Mixer2 (RF System Mixer)
Symbol
Available in ADS and RFDE
Parameters
Name 
Description 
Units 
Default 

SideBand 
Specify the sideband/image option for the mixe: 
None 
BOTH 
OutputSidebandSuppression 
Output sideband suppression (only relevant for SideBand=LOWER, UPPER) 
dB 
−200 
InputImageRejection 
Input image rejection (only relevant for SideBand=LOWER IMAGE REJECTION, UPPER IMAGE REJECTION) 
dB 
−200 
ConvGain 
Conversion gain; use x + j × y, polar(x,y), dbpolar(x,y) for complex value 
None 
dbpolar(0,0) 
RevConvGain 
Reverse conversion gain, use x + j × y, polar(x,y), dbpolar(x,y) for complex value 
None 
polar(0,0) 
SP11 
S11, RF port reflection, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP12 
S12, IF port to RF port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP13 
S13, LO port to RF port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP21 
S21, RF port to IF port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP22 
S22, IF port reflection, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP23 
S23, LO port to IF port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP31 
RF port to LO port leakage (real or complex number)S31, RF port to LO port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value 
None 
polar(0,0) 
SP32 
IF port to LO port leakage (real S32, IF port to LO port leakage, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex value or complex number) 
None 
polar(0,0) 
SP33 
LO port reflection (real or complex numS33, LO port reflection, use x + j × y, polar(x,y), dbpolar(x,y), vswrpolar(x,y) for complex valueor number) 
None 
polar(0,0) 
PminLO 
Minimum LO power before starvation 
dBm 
−100 
DetBW 
Detector bandwidth for LO limiting 
Hz 
1e100 
NF 
Double sideband noise figure 
dB 
None 
NFmin 
Minimum double sideband noise figure at Sopt 
dB 
None 
Sopt 
Optimum Source Reflection for Minimum Noise Figure, use x + j × y, polar(x,y), dbpolar(x,y) for complex value 
None 
None 
Rn 
Equivalent noise resistance 

Z1 
Reference impedance for port 1 (must be a real number) 
50 

Z2 
Reference impedance for port 2 (must be a real number) 
50 

Z3 
Reference impedance for port 3 (must be a real number) 
50 

GainCompType 
Gain compression type: 
None 
LIST 
GainCompFreq 
reference frequency for gain compression if gain compression is described as a function of frequency 


ReferToInput 
Specify gain compression with respect to input or output power of device: 
None 
OUTPUT 
SOI 
Second order intercept 
dBm 
None 
TOI 
Third order intercept 
dBm 
None 
Psat 
Power saturation point (always referred to output, regardless of the value of the ReferToInput parameter) 
dBm 
None 
GainCompSat 
Gain compression at Psat 
dB 
5.0 
GainCompPower 
Power level in dBm at gain compression for X dB compression point, specified by GainComp 
dBm 
None 
GainComp 
Gain compression at GainCompPower 
dB 
1.0 dB 
AM2PM 
Amplitude modulation to phase modulation 
deg/dB 
None 
PAM2PM 
Power level at AM2PM 
dBm 
None 
GainCompFile 
Filename for gain compression data in S2D file format 
None 
None 
ClipDataFile 
Clip data beyond maximum input power: YES=enable, NO=disable 
None 
YES 
ImpNoncausalLength 
Noncausal function impulse response order 
Integer 
None 
ImpMode 
Convolution mode 
Integer 
None 
ImpMaxFreq 
Maximum frequency to which device is evaluated 
GHz 
10 
ImpDeltaFreq 
Sample spacing in frequency 


