Harmonic Balance for Mixers

This topic discusses the use of the Harmonic Balance simulator for standard mixer analyses and the small signal option in the Harmonic Balance dialog box that simplifies mixer analyses. It also discusses noise simulations for mixers.

The Small-signal option makes it possible to do a multitone analysis where sidebands are involved, as in a mixer analysis. By reducing by 1 the number of large-signal tones required in a conventional harmonic balance analysis, and eliminating the need to calculate large-signal products, the simulation is computationally efficient.

The sections focusing on setting up noise simulations for mixers describe how to calculate:

There are two methods of noise simulation. One is to use the parameters on the Noise tab in the harmonic balance dialog box. A second method is to use NoiseCons. Using NoiseCons, you can set up several noise simulations, thereby eliminating the need to change the values on the Noise tab in the Harmonic Balance dialog box. With NoiseCons, you can also set parameters to calculate phase noise.

The starting point of this topic assumes you are familiar with the basics of harmonic balance simulation, as described in Harmonic Balance Basics.

For details on harmonic balance for mixers and small-signal mixer analysis, see the following topics:

When you are familiar with basic mixer simulation, see the following topics for details on mixer noise simulation:

Performing a Basic Mixer Simulation

For a successful analysis:

For details about other fields, click Help from the dialog box.

Examples of Mixer Simulations

This section gives detailed setups to perform harmonic balance simulations for:

Note
The Harmonic Balance simulation uses the Harmonic Balance Simulator license (sim_harmonic) which is included with all Circuit Design suites except RF Designer. You must have this license to run Harmonic Balance simulations. You can work with examples described here and installed with the software without the license, but you will not be able to simulate them.

Finding Mixer Output Tones

The next figure below illustrates an example setup for determining the amplitudes and relationships of mixer output tones.

Note
This design, Mix1.dsn, is in the Examples directory under Tutorial/SimModels_prj. The results are in Mix1.dds. Refer also to the ADS examples in /examples/RFIC/Mixers_prj.

This example is a two-tone harmonic balance analysis. One tone is used for the RF, one for LO. At the IF we expect to find two fundamental tones at f RF - f LO and f RF + f LO. We also expect to find spurious tones (attributed in this case to an intermodulation table). The mixer in this example is a MixerIMT component (available in the System-Data Models palette), which references the intermodulation table dbl1.imt. Conversion gain is determined by the parameter ConvGain, which uses the function dbpolar to set a gain of 6 dB at an angle of 0 degrees. The reflection parameters S 11, S 22, and S 33 have been set to 0.1, 0.33, and 0.1, respectively.

Note
By convention, the mixer input port is port 1, the IF output port is port 2, and the LO input port is port 3. This is different from the automatic component-naming convention used by the P_1Tone sources when they are placed sequentially in the Schematic window.

You can leave SS_Sideband set to its default value of UPPER for this simulation.


Setup for finding output tones

To find the amplitudes and relationships of mixer output tones:

  1. From the Sources-Freq Domain palette, select P_1Tone and place one instance of it at the RF input (PORT1) and another at the LO input (PORT2).
  2. Edit the RF input source:
    • Num = 1. This is port 1.
    • Z = 50 Ohms (default value)
    • P = dbmtow(-10). This converts the 0-dBm input to watts (the power unit used by the system).
    • Freq = RFfreq. This variable will be defined by a VarEqn component.
  3. Edit the LO input source:
    • Num = 2. This is port 2.
    • Z = 50 Ohms (default value)
    • P = dbmtow(7)
    • Freq = LOfreq. This variable will be defined by a VarEqn component.
    • Noise = No. Unless you are doing a noise simulation, this parameter is ignored.
  4. From the Data Items palette, select Var eqn (Variables and Equations), and place the VAR component on the schematic. Edit it and select the default equation, replacing it with the following equations:
    • LOfreq = 1850 MHz
    • RFfreq = 2100 MHz
  5. From the Simulation-HB palette, select the HB simulation component and place it on the schematic. Edit to select the Freq tab and edit the following parameters:
    • Maximum order = 8
    • Frequency = LOfreq. This is Freq[1]. Set its order to 8.
    • Click Add to enter the second fundamental, Freq[2]. Set its frequency to RFfreq, and leave its order set to 8.
      Note
      Normally you would want to set orders for the RF tones lower, to achieve a faster simulation.
  6. Simulate. When the simulation is finished, plot Vif in dBm, then place markers on the two fundamental IF tones, RFfreq - LOfreq (m1) and RFfreq + LOfreq (m2), as shown below:

    Place more markers to view other frequencies and their magnitudes.

