matlab-analyze-rf-mixer

Mixer Intermodulation Analysis

Analyze mixer spurious products using the mixerIMT object. Build intermodulation tables, visualize spur charts, and integrate with rfbudget for cascade analysis.

When to Use

• Building intermodulation tables (IMT) for mixer spur analysis
• Visualizing spur charts at specific RF input frequencies
• Identifying spurious products and designing frequency plans
• Using mixerIMT in rfbudget for cascade analysis with spur-aware mixer models

When NOT to Use

• Simple cascade budgets without spur analysis -- use modulator via matlab-create-rfbudget-elements
• General rfbudget cascade analysis -- use matlab-analyze-rf-budget
• Composing mixer circuits with node wiring -- use matlab-compose-rf-circuit
• Processing complex baseband signals through mixers -- use matlab-process-rf-baseband

Workflow

1. Create mixer — Construct mixerIMT with LO, ConverterType, and power levels
2. Configure IMT — Set custom intermodulation table or use defaults
3. Visualize spurs — Plot spur chart with rfplot(m, fRF)
4. Integrate — Add to rfbudget chain or circuit for cascade/circuit analysis

mixerIMT vs modulator

Both model frequency converters, but serve different purposes:

Feature	modulator	mixerIMT
Gain specification	Gain property (direct)	NominalOutputPower - ReferenceInputPower
Nonlinearity	OIP3-based (poly model)	IMT table (arbitrary spur levels)
Spur visualization	No	Yes — rfplot(m, fRF)
Constructor accepts Gain	Yes	No — settable after construction but ignored by rfbudget
Use case	System-level cascade budgets	Spur analysis and frequency planning

Use modulator for quick cascade budgets. Use mixerIMT when you need to model specific spurious products or visualize spur charts.

Creating a mixerIMT

m = mixerIMT('LO', 2.1e9, 'ConverterType', 'Down', 'NF', 10, 'Name', 'Mixer');

Constructor Parameters

Parameter	Default	Description
LO	1e8	Local oscillator frequency (Hz)
ConverterType	'Up'	'Down' or 'Up' conversion — must set explicitly for down-conversion
NF	0	Noise figure (dB)
ReferenceInputPower	-15	Reference input power (dBm)
NominalOutputPower	-5	Nominal output power (dBm)
Model	auto	'mod' (up) or 'demod' (down) — set automatically from ConverterType
Zin	50	Input impedance (Ohm)
Zout	50	Output impedance (Ohm)
IMT	3x3 default	Intermodulation table
Name	'MixerIMT'	Element name

Post-Construction Properties

These properties are read-write but cannot be passed to the constructor:

m.OIP3 = 20;       % Output IP3 (dBm)
m.OIP2 = 40;       % Output IP2 (dBm)

Do not set m.Gain manually. Although Gain is a settable property, rfbudget ignores it — the budget gain is determined entirely by NominalOutputPower and ReferenceInputPower. Setting m.Gain has no effect on cascade results and can be misleading.

Conversion Gain

Conversion gain in rfbudget is determined by NominalOutputPower and ReferenceInputPower. Although Gain is a settable property, rfbudget ignores it and computes gain from NominalOutputPower - ReferenceInputPower. Do not set m.Gain manually.

Budget gain = NominalOutputPower - ReferenceInputPower

% This gives 10 dB conversion gain in rfbudget:
m = mixerIMT('ReferenceInputPower', -15, 'NominalOutputPower', -5, ...
    'LO', 2.1e9, 'ConverterType', 'Down', 'Name', 'Mixer');
% Budget sees: -5 - (-15) = 10 dB

% To get -6 dB conversion loss in rfbudget:
m2 = mixerIMT('ReferenceInputPower', -10, 'NominalOutputPower', -16, ...
    'LO', 2.1e9, 'ConverterType', 'Down', 'Name', 'Mixer');
% Budget sees: -16 - (-10) = -6 dB

The IMT Table

The intermodulation table (IMT) is an MxN matrix where:

• Rows = RF harmonic order (row 1 = order 0, row 2 = order 1, ...)
• Columns = LO harmonic order (col 1 = order 0, col 2 = order 1, ...)
• Values = relative power in dBc below the reference (non-negative)
• 0 = reference level (desired output)
• 99 = fully suppressed

Default Table

m = mixerIMT;
disp(m.IMT);
%    99    99    99
%    99     0    99
%    99    99    99

The default 3x3 table has only the desired product (RF×1, LO×1) at row 2, col 2 with value 0 (reference). Everything else is suppressed (99).

