

V2/SAA-2N: 5 calibration slots (typically labeled “Cal0” through “Cal4”)
LiteVNA-64: 10+ slots
Hugen H4: 5 slots

Each slot saves a complete sweep range + IF bandwidth + cal data. Switching between bands (HF vs VHF vs UHF) means switching between cal slots, not re-calibrating from scratch.

5.4 Cal verification

After calibration, verify with a known reference:

Connect a known-good 50 Ω load; measure S11 — should be < -30 dB (SWR < 1.06:1) across the calibrated range
Connect a known-good through cable (port 1 to port 2); measure S21 — should be -0.5 to 0 dB across the calibrated range

If verification fails, re-calibrate.

6. Calibration kits — quality matters

6.1 Stock NanoVNA cal kits

The cal kit that ships with most NanoVNAs:

Open: an open SMA jack
Short: an SMA shorting cap
Load: an SMA 50 Ω termination (typically thin-film SMD on a small PCB)

Quality: nominal. The load is typically a poor 50 Ω (often 48-53 Ω due to manufacturing tolerance), giving calibration accuracy to ~1 GHz, less above.

6.2 Aftermarket cal kits

For better measurement accuracy:

Mini-Circuits SMA cal kits: $30-100, calibrated to 3 GHz with documentation
Picotest cal kits: $60-200, calibrated to 6 GHz
Pasternack PE17 SMA kit: $300, calibrated to 18 GHz
Custom kits with traceable certificates: $500+ for serious work

The aftermarket kits give ~0.3 dB accuracy improvement at HF and 1+ dB improvement at VHF/UHF. For amateur antenna work, the stock kit is usually adequate; for compliance measurements, use a calibrated kit.

6.3 Through-line considerations

The through cable for 2-port calibration:

Short coax (15-30 cm) with both ends terminated in clean N or SMA connectors
Impedance match matters — use a 50 Ω cable with proper connectors
Quality: NIST-traceable cables ($100+) are needed for serious work; amateur use can use any clean 50 Ω cable

7. S11 measurement — return loss, SWR, impedance

The 1-port S11 measurement is the antenna-builder’s most-used NanoVNA function.

7.1 The basic S11 sweep

Calibrate the NanoVNA at the operating frequency band
Connect the antenna’s coax to port 1
Set sweep range to cover the band (e.g., 6.5 - 7.5 MHz for 40 m)
Set point density (101-401 points for HF; 401-801 for VHF/UHF)
Read return loss (dB) and SWR at the marker

7.2 What to look for

A good antenna measurement shows:

Return loss minimum < -15 dB (SWR < 1.4:1) at the design frequency
2:1 SWR bandwidth across the operating band
Resistive Z at minimum (the Smith chart marker near horizontal axis)
R near 50 Ω at minimum (the natural impedance match)

7.3 The deployment-day workflow

For each antenna build:

Pre-deployment: calibrate at the rig end of the coax
Deploy the antenna
Sweep the band
Find SWR minimum — identify the resonant frequency
Compare to design frequency — is it too high or too low?
Trim or extend the antenna:
- Too high (resonance below target): wire too long; trim
- Too low (resonance above target): wire too short; extend (cut shorter wire too long initially!)
Re-sweep, iterate

The cm/kHz trim sensitivity tables in Vol 6 §5.3 give the per-band trim amount.

7.4 The impedance reading

The NanoVNA’s Smith chart shows impedance as a complex point. At the SWR minimum:

Real part (R): read directly; this is the resistive impedance
Imaginary part (X): read directly; positive = inductive, negative = capacitive

A clean antenna at resonance shows R = 50 Ω and X ≈ 0. Deviations indicate:

R high (e.g., 75 Ω): antenna over height (elevated to where free-space impedance dominates)
R low (e.g., 35 Ω): antenna below sweet height (ground proximity loading)
X positive: antenna too long (inductive reactance from extra length)
X negative: antenna too short (capacitive reactance from missing length)

8. S21 measurement — gain, loss, filter response

For 2-port DUTs (filters, BALUNs, amplifiers, tuners), S21 measures transmission.