Performing a Small-Signal Simulation of a Mixer

The following figure, similar to the preceding example, illustrates an example setup for performing a small-signal mixer analysis.

Note
This design, Mix1_ssmix.dsn, is in the Examples directory under Tutorial/SimModels_prj. The results of the simulation are in Mix1_ssmix.dds.
For a detailed example refer to ADS examples/RFIC/MixerDiffMode_prj.


Small-signal mixer analysis example

The Small-signal option, which takes less time and uses less memory, makes it possible to do a multitone analysis where sidebands are involved. When you reduce by one the number of large-signal tones required in a conventional harmonic balance analysis (thereby eliminating the need to calculate large-signal products), the simulation is computationally efficient, because it treats the RF signal as a small-signal, and no harmonics of the RF signal are generated, yet it still forms mixing terms with the LO. This procedure requires that a P_1Tone component be placed where you want to simulate the effects of mixing a fundamental with its sidebands.

This procedure uses the upper sideband as an example only. Modify it to view the effects of mixing with the lower sideband. This basic example uses only one RF tone. With two RF tones defined in the RF source, this setup can also be used for a multitone analysis. Small-signal mixer data are indicated by SM.

To perform a small-signal analysis of a mixer:

  1. Start with the circuit used in the example, Performing a Small-Signal Simulation of a Mixer.
    Note
    The Freq parameter of the P_1Tone component should be empty, as it will be ignored in this analysis. If you simulate or modify the Mix1_ssmix.dsn example shown in the previous figure, be sure to change Freq = None.
  2. Edit the HarmonicBalance simulation component as follows:
    • Select the Freq tab and ensure that Frequency is LOfreq and its Order is 4. LOfreq (Freq[1]) is defined in the VAR component.
    • Select the Small-Sig tab and ensure that Sweep Type is Single point, and that Frequency is IFfreq (this is the IF resulting from RFfreq - LOfreq). Also select Use all small-signal frequencies.
      Note
      Because we are specifying the power in the upper sideband in the RF source (PORT1), the frequency for this source is the sum of the difference frequency (250 MHz) and the LO frequency (1850 MHz). The RF input is therefore 2100 MHz.
  3. Simulate. When the simulation is finished, place a Rectangular plot of SM.Vif in dBm, as follows:

    Markers at 250 MHz (f IF) and 3.750 MHz (f LO + f RF) indicate the effects-for our idealistic model-of 6 dB of gain on the input RF power (-60 dBm).

Determining Mixer Conversion Gain

The following figure illustrates an example setup for determining mixer conversion gain.

Note
This design, Mix1_convgain.dsn, is in the Examples directory under Tutorial/SimModel_prj. The results are in Mix1_convgain.dds.


Example determining mixer conversion gain

This example uses the basic Mixer component from the System-Amps & Mixers palette, and also illustrates the use of the Krylov method. While a small-signal mixer analysis can be used to determine conversion gain, the method described here, in which the RF signal is not assumed to be small, should be used if the RF signal is large enough to drive the mixer into compression.

To determine mixer conversion gain:

