Mixers & Modulators — RF Theory

// 01 — Fundamentals

What is a Mixer?

A mixer is a three-port nonlinear device that performs frequency translation. It multiplies two input signals — the RF signal and a Local Oscillator (LO) — to produce an output containing their sum and difference frequencies. The output port is called the IF (Intermediate Frequency) port.

Mixers appear in every transmitter and receiver. In a receiver they downconvert a high-frequency RF signal to a lower IF for filtering and demodulation. In a transmitter they upconvert a baseband or IF signal to the RF carrier frequency.

Figure 1 — Mixer three-port symbol. RF and LO enter; IF exits. The × symbol denotes frequency multiplication (convolution in the frequency domain).

Core Mixer Output Frequencies

Input: V_RF = A·cos(2πf_RFt), V_LO = B·cos(2πf_LOt)
Output: V_IF ∝ V_RF × V_LO = (AB/2)·cos(2π(f_RF−f_LO)t) + (AB/2)·cos(2π(f_RF+f_LO)t)

Desired IF = |f_RF ± f_LO| · Unwanted image and sum terms removed by filtering

// 02 — Mathematics

Mixer Mathematics

Multiplying two sinusoids using the product-to-sum trigonometric identity reveals exactly what frequencies appear at the output.

Product-to-Sum Identity — Full Expansion

cos(A)·cos(B) = ½·cos(A−B) + ½·cos(A+B)

V_RF(t) = A·cos(2πf_RFt)
V_LO(t) = B·cos(2πf_LOt)

V_out(t) = V_RF·V_LO
= (AB/2)·cos(2π·(f_RF−f_LO)·t) ← difference (IF)
+ (AB/2)·cos(2π·(f_RF+f_LO)·t) ← sum (filtered out)

For downconversion: keep f_IF = |f_RF − f_LO|
For upconversion: keep f_RF = f_LO + f_IF (sum term)

A real mixer is not a perfect multiplier. The LO switches the mixer between +1 and −1, which is equivalent to multiplying by a square wave. A square wave contains all odd harmonics:

Square-Wave LO — Output Frequency Products

LO_sq(t) = (4/π)·[cos(ω_LOt) − cos(3ω_LOt)/3 + cos(5ω_LOt)/5 − …]

Output products: f_out = n·f_LO ± f_RF (n = 1,3,5,...)
n=1: f_LO±f_RF (desired, amplitude = 4/π × input) → Conversion Loss = 20·log(π/4) = 3.92 dB
n=3: 3f_LO±f_RF (spurious at −9.5 dB rel. to n=1)
n=5: 5f_LO±f_RF (spurious at −14.0 dB rel. to n=1)

// 03 — Downconversion (Receiver)

Downconversion (RX)

In a receiver, the mixer translates a high-frequency RF signal down to a lower Intermediate Frequency (IF) for amplification, filtering and demodulation. The LO frequency sets where in the spectrum the receiver is tuned. Moving the LO is equivalent to tuning the receiver.

Figure 2 — Downconversion receiver chain. LNA amplifies the RF signal, BPF rejects the image frequency, the mixer translates to IF, and the IF filter selects the desired channel.

Figure 3 — Downconversion spectrum. RF at 2.4 GHz with LO at 2.3 GHz (low-side injection) produces IF = 100 MHz. The image signal at 2.2 GHz also converts to 100 MHz and must be rejected before the mixer by the image-reject filter.

Example 1 — 2.4 GHz WiFi Downconversion

f_RF = 2.400 GHz, f_LO = 2.300 GHz (low-side), desired f_IF = ?

1

f_IF = f_RF − f_LO = 2.400 − 2.300 = 100 MHz

2

Image frequency: f_image = f_LO − f_IF = 2.300 − 0.100 = 2.200 GHz

3

Image separation from RF: f_RF − f_image = 2.400 − 2.200 = 200 MHz = 2×f_IF

4

Sum product at output: f_RF + f_LO = 2.400 + 2.300 = 4.700 GHz (filtered by IF filter)

✓ f_IF = 100 MHz. Image at 2.2 GHz must be rejected by ≥ IRR dB before the mixer. Low IF = harder image rejection (image close to RF).

