Page tree
Skip to end of metadata
Go to start of metadata

Where To Find This Example

Select Help > Open Examples... from the menus and type either the example name listed above or one of the keywords below.

Or in Version 13 or higher you can open the project directly from this page using this button. Make sure to select the Enable Guided Help before clicking this button.

Design Notes

Diode Mixer with Nonlinear and Phase Noise Simulations

This diode mixer demonstrates mixer simulation setup and measurements in the AWR Design Environment including noise figure and phase noise.  


The mixer, shown in the “Diode_Mixer” schematic uses a Lange coupler to distribute the LO and RF signals to the diodes. 

The IF energy is drawn from the summing junction of the two diodes, and passes through a filter to remove any unwanted RF and LO energy.

This project has three test benches setup for different simulations. The “Large_Signal_Test_Bench” schematic is setup to calculate the conversion loss, output spectrum, and output waveform of the circuit.

  The “Noise_Simulation_Test_Bench” schematic is setup to calculate various nonlinear noise measurements of the mixer.

The “Phase_Noise_Test_Bench” schematic is setup to simulate the phase noise out of the mixer with a given phase noise at the LO of the mixer, and to measure the contributions of "noisy" elements to the output phase noise.


This schematic is setup to look at conversion loss, the output spectrum, and the output waveforms.  There is no NLNOISE block for these calculations.   This simulate will sweep the LO power and look at the different characteristics.


This schematic is very similar to the “Phase_Noise_Test_Bench” schematic, except a regular harmonic balance port is used on the LO of the mixer, and the RF input is treated as a small signal.   This schematic will simulate the noise figure of the mixer over the IF output range of 0.5 GHz to 0.6 GHz.

Nonlinear noise analysis requires a NLNOISE component on the schematic.  The NLNOISE component has the following parameters:

PortTo: Output port index, usually the IF port in the case of mixers.

PortFrom: The port where the noise source is applied.  By convention, the source is a resistor at 290K.  This is usually the RF port in the case of mixers.

NFstart: Start of the noise frequency sweep.

NFend: End of the noise frequency sweep.

NFsteps: Number of sweep points in noise analysis.

SwpType: Linear or log sweep.

The noise is analyzed at all frequencies of the form n x f_LO +/- f_noise, where f_LO is the LO frequency and f_noise is the smallest difference between any of the LO products taken into account in the simulation and the specified noise frequency sweep.  Therefore it is easiest to specify noise frequencies as IF frequencies so that the noise will be analyzed at all frequencies of the form n x f_LO +/- f_noise_frequency_sweep.  In the same case, assuming that the mixer is an upper-sideband down converter, the noise input frequency of interest in the NF calculation is f_LO + f_noise_frequency_sweep and the noise output frequency of interest is f_noise_frequency_sweep.  The harmonic index pairs of these two noise components are (1,1) and (0,1).

Some notes on Nonlinear Noise analysis

The RF port must be terminated in order to perform the analysis with RF as a small signal.  One may sweep the LO power, the LO frequency, and the noise frequency simultaneously.  Currently available measurements of interest are noise figure, noise temperature, noise spectral density at the output, and output noise contributors (see Phase_Noise_Test_Bench, below, for example).  Note that the noise of the output termination is excluded from the analysis, as is conventional for the purpose of noise figure simulation.  Approximate conversion gain measurement is available as well.  It is approximate in the sense that it is applicable in cases where the RF signal is so small that it does not contribute to any nonlinearity.  It is a good idea to verify the accuracy of the conversion gain measurement against large signal S parameter measurement to verify that RF can be treated as small signal.  The advantage of the conversion gain measurement is simulation speed, because it requires 1 less large signal tone for harmonic balance simulation.


In this schematic, a source called OSC_W_PH_Noise is used on the LO port of the “Diode Mixer” subcircuit.  This source specifies power, impedance, frequency, and phase noise vs offset frequency.  This vector sets the shape of the phase noise.  A V_NSMTR element is connected between the output and ground to identify the output node for the noise contributor measurements.  The NLNOISE block specifies what noise frequencies will be simulated for the phase noise, and the input and output ports for the noise simulation.  This block also specifies whether and how the noise contributions to the output should be collected.  

