How does sample rate relate to frequency and bandwidth on SDR?

In an SDR, three related but distinct concepts often get mixed together: frequency, bandwidth, and sample rate. Here’s how they actually fit together:

1. Frequency (centre frequency)

This is simply where you tune the SDR.
Example: tuning to 1420.405 MHz for the hydrogen line.

It is independent of sample rate.

---

2. Bandwidth (instantaneous RF bandwidth)

This is how much spectrum the SDR digitises at once.

For most SDRs:

Bandwidth  ≈  SampleRate

Because sampling at Fs captures frequencies in:

- Fs/2   to   + Fs/2

around the tuned centre frequency.

Example:
If you sample at 2 MHz, you get roughly 2 MHz of spectrum (±1 MHz around the centre).

---

3. Nyquist relation

To avoid aliasing:

SampleRate  >=  2 * (SignalBandwidth)

This is why wider signals require higher sample rates.

---

4. Putting it together

Term	Meaning	Depends on
Centre Frequency	Where you tune	You
Bandwidth	How much spectrum is captured	Mostly sample rate + SDR filters
Sample Rate	How fast the ADC takes samples	Limits bandwidth

---

5. Example with SDRplay RSPduo

If you set:

• Centre frequency: 1420.405 MHz
• Sample rate: 2.048 MS/s
• IF bandwidth: 2 MHz

Then your captured RF is roughly:

Centre = 1420.405 MHz

Span   = ± (SampleRate / 2)   ≈ ±1.024 MHz

So the SDR "sees":

1419.381 MHz   to   1421.429 MHz   (approx)

Exact edges depend on internal analog filtering but this is the idea.

---

6. Why this matters in radio astronomy

Higher sample rate gives:

• wider instantaneous bandwidth
• better sensitivity after integration
• ability to take higher-resolution spectra (big FFT sizes)
• more continuum information for calibration

But also:

• higher CPU load
• much larger data files
• more demanding correlation if doing interferometry (RSPduo, gr-corr, etc.)

---