// 04 — Upconversion (Transmitter)

Upconversion (TX)

In a transmitter, the mixer translates a low-frequency baseband or IF signal up to the RF carrier frequency. The sum product of the mixing is selected (rather than the difference). Upconversion must produce a spectrally pure output — any LO leakage or unwanted sidebands corrupt the transmitted signal.

Figure 4 — Upconversion transmitter chain. Baseband or IF signal is upconverted by the mixer. TX BPF rejects the unwanted lower sideband (f_LO−f_IF) and LO feedthrough. PA amplifies the resulting RF signal.

Figure 5 — Upconversion spectrum. IF at 100 MHz mixed with LO at 2.3 GHz produces two sidebands at 2.2 GHz (lower) and 2.4 GHz (upper, desired). The TX BPF selects the upper sideband and rejects the lower sideband and LO feedthrough.

// 05 — LO Injection Side

LO High-Side vs Low-Side Injection

The LO can be placed either below (low-side) or above (high-side) the RF frequency. Both produce the same IF frequency — but the image frequency location, spectrum inversion, and practical trade-offs differ significantly. This choice has major consequences for filter design and system architecture.

Low-Side Injection — LO below RF

Figure 6 — Low-side LO injection. LO sits below RF. Image is at f_LO − f_IF, which is below the RF band. Spectrum orientation is preserved (higher RF frequencies produce higher IFs). Image separation = 2×f_IF.

High-Side Injection — LO above RF

Figure 7 — High-side LO injection. LO sits above RF. Image is at f_LO + f_IF, which is above the RF band and farther from the RF passband. Spectrum is inverted — higher RF frequencies produce lower IF frequencies.

Low-Side vs High-Side — Full Comparison

Parameter	Low-Side (LO < RF)	High-Side (LO > RF)
LO frequency	f_LO = f_RF − f_IF	f_LO = f_RF + f_IF
IF frequency	f_IF = f_RF − f_LO	f_IF = f_LO − f_RF
Image location	f_im = f_LO − f_IF = f_RF − 2×f_IF (below RF)	f_im = f_LO + f_IF = f_RF + 2×f_IF (above RF)
Image separation from RF	2×f_IF	2×f_IF
Image in RX band?	Often yes (e.g. adjacent channel)	Often in adjacent band — easier to reject
Spectrum orientation	Preserved — high RF → high IF	Inverted — high RF → low IF
LO frequency	Lower → easier VCO design	Higher → harder VCO, more phase noise
LO phase noise impact	Lower LO freq → less phase noise	Higher LO freq → more phase noise
LO pulling risk	LO far from antenna → less pulling	LO closer to RF band → more pulling risk
Image reject filter	Must reject signal in same/adjacent band	Image is outside main band — easier rejection
Typical use case	AM/FM broadcast, cable, low-IF	TV tuners, wideband RX, superheterodyne
Example (2.4 GHz, IF=100 MHz)	LO=2.3 GHz, image=2.2 GHz	LO=2.5 GHz, image=2.6 GHz

Example 2 — High vs Low side for 2.4 GHz, IF = 500 MHz

f_RF = 2.4 GHz, f_IF = 500 MHz. Compare both injection sides:

L

Low-side: f_LO = 2.4−0.5 = 1.9 GHz, image = 1.9−0.5 = 1.4 GHz
Image is 1.0 GHz below RF — well outside the 2.4 GHz band, easily rejected ✓

H

High-side: f_LO = 2.4+0.5 = 2.9 GHz, image = 2.9+0.5 = 3.4 GHz
Image is 1.0 GHz above RF — well outside band, also easily rejected ✓. But LO is 500 MHz higher.

✓ With f_IF = 500 MHz both sides work well. With f_IF = 10 MHz (low IF), image is only 20 MHz away — much harder to reject, making filter design critical regardless of injection side.

High IF rule: A higher IF makes image rejection easier — the image is farther from the RF passband. A low IF (e.g. 455 kHz for AM radio) requires very selective pre-filtering. This is the fundamental trade-off in superheterodyne receiver design — high IF for easy image rejection vs. low IF for easy channel filtering.