8.1 The S21 sweep

Calibrate including the through-line
Connect port 1 to DUT input, port 2 to DUT output
Sweep across the relevant band
Read S21 magnitude (in dB) and S21 phase

8.2 S21 interpretation

Filter: S21 should be ~0 dB in the passband, dropping by 20+ dB above the cutoff
BALUN: S21 should be < -0.3 dB in the operating band, with smooth phase response
Amplifier: S21 should be positive (gain) across the operating band
Tuner: S21 of the tuner alone (with appropriate input/output loads) measures the tuner’s insertion loss

8.3 Phase and group delay

S21 phase tells you the time delay of the device:

Group delay = -dφ/dω (the slope of phase vs frequency)
Filters with sharp cutoffs have group-delay peaks (signal envelope distortion)
BALUNs ideally have smooth phase (minimal group-delay variation)

The NanoVNA-Saver can display group delay derived from the S21 measurement.

9. Smith-chart workflow on the device

The Smith chart is a graphical impedance display. The NanoVNA’s built-in Smith mode:

9.1 Switching to Smith mode

In the Display menu: select “Smith Chart” or similar. The S11 measurement now displays as a polar plot in the Γ (reflection coefficient) plane.

The Smith chart’s key features:

Center: 50 Ω matched load (Γ = 0)
Right edge: infinite resistance + j0 (Γ = +1)
Left edge: 0 resistance + j0 (Γ = -1)
Top half: inductive impedance (positive X)
Bottom half: capacitive impedance (negative X)

9.2 Reading impedance on the Smith chart

For each point on the trace:

R (resistance) = read from the constant-resistance circle the point lies on
X (reactance) = read from the constant-reactance arc

Marker readings on the NanoVNA’s screen show R + jX numerically.

9.3 Pattern visualization

A typical antenna’s Smith chart trace forms a loop:

The trace enters the chart at one frequency
Loops through the resistive region
Exits at the higher frequency
The “size” of the loop is related to the antenna’s bandwidth

For a clean half-wave dipole at resonance, the trace passes near the center (Γ ≈ 0); the loop is small. For an off-resonance antenna, the loop is far from center.

9.4 Using the Smith chart for matching network design

For an L-network match (see Vol 17 §6):

Plot the DUT’s Z on the Smith chart
Trace along the constant-resistance circle until you reach the 50 Ω circle (this is the series-reactance step)
Trace along the constant-conductance arc to the center (this is the shunt-reactance step)
Compute the L and C values from the reactance amounts

This visual procedure is faster than the algebra.

10. Time-domain reflectometry — finding faults

TDR mode lets the NanoVNA find faults along a coax cable.

10.1 The TDR principle

The NanoVNA’s TDR mode:

Performs an inverse FFT on the S11 sweep data
Displays the result as reflections vs distance (time)
A short in the coax shows as a strong negative-going pulse at the distance of the fault
An open in the coax shows as a strong positive-going pulse
A discontinuity (kink, connector, joint) shows as a smaller-amplitude pulse

10.2 The velocity factor

The TDR’s distance accuracy depends on the coax’s velocity factor (VF):

LMR-400: VF = 0.85
RG-58: VF = 0.66
RG-8X: VF = 0.78
Heliax LDF4-50: VF = 0.88

Enter the correct VF in the NanoVNA’s settings; the displayed distance is then accurate.

10.3 The TDR workflow for finding cable faults

Disconnect the antenna from the coax (leave the coax as an open at the far end)
Connect the coax to port 1
Switch to TDR mode
Set the cable length parameter (slightly longer than the actual cable)
Set the velocity factor for your coax type
Run the sweep
The open end appears as a large positive pulse at the cable’s actual length
Intermediate reflections (anomalies) appear as smaller pulses at intermediate distances
A short would appear as a strong negative pulse at the fault distance

For a 30 m coax with a fault at 18 m from the connected end, TDR would show:

A pulse at 18 m (the fault location)
A larger pulse at 30 m (the actual end of the cable)

This is the standard tool for finding water-damaged coax sections, broken shields, or pinched connectors.