  1. Start with the circuit in the example, Finding Mixer Output Tones.
  2. Edit the P_1Tone component at the RF source (PORT1)s:
    • P= dbmtow(Power_RF). Power_RF will be established by an equation.
    • Freq = RFfreq
  3. Edit the P_1Tone component at the LO source (PORT2):
    • P= dbmtow(0)
    • Freq = LOfreq
  4. Using the VAR component, enter the following equations (note the use of the underscore before dBm):
    RF_power = -30 _dBm
    LO_power = -10 _dBm
    LOfreq = 1850 MHz
    IFfreq = RFfreq - LOfreq
    RFfreq = 2100 MHz
    Note
    In ADS, dBm is not a recognized unit. The underscore in front of dBm (as in _dBm ) allows this unit to be used for documentation purposes without causing the simulator to issue a warning.
  5. Edit the Harmonic Balance Simulation component, select the Freq tab, and edit the following parameters:
    • Maximum order = 5
    • Freq[1] = LOfreq
    • Copy and paste the frequency above, creating Freq[2], the RF frequency. Then edit it so that its frequency is RFfreq and its order is 3.
  6. Scroll to the Solver tab, then select Krylov Solver.
    Hint
    To pass the variable RF_power to the dataset for subsequent manipulation by equations, use the Output tab. Refer to Selectively Saving and Controlling Simulation Data.
  7. Simulate. When the simulation is finished, do the following:
    • Plot Mix(1) and Mix(2) in List format. This will show you the coefficients that generated the sum and difference tones. In this example we are interested in the fundamental IF tone f LO + f RF. As indicated in the area outlined in the next figure below, these coefficients are 1,1.
      Note
      Inserting a listing column of the frequencies in the simulation also makes it possible to determine which index number to use to select a particular spectral component. For example, if you have a mixer that is upconverting to f LO + f RF, and you want to select that spectral component of Vif, then you select Vif[7], because f LO + f RF (= 3.75 GHz) is the seventh frequency in the array.
    • Using the mix function, write an equation in the Data Display window to find the power in dBm at f LO + f RF, which we know to be 3.750 GHz. That equation is:
      VIFtone = dBm(mix(Vif,{1,1}))
    • Write an equation in the Data Display window that expresses conversion gain as the difference between IF power ( VIFtone ) and RF_power:
      MixConvGain = VIFtone - RF_power[0]
      Any number could be used as the tonal index to the right of RF_power, as this value is a constant.
      Note
      The above equations could also be placed as MeasEqn components in the Schematic window. This would allow MixConvGain to be optimized. For a discussion of optimization, refer to Tuning, Optimization, and Statistical Design.

      The following figure shows the mixing indexes, the equations, and a List plot of MixConvGain. This figure, 6.000, is precisely the value assigned to the parameter ConvGain in MIX1.


Mixer conversion gain indexes, equations, and plot

Determining Mixer Intermodulation Distortion

Teh following figure illustrates an example for determining mixer intermodulation distortion (IMD). The approach used here is commonly referred to as third-order intercept (TOI) analysis.

Note
This design, Mix2_TOI.dsn, is in the Examples directory under Tutorial/SimModels_prj. The results are in Mix2_TOI.dds.

This example is similar to Performing a Small-Signal Simulation of a Mixer, except that it involves three tones as well as equations that determine Freq[2] and Freq[3].


Mixer intermodulation distortion example

To determine mixer intermodulation distortion:

  1. Start with the circuit in the example, Performing a Small-Signal Simulation of a Mixer.
  2. Edit the P_nTone component at the RF source (PORT1) as follows:
    • P[1]= dbmtow(Power_RF). Power_RF will be established by an equation.
    • P[2]= dbmtow(Power_RF)
      Note
      The indexes [1] and [2] are local to this source component only, and are not related to Freq[1] or Freq[2] in the Harmonic Balance Simulation component.
  3. Edit the P_1Tone component at the LO source (PORT2) as follows:
    • P= dbmtow(-5)
    • Freq = 1850 MHz
  4. Enter the following equations in the VAR component:
    Power_RF = -50 _dBm
    RF_Freq = 2100 MHz
    F_Spacing = 100 kHz
  5. Edit the Harmonic Balance Simulation component, select the Freq tab, and edit the following parameters:
    • Maximum order = 8
    • Frequency = 1850 MHz. This is Freq[1], the LO frequency. Set its order to 5.
    • Copy and paste the frequency above, creating Freq[2], the RF frequency. Then edit it so that its frequency is RF_Freq+F_Spacing/2 and its order is 3.
    • Copy the above to create Freq[3], and edit so that its frequency is RF_Freq-F_Spacing/2. Its order is also 3.
  6. Scroll to the Solver tab, then select Krylov Solver.
  7. Simulate. When the simulation is finished, plot the following:
    • Plot Vif, in dBm, in Rectangular format as shown below. A marker placed on the largest tone (f LO + f RF) confirms its frequency on the first line, its power on the second.
    • Plot Vif again, this time changing the scale of the x-axis as follows:
      Select the plot, then choose Edit > Item Options (or double-click). When the window opens, click the Plot Options tab.
      Deselect Auto Scale On/Off, and make sure that xAxis is selected. Set Min = 3.7498e9, Max = 3.7502e9, Step = 4e5, and click OK. The following will appear, showing the tones centered around the fundamental IF tone (f LO + f RF) at 3.750 GHz:

Determining Mixer Noise

The next figure below illustrates an example for determining mixer noise. It is based on the example, Finding Mixer Output Tones.