Custom IMT Table

imt = 99*ones(5,5);       % Start with everything suppressed
imt(2,2) = 0;             % RF×1, LO×1: desired output (reference)
imt(2,1) = 20;            % RF×1, LO×0: RF feedthrough at -20 dBc
imt(1,2) = 15;            % RF×0, LO×1: LO feedthrough at -15 dBc
imt(3,2) = 30;            % RF×2, LO×1: 2RF-LO spur at -30 dBc
imt(2,3) = 25;            % RF×1, LO×2: RF-2LO spur at -25 dBc
m.IMT = imt;

Gotcha: The IMT must be a real 2-D numeric matrix (class double). Setting a cell array, 3-D array, complex matrix, or non-numeric type errors with "An Intermodulation Table must be real two dimensional matrix." Ensure imt is created with standard double operations (e.g., 99*ones(M,N)).

Gotcha: IMT values must be non-negative and ≤ 99 (they represent dBc below reference). Setting a negative value errors with "Value must be nonnegative." Setting a value > 99 errors with "Expected Spur table to be an array with all of the values <= 99." Using inf for "fully suppressed" also errors ("Value must be finite") — use 99 instead.

Reading the IMT Table

IMT Position	RF Order	LO Order	Spur Frequency (Down)
(1,1)	0	0	DC
(2,1)	1	0	fRF (RF feedthrough)
(1,2)	0	1	fLO (LO feedthrough)
(2,2)	1	1	fRF - fLO (desired IF)
(3,2)	2	1	2·fRF - fLO
(2,3)	1	2	fRF - 2·fLO
(3,3)	2	2	2·fRF - 2·fLO

Spur Chart Visualization

m = mixerIMT('LO', 2.1e9, 'ConverterType', 'Down', 'NF', 10, 'Name', 'Mixer');
figure;
rfplot(m, 2.4e9);   % Spur chart at fRF = 2.4 GHz

rfplot(m, fRF) displays a spur chart showing all intermodulation products at the specified RF input frequency, with their relative power levels.

Gotcha: rfplot(m, fRF) takes a scalar frequency, not a vector. Passing a frequency vector errors: "Value must be a scalar."

Gotcha: rfplot(m) with no frequency argument errors: "Not enough input arguments."

Using in rfbudget

lna = amplifier('Gain', 15, 'NF', 2, 'OIP3', 35, 'Name', 'LNA');
m = mixerIMT('ReferenceInputPower', -10, 'NominalOutputPower', -16, ...
    'NF', 10, 'LO', 2.1e9, 'ConverterType', 'Down', 'Name', 'Mixer');
ifAmp = amplifier('Gain', 20, 'NF', 4, 'OIP3', 30, 'Name', 'IFAmp');

b = rfbudget([lna m ifAmp], 2.4e9, -30, 10e6);
fprintf('Total Gain: %.2f dB\n', b.TransducerGain(end));
fprintf('Output Freq: %.3f GHz\n', b.OutputFrequency(end)/1e9);

After the mixer, the output frequency changes to the IF: fIF = fRF - fLO for down-conversion.