// 06 — Image Frequency

Image Frequency

The image frequency is the most insidious problem in heterodyne receivers. It is a second RF frequency that produces the same IF as the desired signal — it cannot be separated from the desired signal after mixing and must be rejected before the mixer using an image-reject (IR) filter or image-reject mixer topology.

Image Frequency Formulas

Low-side injection (f_LO < f_RF):
f_image = f_LO − f_IF = f_RF − 2·f_IF (image is BELOW the desired signal)

High-side injection (f_LO > f_RF):
f_image = f_LO + f_IF = f_RF + 2·f_IF (image is ABOVE the desired signal)

In both cases: image separation = |f_RF − f_image| = 2·f_IF

Image Rejection Ratio (IRR):
IRR = 20·log₁₀(|H(f_RF)| / |H(f_image)|) dB
where H(f) = transfer function of the image-reject filter (or IRF in dB)

Low-side image. Image at 700 MHz, below RF at 1 GHz. If 700 MHz signal present, it corrupts the 1 GHz receive.

High-side image. Image at 1200 MHz, above RF at 1 GHz. Image is above the band and outside the antenna passband — often easier to reject.

f_RF	f_IF	Low-Side LO	Low-Side Image	High-Side LO	High-Side Image
900 MHz	45 MHz	855 MHz	810 MHz	945 MHz	990 MHz
2.4 GHz	100 MHz	2.3 GHz	2.2 GHz	2.5 GHz	2.6 GHz
2.4 GHz	500 MHz	1.9 GHz	1.4 GHz (far)	2.9 GHz	3.4 GHz (far)
5.8 GHz	200 MHz	5.6 GHz	5.4 GHz (in-band)	6.0 GHz	6.2 GHz (out-of-band)
28 GHz (5G)	5 GHz	23 GHz	18 GHz (far)	33 GHz	38 GHz (far)

// 07 — Mixer Topologies

Mixer Topologies

Single-Ended Mixer

Figure 9 — Single-ended mixer. A single diode performs the mixing. Simplest topology but provides no port isolation — LO leaks directly to the RF and IF ports.

Single-Balanced Mixer

Figure 10 — Single-balanced mixer. RF balun feeds two diodes in anti-phase. LO-to-IF isolation improved (LO cancels at IF port). Still passes even-order spurs.

Double-Balanced Mixer (Ring Mixer)

Figure 11 — Double-balanced ring mixer. Four diodes in a bridge with RF and LO baluns. All three ports are isolated from each other. Even-order spurs cancel due to ring symmetry. Industry standard for most microwave applications.

Topology Comparison

Parameter	Single-Ended	Single-Balanced	Double-Balanced	Double-Double-Balanced
Number of diodes	1	2	4	8
LO→RF isolation	Poor (~0 dB)	Moderate (~15 dB)	Good (~30 dB)	Excellent (>40 dB)
LO→IF isolation	Poor	Good (~25 dB)	Good (~30 dB)	Excellent (>40 dB)
RF→IF isolation	Poor	Poor	Good (~25 dB)	Excellent
Even-order spurs	Present	LO evens suppressed	Cancelled (even LO × even RF)	All even orders cancelled
Conversion loss (ideal)	3.9 dB	3.9 dB	3.9 dB	3.9 dB
LO power required	+3 to +7 dBm	+7 to +13 dBm	+7 to +17 dBm	+17 to +27 dBm
Bandwidth	Moderate	Good	Excellent (multi-octave)	Excellent
Complexity / cost	Simplest / lowest	Moderate	Moderate	Complex / highest
Typical use	Detectors, low-cost	Moderate performance	Most microwave apps	High-dynamic-range RX

// 08 — IQ Mixer

IQ Mixer

An IQ (In-phase/Quadrature) mixer uses two mixers driven by the same LO but with a 90° phase difference between them. The I (in-phase) mixer multiplies by cos(ω_LOt), while the Q (quadrature) mixer multiplies by sin(ω_LOt). Together, they produce a complex baseband output that carries both amplitude and phase information — essential for modern digital modulations.