11. NanoVNA-Saver and other host software

PC-side software extends the NanoVNA’s capabilities.

11.1 NanoVNA-Saver

The dominant amateur host software:

Author: mihtjel (GitHub)
Platform: cross-platform (Windows, macOS, Linux)
License: open-source, free
Features: large-screen plots, file export (S2P format), calibration management, sweep configuration, comparison of multiple measurements
Compatibility: V2, SAA-2N, Hugen NanoVNAs, LiteVNA-64

The NanoVNA-Saver workflow:

Connect the NanoVNA to PC via USB
Launch NanoVNA-Saver
The device’s measurements stream to the PC in real-time
Plots are larger and easier to read than the device’s screen
Export measurements to S2P file for archiving or analysis in other software

11.2 NanoVNA-App (DiSlord)

Windows-only, performant:

Smaller binary, lower memory usage than NanoVNA-Saver
Faster sweep rates
Less polished UI but more responsive

11.3 NanoVNA-QT (qrp-labs)

Qt-based, cross-platform:

Used in the Linux community
Similar features to NanoVNA-Saver

11.4 NanoVNAv2 firmware

For firmware updates and source code:

GitHub: ttrftech/NanoVNA-V2-firmware — the V2 reference firmware
GitHub: edy555/NanoVNA-H — older NanoVNA-H firmware
GitHub: nanovna-v2/NanoVNA-V2-firmware — community-maintained version

Firmware updates fix bugs and add features (often calibration improvements, screen UI tweaks, additional measurement modes).

12. Measuring an antenna in the field — the deployment-day workflow

12.1 The complete workflow

Pre-deployment preparation:
- Calibrate the NanoVNA at the rig end of the coax
- Verify cal with a known-good 50 Ω load (S11 < -30 dB across the band)
- Note the calibration frequency range
Antenna deployment:
- Hoist the antenna into operating position
- Verify all mechanical connections
- Connect the coax to the antenna
In-situ measurement:
- Connect the NanoVNA to the rig end of the coax (not the antenna end — the coax acts as a transmission line that complicates measurement at the antenna end)
- Sweep across the design band (typically the entire amateur band + 10% margin)
- Identify the resonant frequency (SWR minimum or Γ = 0 point on Smith chart)
Trim/extend:
- If too high: trim the antenna; re-deploy; re-sweep
- If too low: extend the antenna; re-deploy; re-sweep
- Iterate until resonance matches target
Documentation:
- Save the final S11 sweep to a file (via NanoVNA-Saver if PC-connected)
- Record the resonant frequency and SWR minimum
- This becomes the “as-built” reference for future audit

12.2 At the rig end vs at the antenna end

At the rig end: simpler (the coax is part of the measurement); slightly less accurate (coax effects show up in the measurement); but the calibration is at the operating point.

At the antenna end: more accurate (only the antenna is measured); requires re-calibration at the antenna end (the cal needs to compensate for the coax-end cal kit position).

For amateur use, measure at the rig end. The coax effects are small enough at HF to be tolerable, and the calibration setup is much simpler.

12.3 The “if it isn’t matching” diagnostic

If the antenna’s SWR doesn’t match the predicted curve:

Resonance too far from target: cut length wrong; re-trim
SWR minimum > 2:1: poor ground, common-mode current, BALUN failure, or wrong matching network
Multiple SWR minima: harmonic resonances (EFHW behavior) or the coax common-mode resonance
Wide, shallow minimum: high-loss antenna (the SWR plot resembles a half-flat sea bottom, not a sharp V — significant antenna losses are damping the resonance)

Use the Smith chart visual to diagnose: a clean resonance shows a small loop near the center; a problematic match shows distorted loops far from center.

13. Measuring a BALUN / UNUN / tuner

For matching-network DUTs, 2-port S21 measurement is the key.

