How to Measure SDR Frequency Accuracy with a Reference

[Pavel]

Short answer: Warm the SDR, receive or inject a known stable reference, measure frequency error and express it in hertz and PPM. Repeat at several frequencies and temperatures; a moving transmitter, satellite Doppler or unknown consumer oscillator is not a calibration reference.

Before you start​

• Keep GNSS and drone work passive unless the experiment has explicit authorization and an RF-controlled environment; do not jam, spoof or interfere with navigation or aircraft.
• Record receiver clock, antenna, gain, sample rate, observation location precision and ground truth before interpreting anomalies.

Step-by-step method​

1. Step 1: Select a reference with documented accuracy and connect safely over the air or through attenuation. Keep the SDR within linear range.
2. Step 2: Warm up in the normal enclosure, use a narrow FFT or demodulator and estimate the carrier center without clipping or modulation bias.
3. Step 3: Calculate error as measured minus nominal frequency; calculate PPM as error divided by nominal frequency times one million.
4. Step 4: Apply correction if the software supports it, retest and repeat at another frequency and after temperature change to measure residual and drift.

Concrete example​

A 100 MHz reference appearing at 100.0015 MHz is +1,500 Hz high, or +15 PPM. The required software correction sign must be verified by observing whether the peak moves toward nominal.

How to judge the result​

The measurement is credible when repeated readings agree within stated reference, FFT and temperature uncertainty and multiple frequencies produce a consistent oscillator model.

What to record​

• Separate frequency-reference accuracy from absolute time alignment: 10 MHz disciplines frequency while PPS marks a time boundary.
• Use multiple independent observables for GNSS interference or spoofing claims, such as spectrum, AGC behavior, correlation, navigation consistency and trusted reference data.
• For direction finding or RF classification, report uncertainty and test across channels, distances, devices and temperatures.

Common mistakes​

• Changing several hardware, software or RF variables at once, which removes the controlled comparison needed to identify the cause.
• Treating one autoscaled screenshot or one unusually good result as proof without recording the settings and repeating the test.
• Using a voice transmission's apparent center can introduce modulation and transmitter offset larger than the SDR error being measured.

Final check​

Repeat the passive observation against an independent trusted reference and state uncertainty. Do not make spoofing, interference or aircraft claims from a single receiver screenshot.