ImpMaxOrder 
Maximum allowed impulse response order 
Integer 
32 
ImpWindow 
Smoothing window 
Integer 
0 
ImpRelTol 
Relative impulse response truncation factor 
None 
None 
ImpAbsTol 
Absolute impulse response truncation factor 
None 
None 
Frequently Asked Questions
Q1 : What are the major differences between Mixer and Mixer2?
A1 : Refer to note 1.
Q2 : What are the supported parameter combinations?
A2 : Refer to Range of Usage .
Q3 : What is the range of usage for each parameter combination?
A3 : Refer to Range of Usage.
Q4 : What are the port numbers and the terminology used for the signals at each port?
A4 : Refer to Terminology.
Q5 : What is the basic implementation of Mixer2?
A5 : A Noisy2Port cascaded with an SDD . Refer to Basic Implementation for details.
Q6 : How does Mixer2 operate for the various SideBand modes?
A6 : Refer to Sideband Suppression and Image Rejection.
Q7 : What is ConvGain?
A7 : It is the conversion gain. Refer to Linear Behavior for details.
Q8 : How does Mixer2 apply a complex ConvGain?
A8 : Before the mixing, not after. Refer to Linear Behavior for details.
Q9 : What is RevConvGain?
A9 : It is the reverse conversion gain. Refer to Linear Behavior for details.
Q10 : What are SPij (i,j=1,2,3)?
A10 : They are the reflection/leakage/isolation parameters or Sparameters. Refer to Linear Behavior for details.
Q11 : Are ConvGain, RevConvGain and SPij (i,j=1,2,3) wavebased or voltagebased?
A11 : ConvGain is voltagebased, RevConvGain and SPij are wavebased. Refer to Linear Behavior for details.
Q12 : How is RF to IF compression modeled?
A12 : A polynomial compression model is used. Refer to Nonlinear Behavior and Modeling Basics of the Amplifier2 documentation for details.
Q13 : How is AM to PM conversion modeled?
A13 : A polynomial AM to PM conversion model is used. Refer to AM to PM Conversion of the Amplifier2 documentation for details.
Q14 : How is LO limiting done for different DetBW values?
A14 : Via a filter with a variable bandwidth specified by DetBW or via a Hilbert transform. Refer to LO Limiting for details.
Q15 : What's best, small or large DetBW values?
A15 : Small for LO signals with significant harmonic content, large for bandpass LO signals. Refer to LO Limiting for details.
Q16 : How does PminLO influence the LO limiting?
A16 : It defines a lower limit for the effect of the LO power on the mixer's conversion gain. Refer to LO Limiting for details.
Q17 : Why does the mixer's conversion gain vary depending on the frequency content of the LO signal?
A17 : LO limiting is dynamic. Refer to LO Limiting for details.
Q18 : What noise model is used by Mixer2?
A18 : A Noisy2port , similar to Amplifier2. Refer to Noise for details.
Q19 : What are "NFonly mode" and "(NFmin,Sopt,Rn) mode" for in Mixer2?
A19 : They are two different ways of specifying noise. Refer to Noise for details.
Q20 : How is NFssb/NFdsb calculated, both at low powers and in compression?
A20 : Refer to Noise.
Q21 : Can you provide more details about the noise voltages and noise figures produced by Mixer2 for the different SideBand modes of operation?
A21 : Refer to Noise.
Q22 : How can I lower the noise at the output of Mixer2?
A22: Refer to Noise.
Q23 : Is there a tutorial example for Mixer2?
A23: examples/Tutorial/Mixer2_Example_prj . Refer to note 2 for details.
Q24 : Does Mixer2 support frequency conversion AC (FCAC) analysis?
A24 : No. Refer to note 3 for details.
Q25 : Why doesn't Mixer2 work for complex reference impedances?
A25 : Mixer2 does not support complex reference impedances. Refer to note 4 for details.
Q26 : Can I use Mixer2 for baseband envelope simulations?
A26 : No. Refer to note 6 for details.
Q27 : Why is Mixer2 sometimes slower than Mixer?
A27 : This is a consequence of the implementational differences between Mixer and Mixer2. Refer to note 7 for details.
Q28 : Why don't the predicted second and thirdorder intercept points match SOI and TOI as set on Mixer2?
A28 : You are probably not setting up your validation correctly. Make sure to do a twotone simulation, not a onetone simulation. Refer to note 10 for details.
Q29 : Do sums/differences of two RF frequencies appear at the output of Mixer2?
A29 : It depends on SP21 and SOI. Refer to note 11 for details.
Q30 : Why does Mixer2 ignore the ACDATA block of my S2D file?
A30 : It uses the Sparameters on the component instead. Refer to note 12 for details.
Q31 : When the power range in an S2D file differs from that of a simulation, which power range is used for polynomial fitting?
A31 : The power range in the S2D file. Refer to note 14 for details.
Q32 : Why do I get ringing at high powers when using an S2D file?
A32 : This is a consequence of the polynomial model adopted by Mixer2. Refer to note 15 for details.
Q33 : How do I get rid of this ringing?
A33: Eliminate data for high input powers, rely on extrapolation via ClipDataFile=yes, or break the S2D file into two. Refer to note 15 for details.
Range of Usage
NF ≥ 0 dB
NFmin ≥ 0 dB
0 < Sopt < 1
Rn > 0
GainCompFreq > 0
0 dB < GainComp < 3 dB