The additional steps in this example configure the simulation for noise analysis. The MixerIMT2 parameters are the same as those in the previous example, except that NF (double sideband noise figure in dB) has been set to 3 dB. The noise figure computed by a harmonic balance noise simulation is a single sideband noise figure.

Note
Port sources and an output termination (Term component) are necessary only if a noise figure simulation is being performed. The model used in the current design is a 2-port model, and so ignores LO noise. In a circuit-based model, take LO noise into account.

To determine mixer noise:

  1. Start with Finding Mixer Output Tones.
  2. From the Simulation-HB palette, select a Term, place it at the output, and edit it:
    • Num = 2. This is port 2.
    • Z = 50 Ohms (default value)
    • Noise = No

      Determining mixer noise example
  3. Edit the HarmonicBalance simulation component. In the dialog box, select the Noise tab, enable Nonlinear noise, then click Noise (1) to edit the settings in the following fields:
    • Sweep Type = Single Point
    • Frequency = RFfreq+LOfreq (f RF + f LO). This is the mixer output frequency.
    • Input frequency = RFfreq (frequency of the RF input)
    • Noise input port = 1 (the RF input port)
    • Noise output port = 2 (the output or Term port)
  4. From the Simulation-HB palette, select an Options component and place it on the schematic. Set Temp to 16.85.
    Note
    In nonlinear noise analyses, we recommend that the Options component be used to establish a global simulation temperature of 16.85°C if a noise calculation is to be performed. This can be done by editing Temp=16.85 in the Schematic window, or by selecting the Misc tab and editing Simulation temperature to that value. This temperature, 16.85°C or 290 K, is the standard temperature for noise figure measurement as defined by the IEEE definition for noise figuration.
  5. Simulate. When the simulation is finished, plot nf(2) and te(2) in dB, as shown below. These are the single sideband noise figure and equivalent noise temperature (in Kelvin) at port 2, respectively.

    The single sideband noise figure of 6.3 dB indicates that the mixer is adding noise based on the MixerIMT2 component setting NF = 3 dB (double sideband).

Simulating Mixer Noise with NoiseCons

When you use a NoiseCon nonlinear noise controller with a mixer, you have more flexibility in controlling how the noise simulation is performed than you do with Noise(1) and Noise(2). You can compute noise figure, noise at different nodes and frequencies, and integrate the effects of oscillator phase noise.

The next figure below shows a mixer that will be used to demonstrate the use of NoiseCons components with a mixer.

Note
Refer to the ADS project examples/Tutorial/Noisecon_prj for an illustration of how to use NoiseCons for mixer noise simulation. See mixer.dsn for the design and mixer.dds for the results.


Example demonstrating NoiseCons with a mixer

Mixer Noise Figure

The NoiseCon components, NC_USB and NC_LSB, are used to compute the noise figure of the mixer. NC_USB is used to compute the noise figure to the up conversion (RF+LO) frequency. NC_LSB is used to compute the noise figure of the down conversion (RF-LO) frequency.

  1. From the Simulation-HB palette, select a NoiseCon, place it in the design and name it NC_USB.
  2. Select the Freq tab, select Single point for the frequency and set it to 1750 MHz, the up-conversion frequency for this mixer.
  3. Select the Misc tab, set Noise input port to 1 and Noise output port to 2. Set the Input frequency to 900 MHz.
  4. Place another NoiseCon and name it NC_LSB.
  5. Select the Freq tab, select Single point for the frequency and set it to 50 MHz, the down-conversion frequency for this mixer.
  6. Select the Misc tab, set Noise input port to 1 and Noise output port to 2. Set the Input frequency to 900 MHz.
  7. Edit the Harmonic Balance controller and select the NoiseCons tab. Select NC_USB and NC_LSB and Add them to the list of NoiseCons to be simulated.

The simulation results are shown next.

These values are slightly different due to high-frequency rolloff in the mixer.

Noise at Different Nodes and Frequencies

The NoiseCon components, NC_RF and NC_IF, are used to compute noise voltage at different nodes and frequencies in one simulation. NC_RF also illustrates the use of a differential noise measurement.