Figure 12 — IQ mixer block diagram. RF signal is split to two identical mixers. The LO is split 90°, feeding cos(ω_LOt) to the I mixer and sin(ω_LOt) to the Q mixer. Baseband I(t) and Q(t) outputs carry the full complex envelope of the RF signal.

IQ Mixer Output — Mathematics

RF input: V_RF(t) = A(t)·cos(ω_RFt + φ(t))

I path: V_RF × 2cos(ω_LOt) → LPF →
I(t) = A(t)·cos((ω_RF−ω_LO)t + φ(t))

Q path: V_RF × 2sin(ω_LOt) → LPF →
Q(t) = −A(t)·sin((ω_RF−ω_LO)t + φ(t))

Complex baseband: s(t) = I(t) + j·Q(t) = A(t)·e^{j((ω_RF−ω_LO)t + φ(t))}

The complex signal preserves both amplitude A(t) and phase φ(t) — essential for QAM, OFDM, QPSK demodulation.

Direct Conversion (Zero-IF)

When f_LO = f_RF exactly, the IQ mixer performs direct downconversion to baseband. The IF = 0 Hz — there is no intermediate frequency stage. This eliminates the image problem and is widely used in 4G/5G and WiFi chipsets. However it introduces new problems:

Parameter	Superheterodyne (IF≠0)	Direct Conversion (Zero-IF)
Image rejection	Requires IR filter before mixer	No image problem (LO=RF)
DC offset	Not a problem	LO self-mixing creates DC at output
1/f (flicker) noise	IF above 1/f corner	Baseband at 1/f corner — degrades NF
IQ imbalance sensitivity	Less critical	Critical — I/Q errors corrupt demodulation
Number of filters	Multiple (image, IF, channel)	Fewer (just channel select)
Integration	Hard to fully integrate	Fully integrated CMOS possible
Typical use	Traditional radio, base stations	WiFi, Bluetooth, 4G/5G UE, SDR

I/Q Imbalance

In a real IQ mixer the two paths are never perfectly matched. Amplitude imbalance (ε) and phase imbalance (Δφ) corrupt the demodulated signal by folding the image onto the desired signal. The Image Rejection Ratio (IRR) quantifies how much the unwanted sideband is suppressed:

IQ Imbalance — Image Rejection Ratio

Amplitude imbalance: ε (fractional, e.g. 0.01 = 1%)
Phase imbalance: Δφ (degrees)

IRR = 10·log₁₀[ (1 + 2ε·cos(Δφ) + ε²) / (1 − 2ε·cos(Δφ) + ε²) ] dB

Small imbalance approximation: IRR ≈ 20·log₁₀(2/√(ε²+Δφ²[rad]))

Examples (perfect amplitude, only phase error):
Δφ = 1° → IRR ≈ 40 dB
Δφ = 3° → IRR ≈ 30 dB
Δφ = 10° → IRR ≈ 20 dB (marginal for 64-QAM)
Δφ = 0.1° → IRR ≈ 55 dB (required for 256-QAM)

Amp. Error ε (%)	Phase Error Δφ	IRR (approx)	Adequate for
0%	0.1°	55 dB	256-QAM, 5G NR
0%	0.5°	47 dB	64-QAM, LTE
0%	1.0°	40 dB	16-QAM
0%	3.0°	30 dB	QPSK, Bluetooth
1%	1.0°	34 dB	16-QAM marginal
3%	3.0°	24 dB	BPSK only
5%	5.0°	19 dB	Poor — requires calibration

// 09 — Key Specifications

Mixer Specifications

Conversion Loss (CL) & Conversion Gain

Conversion Loss

CL = P_RF,in (dBm) − P_IF,out (dBm) (positive = loss)
Ideal passive (diode) mixer: CL = 20·log₁₀(π/4) = 3.92 dB (theoretical minimum)
Real passive mixer: CL = 5–8 dB (diode Ron, balun losses, mismatch)
Active (Gilbert cell) mixer: CL < 0 dB (conversion gain)