Using in circuit

mixerIMT is a 2-port (4-terminal) element compatible with circuit:

ckt = circuit('MixerCircuit');
add(ckt, [1 2 0 0], m);
setports(ckt, [1 0], [2 0]);
s = sparameters(ckt, freq);

S2D Data Files

Load mixer characterization from an S2D file:

m = mixerIMT('FileName', 'mixer_data.s2d', 'UseDataFile', true, ...
    'LO', 2.1e9, 'ConverterType', 'Down', 'Name', 'MeasuredMixer');

OpenIF -- IF Frequency Planning

The OpenIF object finds spur-free intermediate frequencies in multiband receiver/transmitter architectures. Use it to identify open IF bands where no mixer spurs fall.

hif = OpenIF(20e6);                    % 20 MHz IF bandwidth
imt1 = [99 99 99; 99 0 99; 99 99 99]; % Default 3x3 IMT
addMixer(hif, imt1, 2.4e9, 200e6, 'low', 20e6);   % RF center, RF BW, type, IF BW
addMixer(hif, imt1, 5.2e9, 200e6, 'low', 20e6);
show(hif);                             % Graphical spur-free zone display
report(hif);                           % Command-window summary

addMixer Syntax

addMixer(hif, imt, rfCenterFreq, rfBW, mixType, ifBW)

Argument	Type	Description
hif	OpenIF	The OpenIF object
imt	MxN double	Intermodulation table (same format as mixerIMT.IMT)
rfCenterFreq	scalar	RF center frequency (Hz)
rfBW	scalar	RF bandwidth (Hz)
mixType	char	'low' or 'high' when IFLocation='MixerOutput'; 'sum' or 'diff' when IFLocation='MixerInput'
ifBW	scalar	IF bandwidth for this mixer (Hz)

'low' means fIF = fRF - fLO (standard down-conversion); 'high' means fIF = fRF + fLO.

Property	Default	Description
IFBW	constructor arg	System-wide IF bandwidth (Hz)
SpurFloor	99	Maximum spur level to consider (dBc)
IFLocation	'MixerOutput'	'MixerOutput' or 'MixerInput'

OpenIF complements mixerIMT: use mixerIMT + rfplot(m, fRF) to visualize spurs at a single RF frequency; use OpenIF to find spur-free IF zones across an entire multiband architecture.

Gotchas

1. Do not set m.Gain — Although Gain is settable, rfbudget ignores it and computes gain from NominalOutputPower - ReferenceInputPower. Setting it manually has no effect on the budget and can be misleading.
2. Gain not in constructor — mixerIMT('Gain', -6) errors. Gain cannot be passed as a constructor argument.
3. IMT values must be 0-99 — They represent dBc below the reference (0 = reference, 99 = fully suppressed). Negative values error ("Value must be nonnegative"), values > 99 error ("values <= 99"), and inf errors ("Value must be finite"). Always use 99 for suppressed spurs, never inf.
4. rfplot takes scalar frequency — rfplot(m, fRF) requires a single frequency, not a vector.
5. Default conversion gain is 10 dB — NominalOutputPower(-5) - ReferenceInputPower(-15) = 10 dB. This is rarely what you want — set both explicitly.
6. ConverterType sets Model automatically — 'Down' sets Model='demod', 'Up' sets Model='mod'.
7. ConverterType defaults to 'Up' — If you omit ConverterType, the mixer performs up-conversion (fOut = fRF + fLO). Always set 'ConverterType', 'Down' explicitly for down-conversion receivers.
8. IMT must be a real 2-D double matrix — Cell arrays, 3-D arrays, complex matrices, or non-numeric types error with "An Intermodulation Table must be real two dimensional matrix." Build with 99*ones(M,N) then set individual entries.
9. Clone elements before reusing in a second rfbudget — An element object can only belong to one rfbudget at a time. Reusing it errors with "Element is already in another rfbudget." Use clone(element) to create an independent copy.

Conventions

• Use tiledlayout/nexttile for multi-panel spur analysis figures
• Always label spur charts with the RF input frequency
• Set ReferenceInputPower and NominalOutputPower explicitly — do not rely on defaults
• Use modulator for simple cascade budgets; use mixerIMT when spur analysis matters

---

Copyright 2026 The MathWorks, Inc.

---