  1. From the Simulation-HB palette, select a NoiseCon, place it in the design and name it NC_RF.
  2. Select the Freq tab, select Single point for the frequency and set it to 900 MHz, the RF frequency for this mixer.
  3. Select the Nodes tab to compute both noise at a single node and differential noise between two nodes. The nodes mix1.RFampp and mix1.RFampn are two internal nodes in the mixer subnetwork at the output of the internal RF preamplifier.
    • First, select the node name mix1.RFampn from the Pos Node drop-down list. Make sure the Neg Node entry field is blank. Now click Add to create the first entry in the list of nodes. This computes the noise at one node.
    • Next, select the node name mix1.RFampp from the Pos Node drop-down list and select the node name mix1.Rfampn from the Neg Node drop-down list. Click Add to create the second entry in the list of nodes. This computes the differential noise between two nodes.
  4. Select the Misc tab and erase the default values for Noise input port and Noise output port so the noise figure is not computed with this NoiseCon.
  5. Place another NoiseCon and name it NC_IF.
  6. Select the Freq tab, select Single point for the frequency and set it to 50 MHz, the IF frequency for this mixer.
  7. Select the Nodes tab, enter the node name vif for Pos Node and click Add.
  8. Select the Misc tab, erase the default entries for Noise input port and Noise output port so the noise figure is not computed with this NoiseCon.
    Note that this could have also been done on the NC_LSB NoiseCon.
  9. Edit the Harmonic Balance controller and select the NoiseCons tab. Select NC_RF and NC_IF and Add them to the list of NoiseCons to be simulated.

Note the following when viewing the results.

Effects of LO Phase Noise on Noise Figure

You can set up an analysis, using a NoiseCon, that includes the effect of LO phase noise on the noise figure of the mixer. The figure above, Determining mixer noise example, shows the setup to be used. This design is similar to the first design except the P_1Tone ideal LO source has been replaced with an OSCwPhNoise source that includes phase noise.

Note
Refer to the ADS project examples/Tutorial/Noisecon_prj for an illustration of how to use NoiseCons for mixer noise simulation. See mixer_pn.dsn for the design and mixer_pn.dds for the results.


Setup using NoiseCons with a mixer and effects of LO phase noise

Two NoiseCons are used to set up this simulation. The first computes the mixer noise figure without phase noise. This is done by setting the mixer input frequency and (output) noise frequency to be exactly equal to the large signal frequencies. No LO phase noise is considered. The second computes the mixer noise figure including the phase noise from the LO. The input and output frequencies will be the same, but the total noise power at the output will be obtained by integrating over a bandwidth of 30 kHz centered around the IF frequency.

  1. From the Simulation-HB palette, select and place a NoiseCon and name it NC1. This will be set up the same as the NC_LSB that was used previously.
  2. Select the Freq tab, select Single point for the frequency and set it to 50 MHz, the output frequency for this mixer.
  3. Select the Misc tab, set the Noise input port to 1 and the Noise output port to 2. Set the Input frequency to 900 MHz.
  4. Place another NoiseCon and name it NC2. This will be used to compute the mixer noise figure including the effects of phase noise integrated over a 30 kHz bandwidth.
  5. Select the Freq tab to set up the frequency that will serve as the offset frequency for the phase noise analysis that will be swept over the 30 kHz bandwidth. The mixer output frequency for this analysis will be set later on the PhaseNoise tab. Select a Log sweep, set the Start frequency to 1 Hz and the Stop frequency to 100 kHz. Set the Pts./decade to 3.
  6. Select the Misc tab, set the Noise input port to 1 and the Noise output port to 2. Set the Input frequency to 900 MHz.
  7. Select the PhaseNoise tab, select a Phase Noise Type of Integrate over bandwidth. The phase noise carrier frequency, which for this example is the output frequency of the mixer, is specified using Carrier mixing indices. For an IF frequency of 50 MHz, the mixing indices of interest is 1,-1. Enter the first Index of 1 and click Add ; then enter the second Index of -1 and click Add.
  8. Edit the Harmonic Balance controller and select the NoiseCons tab. Select NC1 and NC2 and Add them to the list of NoiseCons to be simulated.

The simulation results are shown next.

Note that the second value is larger than the first, demonstrating the degradation of the mixer noise figure due to the extra noise injected by the LO phase noise.

A standard noise figure computation is performed using the IEEE standard definition of single sideband noise figure:

where

k is Boltzmann's constant (1.389658x10 -23 )
T 0 is the IEEE standard temperature for noise figure (290 K)
G 1 is the conversion gain of the mixer
G 2 is the image conversion gain of the mixer
G 3,..., G n are conversion gains of higher order mixing products
R is the resistance of the output termination
v n is the noise voltage at the output port at the output frequency where the input and output terminations do not contribute any noise

When the effects of phase noise are included, the v n 2 (f)/R term is replaced with total noise power at the output integrated over a bandwidth B centered at the output IF frequency fIF:

In a mixer, the LO will generate noise over a range of frequencies from f LO - B/2 to f LO + B/2. These will mix with the RF tone at f RF to produce noise from f IF -B/2 to f IF + B/2, where f IF = |f RF - f LO|.