Example: P_RF = −30 dBm, CL = 6 dB → P_IF = −36 dBm

Port Isolation

Port Isolation Definitions

LO→IF isolation = P_LO − P_{LO-at-IF-port} (dB) — LO leaking to IF
LO→RF isolation = P_LO − P_{LO-at-RF-port} (dB) — LO leaking back to antenna
RF→IF isolation = P_RF − P_{RF-at-IF-port} (dB)

Poor LO→RF isolation: LO leaks back through the antenna and interferes with adjacent receivers (a legal and regulatory issue in licensed bands!)
Poor LO→IF isolation: LO appears at the IF output and can saturate downstream circuits

Linearity — IIP3 and P1dB

Mixer Linearity Parameters

Mixers are inherently nonlinear (that is how they work!) but should be linear in the RF signal amplitude for given LO drive.

IIP3: input-referred third-order intercept (same definition as amplifiers)
IIP3 = P_in + ΔIM3/2 (measured with two tones at RF port)

P1dB: 1 dB compression point — RF input power where CL increases by 1 dB

Rule of thumb: P1dB_in ≈ IIP3 − 9.6 dB (same as amplifiers)

LO drive impacts linearity: higher LO → lower CL → better P1dB and IIP3
Optimal LO power is a key design parameter (too high → IMD from LO harmonics)

Noise Figure

Mixer Noise Figure

NF_mixer = CL + NF_excess (dB)
For a passive mixer: NF_DSB ≈ CL (double-sideband NF equals conversion loss)
For an SSB NF: NF_SSB = NF_DSB + 3 dB (noise from image band folds in)

Example: Passive mixer with CL = 6 dB:
NF_DSB = 6 dB → NF_SSB = 9 dB

Active (Gilbert cell) mixer: NF = 8–15 dB (conversion gain offsets loss, but transistor noise adds)

Spurious Products — Spur Table

A real mixer generates outputs at n·f_LO ± m·f_RF for all integer n, m. These spurious products (spurs) can fall in-band and corrupt the desired IF. The dominant spurs are the odd-order products (n=1,3,5 for passive mixers).

Figure 13 — Mixer spur chart. f_RF=2.4 GHz, f_LO=2.3 GHz. The desired IF=100 MHz. The dangerous in-band spur at 2.2 GHz (n=2,m=1: 2f_LO−f_RF) can fall inside the RF passband and must be managed through LO power and filter design.

Key spur products for f_RF=2.4 GHz, f_LO=2.3 GHz

n (LO order)	m (RF order)	f_out = \|n·f_LO ± m·f_RF\|	Relative level	Risk
1	1	100 MHz & 4700 MHz	0 dBc (reference)	Desired IF
2	1	2200 MHz, 7000 MHz	−6 dBc	Can fall in RF band
1	2	2500 MHz	−6 dBc	Near RF band
3	1	4500 MHz, 11400 MHz	−9.5 dBc	Out of band
2	2	200 MHz	−12 dBc	Near IF
3	2	2100 MHz	−12 dBc	In RF band
1	0	2300 MHz (LO)	~−30 dBc (isolation)	LO feedthrough
0	1	2400 MHz (RF)	~−25 dBc (isolation)	RF feedthrough

Typical Mixer Specs — Technology Comparison

Parameter	Passive Diode DBM	Active (Gilbert Cell CMOS)	MMIC Active (GaAs)
Conversion Loss/Gain	5–8 dB loss	5–15 dB gain	3–10 dB gain
NF (SSB)	6–10 dB	10–18 dB	8–15 dB
IIP3	+15 to +30 dBm	0 to +15 dBm	+5 to +25 dBm
P1dB	+5 to +20 dBm	−10 to +5 dBm	−5 to +15 dBm
LO Port Isolation	30–40 dB	20–30 dB	25–35 dB
LO Drive Required	+7 to +17 dBm	−10 to +5 dBm	0 to +10 dBm
DC power	Zero (passive)	5–100 mW	50–500 mW
Frequency range	DC to 100+ GHz	DC to 10 GHz	DC to 100+ GHz
Integration	Discrete only	Full SoC integration	MMIC chip
Typical application	Base stations, test equipment, radar	WiFi, 4G/5G handsets, IoT	Satellite, 5G mmWave, defense