When specifying gain compression using model parameters, only certain combination of parameters will produce stable polynomial curve fitting. If unrealistic parameter values are used, the polynomial will become unstable, resulting in oscillations. The recommended parameter combinations are listed here:
 Thirdorder intercept parameter:
Parameters: TOI
Range of validity: N/A  Gain compression parameters:
Parameters: GainCompPower, GainComp
Range of validity: N/A  Power saturation parameters:
Parameters: Psat, GainCompSat
Range of validity: N/A  Thirdorder intercept and 1dB gain compression parameters:
Parameters: TOI, GainCompPower with GainComp=1dB
Range of validity: TOI > GainCompPower + 10.8  Thirdorder intercept and power saturation parameters:
Parameters: TOI, Psat, GainCompSat
Range of validity: TOI > Psat + 8.6  1dB gain compression and power saturation parameters:
Parameters: GainCompPower with GainComp=1dB, Psat, GainCompSat
Range of validity: Psat > GainCompPower + 3  Thirdorder intercept, 1dB gain compression and power saturation parameters:
Parameters: TOI, GainCompPower with GainComp=1dB, Psat, GainCompSat
Range of validity: Psat > GainCompPower +3, TOI > GainCompPower + 10.8  Secondorder intercept and thirdorder intercept parameters:
Parameters: SOI, TOI
Range of validity: N/A
Terminology
This section outlines the terminology that will be used in the remainder of the Mixer2 documentation.
Mixer2 is a 3port device. Its input port is numbered port 1, its output port is numbered port 2, and its local oscillator port is numbered port 3. The signals at the three ports will be denoted by RF, IF, and LO, respectively, as illustrated in Mixer Terminology.. This follows the receiver/downconversion convention. This convention has RF and IF flipped relative to the transmitter/upconversion convention. When Mixer2 is used for transmitter/upconversion applications, RF therefore refers to the input signal and IF refers to the output signal.
Mixer Terminology.
The primary signal of interest at the input of the mixer is the RF signal. A secondary signal of interest at the input of the mixer is the RF image, denoted RFimg, defined as the other input signal which results in IF as a secondorder mixing product. The input image is typically only of interest for receiver/downconversion applications, in which case it is given by RFimg=2*LORF. RF and RFimg=2*LORF both mix down with LO to generate IF=LORF at the output. The signals RF and RFimg are called images of each other with respect to LO. If RFimg is attenuated relative to RF, we talk about image rejection .
The primary signals of interest at the output of the mixer are LORF and LO+RF. Typically, one of these is desired and the other is undesired. We refer to these two signals as lower and upper sidebands . If one sideband is attenuated relative to the other, we talk about sideband suppression .
Throughout the Mixer2 documentation, we will use the terms image and rejection when describing input signals to Mixer2 and use the terms sideband and suppression when describing output signals from Mixer2. Other choices are possible but the terminology adopted here is in line with accepted industry literature, see for example the exhaustive mixer applications notes from WJ Communications. The only exception is that the Mixer2 SideBand parameter controls both the image and the sideband operation of the mixer, not just the sideband operation. The reason for this is compatibility with previous versions of Mixer2.
Basic Implementation
The basic implementation of Mixer2 is a Noisy2Port cascaded with an SDD. The parameters NF, NFmin, Sopt, Rn and Z1 are passed to the Noisy2Port and noise is generated. The signal is not affected by the generation of noise but goes unaltered through the Noisy2Port. All parameters except NF, NFmin, Sopt and Rn are then passed to the SDD and the signal and noise goes through the SDD. The SDD implements a linear mixer which allows compression from RF to IF and from RF to LO. Different circuit topologies are used for realizing sideband suppression and image rejection mixers, but the overall implementation is the same  a Noisy2Port cascaded with an SDD. This basic knowledge of the implementation may help understand some of the following sections.
Sideband Suppression and Image Rejection
A mixer can pass both the lower and the upper input image, reject the lower input image, or reject the upper input image. Similarly, a mixer can generate both the lower and the upper output sideband, generate the lower sideband only, or generate the upper sideband only. Various combinations of these are possible. To explain the image rejection and sideband suppression features for Mixer2, we will look at a receiver/downconversion example.
Given RF and LO for a downconverting mixer, we have IF=LORF. Also of interest is the RF image given by RFimg=2*LORF. If RF is below LO, RFimg is above LO. Similarly, if RF is above LO, RFimg is below LO. We will use RF1 to denote the signal below LO and RF2 to denote the signal above LO. If RF=RF1, then RFimg=RF2. Similarly, if RF=RF2, then RFimg=RF1. RF1 and RF2 both mix down to LORF1=RF2LO. RF1 and RF2 mix up to RF1+LO and RF2+LO. This is illustrated in Frequency Plan for Image Rejection and Sideband Suppression..
