How does undersampling work in a bandpass sampling receiver architecture?
Bandpass Sampling Principles
Traditional sampling theory requires the sample rate to exceed twice the highest frequency in the signal (Nyquist criterion). However, for a narrowband bandpass signal centered at a high frequency, this requirement is unnecessarily strict. The actual information bandwidth is the signal bandwidth B, not the carrier frequency fc. Bandpass sampling exploits this by sampling at a rate of at least 2B, allowing the signal to alias into the first Nyquist zone.
| Parameter | Superheterodyne | Direct Conversion | Digital IF |
|---|---|---|---|
| Image Rejection | 60-90 dB (filter) | 30-50 dB (mismatch) | N/A (digital) |
| DC Offset | No issue | Major issue | No issue |
| LO Leakage | Low | High | Low |
| Integration | Difficult | Easy (single chip) | Moderate |
| Dynamic Range | 80-120 dB | 60-90 dB | 70-100 dB |
Noise Sources
The valid sampling rates for a bandpass signal centered at fc with bandwidth B are constrained to avoid the signal straddling a Nyquist zone boundary. The signal must fall entirely within one Nyquist zone (each zone spans fs/2 in width). Valid sample rates satisfy: 2fc/(2k+1) ≤ fs ≤ 2fc/(2k), where k is an integer representing the Nyquist zone number. Only certain discrete ranges of sample rates produce valid, non-overlapping aliases.
Cascade Analysis
The primary advantage is eliminating the analog mixer and LO synthesis, replacing them with the sampling process itself. This simplifies the hardware significantly. However, the ADC must have sufficient analog bandwidth (input bandwidth, not sample rate) to accurately capture the high-frequency signal. Many ADCs have analog input bandwidths of several GHz while supporting sample rates of only hundreds of MHz.
- Performance verification: confirm specifications against the application requirements before finalizing the design
- Environmental factors: temperature range, humidity, and vibration affect long-term reliability and parameter drift
- Cost vs. performance: evaluate whether the application demands premium components or standard commercial grades
- Interface compatibility: verify impedance, connector type, and mechanical form factor match the system architecture
- Margin allocation: include sufficient design margin to account for manufacturing tolerances and aging effects
Measurement Techniques
The main challenge is the anti-aliasing bandpass filter requirement. The filter must sharply limit the input to the desired Nyquist zone to prevent signals from adjacent zones from overlapping with the desired signal after aliasing. This filter operates at the RF or IF frequency and must have steep skirts and sufficient out-of-band rejection.
Frequently Asked Questions
Does undersampling degrade SNR?
Each aliasing fold brings noise from the folded Nyquist zones into the baseband. If the input noise is wideband, the SNR degrades by approximately 10·log10(number of Nyquist zones folded). The anti-aliasing filter is critical for limiting noise to the desired zone.
What ADC specifications matter for undersampling?
Analog input bandwidth (must exceed the signal frequency), jitter (clock phase noise becomes more critical at higher frequencies), and SFDR at the input frequency (not the sample rate). The ADC's effective number of bits (ENOB) typically decreases at higher input frequencies.
Can I use undersampling for wideband signals?
Only if the signal bandwidth is narrow relative to the carrier frequency. A 10 MHz signal at 2 GHz is an excellent candidate for undersampling (0.5% fractional bandwidth). A 500 MHz signal at 2 GHz would require nearly 1 GS/s sampling, offering minimal benefit over direct sampling.