Double Conversion

Two IF stages for selectivity and image rejection.

Darshan N
Updated: 19 March 2026
8 min read

The double conversion receiver is an enhanced superheterodyne architecture that uses two successive frequency conversion stages instead of one. It was developed to solve the fundamental conflict between image rejection and channel selectivity that exists in single conversion receivers, and remains the preferred architecture for high-performance HF communication receivers and wideband spectrum monitoring equipment.

Double Conversion Receiver ArchitectureRFPreselectMixer 1(1st Conv)LO1fLO1 = fs + fIF11st IF BPFHigh fIF11st IFAmplifierMixer 2(2nd Conv)LO2Fixed fLO22nd IF BPFLow fIF22nd IFAmplifierDetectorOutput1st conversion: Image rejectionHigh fIF1 (e.g. 10.7 MHz or 70 MHz)Image far from signal, easy to reject2nd conversion: SelectivityLow fIF2 (e.g. 455 kHz)High-Q filter at low IF, excellent selectivity
Figure 1: Double conversion architecture separates the image rejection function (1st IF) from the selectivity function (2nd IF)

Why Double Conversion Is Needed

A single conversion superheterodyne receiver faces a fundamental constraint: the choice of IF determines both image rejection and channel selectivity, and these two requirements pull in opposite directions. A high IF places the image frequency far from the desired signal, making image rejection easy for the RF preselector. However, a high IF makes it difficult to build a sharp, narrow-bandwidth IF filter because the fractional bandwidth becomes small and the Q required becomes very high. A low IF gives excellent selectivity in the IF filter but brings the image frequency so close to the desired signal that the RF preselector cannot reject it adequately.

The double conversion architecture resolves this conflict by using two IF stages, each optimized for a different function. The first IF is chosen high, providing excellent image rejection because the image is well separated from the signal. The second IF is chosen low, where sharp bandpass filters with high Q can be built practically to achieve excellent channel selectivity. The two requirements are thus decoupled and each can be independently optimized.

First and Second Conversion Analysis

In the first conversion stage, the incoming RF signal at fs is mixed with LO1 to produce the first intermediate frequency fIF1. LO1 is tunable and tracks the desired station: fLO1 = fs + fIF1 (high-side injection). The first IF filter is a moderate-bandwidth bandpass filter centered at fIF1. Its role is primarily to reject the image frequency of the first conversion: fimage1 = fs + 2*fIF1.

In the second conversion stage, the first IF signal at fIF1 is mixed with a fixed-frequency LO2 to produce the second intermediate frequency fIF2. Because the first IF is already at a fixed frequency fIF1, LO2 does not need to be tunable. This simplifies the design significantly. The second IF filter is a high-Q narrowband filter centered at fIF2, which provides the channel selectivity and adjacent channel rejection. The image of the second conversion falls at fIF1 + 2*fIF2 and must be rejected by the first IF filter.

A typical HF communication receiver might use fIF1 = 70 MHz (for excellent image rejection across the 3-30 MHz HF band) and fIF2 = 455 kHz (for excellent selectivity with a ceramic or crystal filter). The second conversion image is at 70 MHz + 2*0.455 MHz = 70.91 MHz, which the moderate-bandwidth first IF filter easily rejects.

Mathematical Relationships

For the first conversion with high-side injection: fLO1 = fs + fIF1, and first stage image fimage1 = fs + 2*fIF1. For the second conversion: fLO2 = fIF1 + fIF2 (if LO2 > fIF1), second stage image fimage2 = fIF1 + 2*fIF2. The overall image rejection of the receiver is determined by how well the RF preselector attenuates fimage1 and how well the first IF filter attenuates fimage2. The final channel selectivity is determined entirely by the second IF filter bandwidth.