13.1 The setup

Calibrate including the through
Connect: port 1 → DUT input; DUT output → known load (or appropriate impedance terminator)
Sweep across the operating band

13.2 What to expect

1:1 current BALUN: S21 < -0.3 dB across 1.8 - 30 MHz with 50 Ω load on output; common-mode rejection (S21 with input/output swapped) > 25 dB
4:1 current BALUN: S21 < -0.5 dB across 1.8 - 30 MHz with 200 Ω load on output
9:1 UNUN: S21 < -0.5 dB with 450 Ω load on output (use a non-inductive 450 Ω resistor for measurement)
49:1 UNUN: S21 < -0.5 dB with 2450 Ω load — the high-Z load is a challenge to construct (typically a 2.45 kΩ non-inductive metal-film resistor)

13.3 The common-mode rejection test

To verify BALUN’s CMRR:

Set up: port 1 → BALUN coax input; BALUN antenna output left disconnected
Connect a small loop antenna (1-foot wire loop) to port 2
Tape the loop near the BALUN’s coax shield
Sweep
Good CMRR: S21 < -30 dB across the band (the BALUN doesn’t couple common-mode to the coax shield)
Poor CMRR: S21 = -10 to -20 dB (the BALUN allows common-mode current to flow on the shield)

This test confirms the BALUN’s primary job — preventing the coax shield from radiating.

14. Buying — current good NanoVNA models in mid-2026

Sorted by tier:

Tier	Model	Frequency	DR	Cost	Notes
Budget	NanoVNA-H V3 (gen3sw)	50 kHz – 1.5 GHz	50-60 dB	$60-90	Most basic; usable for HF
Budget	V2.2 (Chinese clones)	50 kHz – 1.5 GHz	50 dB	$40-70	Risk of clone quality issues
Budget	NanoVNA-H	50 kHz – 900 MHz	50 dB	$30-50	Predecessor to H V3
Mid	NanoVNA-H4	50 kHz – 4 GHz	60 dB	$130-200	4” screen; good general-purpose
Mid	SAA-2N	50 kHz – 3 GHz	60-70 dB	$100-150	Higher DR, smaller screen
Mid	NanoVNA-F V3	1 MHz – 3 GHz	70 dB	$200	5” non-touch tabletop
Premium	LiteVNA-64	50 kHz – 6.3 GHz	70-80 dB	$200-300	Best amateur consumer NanoVNA
Premium	NanoVNA SAA-2N Pro	50 kHz – 3 GHz	75 dB	$200	Higher-end SAA-2N variant
Premium	NanoVNA-V2 Plus5	50 kHz – 4 GHz	70 dB	$200	Plus5 generation

14.1 Recommendation

For most amateur use: SAA-2N ($100-150) is the sweet spot. Adequate dynamic range, full 3 GHz coverage, reliable build quality.

For VHF/UHF/Wi-Fi-band work: LiteVNA-64 ($200-300) — the 6.3 GHz upper range covers everything an amateur is likely to measure.

For tabletop bench work: NanoVNA-F V3 ($200) — the 5” non-touch screen with physical controls is more bench-friendly than the small touchscreens.

14.2 Used market

Used NanoVNA-Hugen H4 and SAA-2N are common on eBay at $80-150. Quality varies; look for documented sellers and recent firmware.

14.3 What to avoid

Very cheap eBay V2 clones (< $40) — quality is inconsistent; calibration kits often missing or wrong
“8 GHz NanoVNA” listings — physically impossible for the price; usually mislabeled
Used NanoVNAs without included calibration kit — buying a separate kit is more expensive than buying the NanoVNA bundle

15. Limitations — what the NanoVNA isn’t good at

15.1 Dynamic range

NanoVNA dynamic range is 50-80 dB. Premium VNAs (R&S, Keysight) reach 100-120 dB. The implications:

Cannot measure highly-isolated devices: a coupler with > 60 dB isolation is at the NanoVNA’s noise floor
Cannot measure low-loss filters precisely: a filter with < 0.1 dB passband loss is below the NanoVNA’s measurement noise
Cannot measure compression: the stimulus is fixed at ~0 dBm; can’t characterize the DUT’s compression behavior

For amateur antenna work, the NanoVNA’s DR is adequate. For commercial RF design or compliance work, step up to a real VNA (Vol 25).