The single sideband noise figure is computed with the input port specified by the Noise input port parameter, and the output port specified by the Noise output port parameter. It is saved in the dataset as the variable nf(i), where i is the number of the noise output port. ADS also saves this value as NFssb in the output dataset.

ADS also computes the double sideband noise figure using the following equation:

This is the value that would be computed by a noise figure meter. It is saved in the dataset as NFdsb.

Small-Signal Mode Description

When the Small-signal mode option is selected, the simulator uses the frequency (or frequencies, in the case of a sweep) entered under the Freq tab to perform a normal large-signal simulation. It then takes the resulting data and performs a small-signal mixer simulation at all the original large-signal frequencies plus and minus the small-signal frequency or frequencies. The simulator essentially linearizes the circuit around the precomputed large-signal operating point.

All the time-varying nonlinearities that result from the presence of large signals are simulated, but with the assumption that the small signal does not change these nonlinearities. For example, a mixer will appear to be perfectly linear with respect to the amplitude of the small-signal input, but all the frequency translation terms will be computed. If a sweep is being performed, then the large-signal analysis is performed only once, followed by the required number of small-signal analyses.

The following assumptions and restrictions apply to mixer analysis:

Small-Signal Noise Simulation

The small-signal noise simulator is a subset of the nonlinear noise simulator. (For a discussion of nonlinear noise, refer to Nonlinear Noise Simulation Description.) In a small-signal noise analysis, there is no frequency translation of the noise. Since the noise is assumed to be a small-signal perturbation, there can be no noise mixing without a large-signal harmonic balance source. The only mixing frequencies that are accounted for are the upper and lower sidebands of the normal harmonic-balance analysis.

Nonlinear Spot-Noise Simulation

A nonlinear spot-noise simulation computes the noise at a specific noise frequency, as a function of another variable (such as LO power). For noise figure to be calculated, noise must be injected at the input port. The noise input frequency is entered into the Input frequency field of the Noise (1) dialog box. Since frequency translation occurs in a mixer, you must specify at which input frequency noise is to be injected.

For example, for a downconverter you would specify

InputFreq = noisefreq + LOfreq

where LOfreq is the local oscillator frequency input to the mixer.

The downconverter will shift INPUTFREQ to

InputFreq - LOfreq = (noisefreq + LOfreq) - LOfreq = noisefreq

For an upconverter, INPUTFREQ would be

InputFreq = noisefreq - LOfreq

Single-Frequency Nonlinear Spot Noise

The simulation results of a single-point analysis have no independent variable and therefore cannot be plotted on a rectangular plot, however, the data can be displayed in a listing column. Note that noise voltages are always rms.

If the amplitude of the RF signal is small compared to that of the LO, then the noise mixed by the LO and its harmonics will dominate, and simulations can be run much more quickly by simply including the LO and its harmonics in the harmonic balance noise analysis.

Multiple-Frequency Nonlinear Spot Noise

A sweep can also be used to compute noise at the specified noise frequency as a function of a variable. If an independent variable is in the nonlinear noise data, the data can be plotted on a rectangular plot.

Swept-Noise Simulation

A nonlinear swept-noise simulation computes the noise at a swept list of noise frequencies with another optional swept variable (for example, LO power). The result is analogous to a spectrum analyzer display. Nonlinear swept-noise analysis differs from nonlinear spot-noise analysis in that the noise frequency (in the Noise frequency area in the Noise (1) dialog box) is allowed to be swept in a manner similar to the way a spectrum analyzer sweeps a noise floor.

The noise frequency can be an expression if the noise frequency is swept. For example, to compute the noise figure of an upper sideband mixer in which the noise output frequency is swept but the LO frequency is fixed, set the parameter Input frequency as follows:

InputFreq = noisefreq + LOfreq

where noisefreq is the noise frequency and LOfreq is an equation defining the fixed LO frequency.

The results of the analysis can be plotted on a rectangular plot. The independent variable of the swept nonlinear noise analysis is always noisefreq.

 

Privacy Statement  | Terms of Use  | Legal | Contact Us  | © Agilent 2000-2008 

Contents
Additional Resources