// 10 — Practical Design

Practical Design Examples

Example 3 — 2.4 GHz WiFi Receiver Mixer Budget

Example 3 — Passive DBM for 2.4 GHz WiFi receiver

f_RF=2.4 GHz, f_LO=2.3 GHz (low-side), f_IF=100 MHz. Mixer: Mini-Circuits ADE-1 (typical specs)

1

Conversion loss: CL = 6.5 dB at LO drive = +7 dBm

2

Noise figure (SSB): NF = CL + 3 = 9.5 dB

3

IIP3 (input-referred): IIP3 = +13 dBm

4

Image frequency: f_im = 2.3−0.1 = 2.2 GHz (must reject by >30 dB)

5

LO→RF isolation: 35 dB. LO at +7 dBm → leakage to RF port = 7−35 = −28 dBm (acceptable)

6

Cascaded NF with LNA (NF=2 dB, G=18 dB): F_mixer = 10^(0.95) = 8.91. Friis: F_total = 1.585 + (8.91−1)/63.1 = 1.585+0.125 = 1.71 → NF=2.33 dB

✓ With 18 dB LNA before the mixer, system NF = 2.33 dB despite the mixer's own 9.5 dB NF. This illustrates why high LNA gain is critical — it suppresses the mixer NF contribution in the Friis cascade.

Example 4 — IQ Mixer Image Rejection Budget

Example 4 — 900 MHz direct-conversion receiver IQ balance

Requirement: IRR ≥ 40 dB for LTE QPSK demodulation

1

IRR from phase error alone (at 40 dB): 40 dB → Δφ ≈ arcsin(1/100) ≈ 0.57° maximum phase error

2

IRR from amplitude error alone (at 40 dB): 20·log(1+ε)/(1−ε)=40 dB → ε = ≈1% maximum amplitude imbalance

3

Combined budget: If Δφ=0.3° and ε=0.5%, combined IRR ≈ 46 dB — meets requirement with margin

4

For 256-QAM (5G NR): Required IRR ≥ 55 dB → Δφ < 0.1°, ε < 0.2% → requires digital IQ calibration

✓ Practical IQ mixers achieve 35–40 dB IRR without calibration. Digital calibration in the baseband DSP can improve this to 55–70 dB for high-order modulations.

Common mixer design mistakes:
· Insufficient LO drive — too low LO power increases CL and degrades IIP3
· No image-reject filter before the mixer — image noise degrades NF by 3 dB (DSB vs SSB)
· LO frequency on wrong side for the modulation scheme — inverts the spectrum, DSP must compensate
· Ignoring LO→antenna leakage — can violate spectrum regulations (e.g. 2.3 GHz LO leaking into 2.3–2.4 GHz licensed bands)
· Mixing LO signal from a PLL with high reference spurs — reference spurs appear as spurious outputs at IF ± n×f_ref

Cascaded NF — include mixer as a stage → Plot mixer S-params from .s2p → Check stability with mixer in chain →

// 11 — Try the Tools

Put This Theory Into Practice

Use these RFLab tools to design and verify receiver chains containing mixers — from cascaded noise figure to S-parameter analysis of the complete signal chain.

Signal Chain Calculators

⚡

CALCULATOR

Cascaded Noise Figure

Enter the mixer as a stage with its CL as loss (NF=CL) and no gain (G=−CL). Compute the full Friis cascade for LNA → BPF → Mixer → IF Amp. See Example 3 above — the LNA gain of 18 dB reduces the mixer NF contribution from 9.5 dB to just 0.13 dB.

⚡

CALCULATOR

Attenuator Design

Use a small pad between LNA and mixer input to improve input match and reduce LNA-to-mixer interaction. A 2–3 dB pad stabilises the impedance the mixer presents to the LNA and reduces re-reflections that cause gain ripple.