Example
Given:
HF communication receiver
Signal frequency fs = 14.2 MHz (amateur 20m band)
First IF fIF1 = 70 MHz
Second IF fIF2 = 455 kHz
High-side injection for first conversion

Why this formula applies:
First conversion: fLO1 = fs + fIF1 (tunable LO)
First stage image: fimage1 = fs + 2*fIF1
Second conversion: LO2 is fixed = fIF1 + fIF2
Second stage image (referred to 1st IF): fimage2 = fIF1 + 2*fIF2

Formula:
fLO1 = fs + fIF1
fimage1 = fs + 2*fIF1
fLO2 = fIF1 + fIF2 (fixed)
fimage2 = fIF1 + 2*fIF2

Substitution:
fLO1 = 14.2 + 70 = 84.2 MHz
fimage1 = 14.2 + 2*70 = 14.2 + 140 = 154.2 MHz
fLO2 = 70 + 0.455 = 70.455 MHz (fixed)
fimage2 = 70 + 2*0.455 = 70 + 0.91 = 70.91 MHz

Final Answer: LO1 = 84.2 MHz (tunable), Image1 = 154.2 MHz (140 MHz from signal, easy to reject with preselector). LO2 = 70.455 MHz (fixed), Image2 = 70.91 MHz (910 kHz from 1st IF center, easily rejected by moderate-Q first IF filter). Second IF at 455 kHz allows use of standard ceramic filters for excellent channel selectivity.
Exam Tip: In double conversion, the first IF is chosen HIGH for image rejection and the second IF is chosen LOW for selectivity. LO2 is fixed (not tunable) because the first IF is already at a fixed frequency. The second stage image fimage2 = fIF1 + 2*fIF2 must be within the rejection band of the first IF filter. This is a common multi-part GATE calculation.
Double Conversion: Frequency Planffs14.2 MHzfLO184.2 MHzfimg1154.2 MHzfIF1 = 70 MHzfIF1 = 70 MHz1st IF = 70 MHz (HIGH)Image 140 MHz away from signal2nd IF = 455 kHz (LOW)Sharp ceramic filter, excellent selectivity
Figure 2: Frequency plan showing how high first IF separates image far from signal, and low second IF enables sharp channel filtering
  • First IF is high (e.g., 70 MHz or 10.7 MHz) to push image frequency far from the desired signal, ensuring easy rejection by the preselector.
  • Second IF is low (e.g., 455 kHz) where high-Q ceramic, crystal, or mechanical filters provide sharp channel selectivity.
  • LO1 is tunable and tracks the desired station. LO2 is fixed because the first IF output is already at a constant frequency.
  • Second stage image fimage2 = fIF1 + 2*fIF2 must be rejected by the first IF bandpass filter bandwidth, which is an important design check.
  • Double conversion is widely used in HF SSB receivers, spectrum analyzers, and professional broadcast monitoring equipment where both high image rejection and high selectivity are simultaneously required.

Quick Revision

  • Double conversion uses two mixers, two LOs, and two IF stages. LO1 is tunable, LO2 is fixed.
  • First IF chosen HIGH: fIF1 >> fIF2. Purpose = image rejection. Image1 = fs + 2*fIF1.
  • Second IF chosen LOW: fIF2 << fIF1. Purpose = channel selectivity. Image2 = fIF1 + 2*fIF2.
  • fLO1 = fs + fIF1 (tunable). fLO2 = fIF1 + fIF2 (fixed). Both typically use high-side injection.
  • Advantage: decouples image rejection and selectivity requirements, allowing each to be independently optimized.
  • Disadvantage: more complex, more components, more spurious responses possible from additional mixing products.
  • Exam trap: LO2 is NOT tunable in double conversion. If a question states LO2 changes with station, that is an error.

Double Conversion Quiz

Test your knowledge of dual-conversion superheterodyne architecture and its selectivity-image rejection tradeoff.

Question 1 of 3

Q1.A double-conversion receiver uses two IF stages. The first IF is chosen to be high, and the second IF is chosen to be low. The reason for this arrangement is: