Why convert to an intermediate frequency at all?

Building a tunable narrowband filter at RF is extremely difficult. A receiver covering 100 MHz to 6 GHz would need a filter that maintains constant 200 kHz bandwidth while tuning across a 60:1 frequency range. No practical technology can do this. Instead, the superheterodyne receiver converts any RF frequency to a single fixed IF (say 70 MHz) by choosing the appropriate LO frequency. The IF filter is fixed, optimized for one frequency, and can use high-Q technologies like crystal or SAW resonators that achieve sharp selectivity impossible at RF. Changing the received frequency only requires retuning the LO, which is trivially done with a PLL synthesizer. This separation of tuning (LO) from selectivity (IF filter) is the fundamental insight of the superheterodyne architecture, patented by Edwin Armstrong in 1918.

What is the image frequency and why is it a problem?

The mixer produces IF output from any input at f_LO plus f_IF or f_LO minus f_IF. If the desired signal is at f_RF = f_LO + f_IF, then a signal at the image frequency f_image = f_LO minus f_IF also produces the same IF output and cannot be distinguished after mixing. The image frequency is separated from the desired signal by 2 times f_IF. With a 70 MHz IF: the image is 140 MHz away. With a 10 MHz IF: the image is only 20 MHz away. A higher IF makes image rejection easier (more separation for the preselector filter to exploit) but makes IF filter design harder (lower fractional bandwidth). This is the fundamental superhet trade-off. Dual conversion uses a high first IF for easy image rejection and a low second IF for sharp channel selectivity.

How does a dual-conversion receiver work?

A dual-conversion receiver uses two mixing stages. The first mixer converts the RF to a high first IF (e.g., 1 GHz) for easy image rejection since the image is now 2 GHz away from the desired signal. A fixed filter at the first IF provides initial selectivity. The second mixer converts the first IF down to a low second IF (e.g., 70 MHz or 10 MHz) where narrow crystal or SAW filters provide the final channel selectivity. The first LO tunes to select the desired RF channel; the second LO is fixed. This architecture provides both excellent image rejection (from the high first IF) and excellent selectivity (from the low second IF), at the cost of additional complexity, spurious responses from the second mixing stage, and the need to manage intermodulation between the two IF chains.

Receiver Design

Superheterodyne

Edwin Armstrong patented it in 1918. Over a century later, the superheterodyne is still the dominant receiver architecture for everything from AM radios to spectrum analyzers. The idea is elegant: instead of building a tunable narrowband filter at every possible RF frequency, convert any RF frequency down to a single fixed intermediate frequency using a mixer and a tunable local oscillator. The IF filter provides the selectivity: sharp, stable, and optimized for one frequency. The LO provides the tuning: sweeping across any desired band. Separate the tuning problem from the selectivity problem, and both become tractable. This insight remains as powerful today as it was in 1918.

Category: Receiver Design

Invented: 1918 (E. Armstrong)

Key Trade-off: IF height vs. image rejection

Architecture Comparison

Architecture	Conversions	Image Rejection	Selectivity	Complexity	Use Case
Single conversion (high IF)	1	Excellent	Moderate	Low	Narrowband receivers
Single conversion (low IF)	1	Moderate	Excellent	Low	AM/FM radio
Dual conversion	2	Excellent	Excellent	Medium	HF/VHF comms, SA
Triple conversion	3	Excellent	Excellent	High	Spectrum analyzers
Direct conversion (zero-IF)	1	N/A (no image)	Digital only	Low (but DC issues)	Cellular, WiFi, SDR
Direct RF sampling	0	N/A	Digital only	High (fast ADC)	5G, radar, SDR

Mixing equation:
f_IF = |f_RF − f_LO|

Image frequency:
f_image = f_RF ± 2×f_IF
RF = 1 GHz, IF = 70 MHz: image at 860 MHz (140 MHz away)
RF = 1 GHz, IF = 10 MHz: image at 980 MHz (20 MHz away)

The trade-off:
Higher IF = easier image rejection but harder IF filter design
Lower IF = sharper channel selectivity but worse image rejection

Common Questions

Frequently Asked Questions

Why convert to IF?

A tunable narrowband filter across 100 MHz to 6 GHz is impossible. Convert to fixed IF (70 MHz) using a tunable LO; the IF filter is optimized at one frequency using crystal/SAW resonators. LO tuning is trivial (PLL). This separates tuning from selectivity.

How does the image frequency arise?

f_image = f_RF ± 2×f_IF. Both produce the same IF output. Higher IF = more image separation = easier rejection. Lower IF = less separation but sharper channel filtering. The fundamental superhet trade-off.

Dual conversion?

First mixer: high IF (1 GHz) for easy image rejection. Second mixer: low IF (70 MHz) for sharp selectivity. First LO tunes; second LO is fixed. Best of both worlds at the cost of additional spurs and complexity.

Related Terms

← Stripline Surface Mount →

Receiver Planning

IF & Image Frequency Calculator

Enter RF frequency and candidate IF values. Instantly see image frequency, required preselector rejection, and IF filter fractional bandwidth for single and dual conversion architectures.

Plan Your IF Chain