⚡

CALCULATOR

LC Filter Design

Design the image-reject filter that must attenuate the image frequency before the mixer. For f_RF=2.4 GHz, f_IF=100 MHz, the image is at 2.2 GHz — the filter must provide >30 dB at 2.2 GHz while passing 2.4 GHz with <1 dB loss.

⚡

CALCULATOR

Noise Budget Analysis

Model the complete RX chain for IQ sensitivity budget: Antenna → Cable → BPF → LNA → IR Filter → Mixer (CL=6.5 dB, NF=9.5 dB) → IF Amp. Verify system NF meets the sensitivity requirement from the Noise Theory page.

S-Parameter & Analysis Tools

📡

S-PARAMS

S-Parameter Plotter

Upload mixer datasheet .s2p file. Plot S11 at RF port to verify the mixer input impedance is close to 50 Ω — poor match degrades conversion loss and causes gain ripple in the signal chain. Also check S22 at IF port to verify IF filter termination.

📡

S-PARAMS

Cascade S-Parameters

Cascade LNA .s2p + IR Filter .s2p + Mixer .s2p to see the combined S21 (overall CL from antenna to IF) and S11 (system input match). Verify the LNA gain and IR filter loss leave the mixer operating in its linear region.

📡

S-PARAMS

Amplifier Stability

Check LNA stability with the mixer and IR filter connected. The mixer presents a reactive impedance at the LNA output at frequencies outside the IF band — an LNA stable alone can become conditionally unstable when the mixer is attached. Always check K and μ with the full chain.

📡

S-PARAMS

Smith Chart Plotter

Plot mixer RF port impedance on the Smith chart. Most passive DBMs have S11 near 50 Ω across a wide bandwidth, but active mixers can have significant reactive impedance. Use to design the matching network at the mixer RF input if needed.

📡

S-PARAMS

De-embedding Tool

When measuring mixer conversion loss on a board, connectors and PCB launches add insertion loss and affect port match. De-embed the test fixture to get the true mixer CL — removing the board parasitics from the measurement.

📡

S-PARAMS

Group Delay & TDR

Group delay variation at the IF output of the mixer affects signal fidelity for wideband modulations. Upload the IF filter + mixer Touchstone file and measure GD flatness. Excess GD variation causes ISI, directly degrading EVM in OFDM systems.

Related Theory Pages

📘

THEORY

Noise in RF Systems

Mixer NF and cascaded IIP3 are explained in full detail. The Friis formula shows exactly how mixer NF is suppressed by LNA gain. IIP3 cascaded through LNA→Mixer is the primary dynamic range limit of a receiver — both are on the Noise page.

📘

THEORY

RF Filter Theory

The image-reject filter before the mixer is one of the most demanding filter specifications in a receiver. Filter Theory explains how to choose order and approximation type, and calculate the Chebyshev or Butterworth component values for the required attenuation at the image frequency.

📘

THEORY

Impedance Matching

Matching the mixer RF input to 50 Ω minimises conversion loss and ensures the LNA sees a stable load impedance. For active mixers with reactive input impedances, a stub or L-network matching network is required — design it using the Impedance Matching page.

// Suggested workflow — 2.4 GHz WiFi receiver mixer selection and integration

1

Choose LO side: low-side (f_LO=2.3 GHz) gives normal spectrum orientation. Image at 2.2 GHz, 200 MHz from RF.

2

LC Filter Calculator → design 2.4 GHz BPF that provides >30 dB at 2.2 GHz image. Chebyshev n=3, f_c=2.4 GHz, check 2.2 GHz attenuation.

3

Noise Figure Calculator → enter LNA (NF=1.5 dB, G=18 dB) → BPF (NF=1.5 dB) → Mixer (NF=9.5 dB, G=−6.5 dB) → IF Amp → verify total NF meets system spec.

4

Upload mixer datasheet .s2p to S-Param Plotter → verify S11 at 2.4 GHz is near 50 Ω. Plot S22 at 100 MHz IF port.

5

Cascade LNA + BPF + Mixer .s2p → check combined gain, input match, and stability with Amplifier Stability tool.