Coupler Schematic (Subcircuit)

The Lange coupler is simulated independently from its own schematic, which is then used as a sub-circuit in the "Diode Mixer" schematic.  Any changes made in "Coupler" will affect all circuits and do not have to be repeated.  Note that the reference to the coupler in the mixer schematic (and subcircuit references in general) does not have to be represented by a "black box", but can be any existing symbol with the correct number of ports - in this case it is represented by the Lange Coupler symbol.

SOT23_Diode Schematic (Subcircuit)

This schematic is used as subcircuit in the “Diode Mixer" schematic.  This subcircuit adds the package parasitic effects to the nonlinear SPICE diode model.

User Folders

Four different User Folders are created to organize the project. Each folder contains corresponding schematics and graphs. User Folders can be created simply by right clicking User Folders node and choosing Add New > Folder. The corresponding schematics and graphs are dragged and dropped inside the User Folders or they can also be created inside the folder by right clicking the folder name and choosing Add New > .....

Modeling and Simulation

The harmonic balance simulation for the mixer is set up as follows:

• Tone 1 applied to PORT3 defines the LO frequency and swept power.  The tone information is set by double clicking on the PORT to open the Element Options dialog box and then choosing the Port tab.  This frequency is set to 3.75 GHz by right clicking on the “Large Signal” schematic and choosing Options and then the Frequencies tab.

• Tone 2 applied to PORT1 defines the RF frequency and power.  The frequency is defined explicitly as 4.25 GHz.


Various graphs are created for various measurements of which a few of them are explained below.

Mixer Conversion Loss Graph

This graph shows the conversion loss that is calculated using both the 2-tone and single-tone simulations in "Large_Signal_Test_Bench" and "Noise_Simulation_Test_Bench".  For the 2-tone simulation, the large signal s-parameter measurement, LSSnm is used, in which the IF output signal at PORT_3 is specified as:

       -1 * tone1 frequency  + 1 * tone2 frequency, or

       -3.75GHz  + 4.25 GHz = 0.5 GHz

and the input signal at PORT_1 (RF) is specified as:

       0 * tone1 frequency  + 1 * tone2 frequency, or

       0 GHz  + 4.25 GHz = 4.25 GHz

The LO (PORT3) power sweep is made the x-axis for the measurement by selecting Use for X axis from the pull-down list under PORT_2.  The frequency of the LO is set to the schematic frequency (see above) by selecting Sweep Freq (FDOC) using the small up and down buttons next to the frequency pull-down.  In contrast, the Lange Coupler s-parameter measurements use the project frequency settings.

For the single-tone simulation, the NLNOISE block sets the IF frequencies (0.5 GHz to 0.6 GHz) as the sideband offsets.  The ConvG measurement selects the first frequency index, corresponding to 0.5 GHz, for the sideband offset.  It then identifies the IF output frequency at port 2 as the "Upper" sideband of DC, and the RF input frequency at port 1 as the "Upper" sideband of the LO fundamental frequency. That is:

       IF output frequency = 0 * tone1 frequency + 0.5 GHz = 0.5 GHz

       RF input frequency = 1 * tone1 frequency + 0.5 GHz = 4.25 GHz

Mixer IF Output Power Spectrum Graph

This graph shows the mixer IF Output Power Spectrum at a single LO Drive level.  The LO drive level can be adjusted using the Tuner (Simulate > Tune) for post simulation tuning.

Phase Noise vs Offset from IF Graph

The PH_Noise_NL_F measurement is used to set the output harmonic of interest.  So this will simulate at the IF frequency of 0.5 GHz (set by the measurement harmonics) with offsets from this frequency of 1e-5 to 1e-1 GHz (set by the NLNOISE block).  The "Plot vs. Offset Frequency" option in the measurement is used to subtract the IF frequency from the x-axis values, and the results are thus plotted vs. the offset frequency, which is displayed logarithmically for easier viewing.  The result is shown in the "Phase Noise vs Offset from IF" graph.

Output Phase Noise Contributors Graph (Table)

This table shows the contributions of noisy elements to the output phase noise in the Phase_Noise_Test_Bench schematic.  The ModNoiseCon_dBc measurement is used, so the noise contribution of each element at the specified offset (Noise Frequency Index) is provided relative to the IF output power level.