15.2 Calibration drift

The NanoVNA’s calibration drifts with:

Temperature: ~0.5 dB per 10°C
Time since calibration: gradual drift, especially at the band edges
Connector wear: 100+ mate/unmate cycles cause measurable degradation

For serious work, re-calibrate frequently. For amateur work, re-cal once per session.

15.3 Microwave performance

Above 3-6 GHz, the NanoVNA’s accuracy drops:

LiteVNA-64 at 5 GHz: ±1.5 dB accuracy typical
LiteVNA-64 at 6 GHz: ±3 dB accuracy
LiteVNA-64 above 6 GHz: out of spec

For amateur work below 6 GHz (which covers HF, VHF, UHF, 23 cm, Wi-Fi 2.4 + 5), the NanoVNA is fine. For higher microwave amateur work, step up to a real VNA.

15.4 Phase noise and stability

The NanoVNA’s source is not phase-locked:

Phase noise: -80 to -100 dBc/Hz at 1 kHz offset
Frequency stability: ±20 ppm typical
For amateur antenna work: fine
For sensitive measurements (low-phase-noise oscillator testing): inadequate

16. Common gotchas and myths

“I calibrated last week, no need now” — temperature drift makes that calibration useless. Re-cal at the start of each measurement session.
“Higher point count = better measurement” — only if scaling sweep speed allows. The IF bandwidth limit means dense sweeps at fast scan can drop precision. For HF, 101-401 points is usually optimal.
“S11 of a vertical sweeps the same on the bench as deployed” — false. Ground proximity, radials, height all change the antenna’s Z significantly. Always measure as-deployed.
“NanoVNA cal kits are equivalent to lab cal kits” — false. Stock NanoVNA loads are 48-53 Ω typically; lab-grade loads are 50.0 ± 0.1 Ω. The accuracy difference is several dB at VHF.
“NanoVNA at $50 is the same as $200” — no. The cheaper units have worse dynamic range, less stable calibration, and lower-quality screens. The $150-200 mid-tier is the sweet spot.
“The TDR distance is exact” — only if the velocity factor is set correctly. Wrong VF gives wrong distance.
“I can measure my antenna at the rig and the coax doesn’t matter” — partially true. The coax loss is real and contributes to the measured SWR (a 30 m LMR-400 at 144 MHz has 1.5 dB round-trip loss). For accurate measurement, calibrate at the antenna end or compensate for coax loss.
“All NanoVNAs run the same firmware” — false. V2, SAA-2N, LiteVNA-64 all have different firmware lineages. Use the right firmware for the hardware.
“NanoVNA can replace an oscilloscope” — false. The NanoVNA measures frequency-domain S-parameters; an oscilloscope measures time-domain voltage. Different tools.
“S21 measurement always shows real loss” — true if the calibration includes the through cable. Without through calibration, S21 includes the cable’s loss, distorting the measurement.
“I can measure SWR up to 100:1 with a NanoVNA” — true in principle; the NanoVNA’s display can show it. But the accuracy degrades at extreme SWR; the meaningful range is 1:1 to ~10:1.
“NanoVNA-Saver is required for serious work” — false. The device’s standalone screen is adequate for most measurements. NanoVNA-Saver is a convenience for the PC-connected workflow.

17. Resources

NanoVNA users group: https://groups.io/g/nanovna-users — active community forum
NanoVNA-Saver GitHub: https://github.com/NanoVNA-Saver/nanovna-saver — official repository
Hugen / edy555 V2 firmware history — GitHub repositories for the V2 series
LiteVNA documentation — official LiteVNA support pages
Pozar, Microwave Engineering Ch. 4 (Network Analysis) — academic VNA reference
Ham Radio NanoVNA wiki — community-maintained NanoVNA resource
NanoVNA tutorials by W5BSF, K6XK — popular YouTube series for amateur learners
NanoVNA-App by DiSlord — Windows-only NanoVNA host
NanoVNA-QT by qrp-labs — Qt-based cross-platform NanoVNA host
VNA fundamentals app notes from R&S, Keysight, NI — vendor-published VNA tutorials applicable to the NanoVNA

Contents