Frequency Plan for Image Rejection and Sideband Suppression.
We will use this to explain the five supported SideBand parameters and the use of OutputSidebandSuppression and InputImageRejection. Note that this illustration has a receiver/downconversion bias as LO~RF. For a transmitter/upconversion application, RF would generally be much closer to zero and LO would generally be much closer to LORF and LO+RF. However, all principles remain the same and we can use the above example to explain all SideBand options.
For SideBand=BOTH, both input images are passed through the mixer without rejection and both output sidebands are generated without suppression. This mode is the default for Mixer2. A normal circuit level mixer without external filtering operates in this manner. OutputSidebandSuppression and InputImageRejection are ignored for this mode. This is illustrated in Mixer2 image/sideband operation for SideBand=BOTH..
Mixer2 image/sideband operation for SideBand=BOTH.
For SideBand=LOWER, both input images are passed through the mixer without rejection and the undesired upper output sideband is suppressed by the amount given by OutputSidebandSuppression relative to the desired lower output sideband. This mode is a downconversion mode used for receiver applications. The default value for OutputSidebandSuppression is 200 dB. InputImageRejection is ignored for this mode. This is illustrated in Mixer2 image/sideband operation for SideBand=LOWER..
Mixer2 image/sideband operation for SideBand=LOWER.
For SideBand=UPPER, both input images are passed through the mixer without rejection and the undesired lower output sideband is suppressed by the amount given by OutputSidebandSuppression relative to the desired upper output sideband. This mode is an upconversion mode used for transmitter applications. The default value for OutputSidebandSuppression is 200 dB. InputImageRejection is ignored for this mode. This is illustrated in Mixer2 image/sideband operation for SideBand=UPPER..
Mixer2 image/sideband operation for SideBand=UPPER.
For SideBand=LOWER IMAGE REJECTION, the lower input image is rejected by the amount given by InputImageRejection relative to the upper input image and both output sidebands are suppressed by the amount given by InputImageRejection. This upper sideband behavior is not intuitive but since this mode is almost exclusively used for receiver/downconversion applications, the upper output sideband is generally not of much interest. The default value for InputImageRejection is 200 dB. OutputSidebandSuppression is ignored for this mode. This is illustrated in Mixer2 image/sideband operation for SideBand=LOWER IMAGE REJECTION..
Mixer2 image/sideband operation for
SideBand=LOWER IMAGE REJECTION.
For SideBand=UPPER IMAGE REJECTION, the upper input image is rejected by the amount given by InputImageRejection relative to the lower input image and both output sidebands are generated without suppression. Since this mode is almost exclusively used for receiver/downconversion applications, the upper output sideband is generally not of much interest. The default value for InputImageRejection is 200 dB. OutputSidebandSuppression is ignored for this mode. This is illustrated in Mixer2 image/sideband operation for SideBand=UPPER IMAGE REJECTION..
Mixer2 image/sideband operation for
SideBand=UPPER IMAGE REJECTION.
Other combinations of input image rejection and output sideband suppression exist which are not supported by Mixer2. None of these modes are of much practical relevance.
The InputImageRejection and OutputSidebandSuppression parameters do not support image or sideband enhancement. The input image rejection and output sideband suppression is abs (InputImageRejection) and abs (OutputSidebandSuppression). The signs of InputImageRejection and OutputSidebandSuppression do not matter.
The operation of Mixer2 for SideBand=BOTH emulates the behavior of a single mixer. The operation of Mixer2 for SideBand=LOWER and SideBand=UPPER emulates the behavior of a single mixer with a filter at the output. The operation of Mixer2 for SideBand=LOWER IMAGE REJECTION and SideBand=UPPER IMAGE REJECTION emulates the behavior of a mixing subsystem constructed from two single mixers, a phase shifter, and possibly other components.
Since circuitlevel mixers pass both input images and generate both output sidebands, a practical design process will typically end up using Mixer2 with SideBand=BOTH when the design is finalized. This, however, does not render the other four modes any less valuable. If a design uses a mixer suppressing the lower/upper output sideband, SideBand=UPPER or SideBand=LOWER will typically be used in the beginning stages of the design process. As the design is finalized and more physical realism is desired, a switch to SideBand=BOTH can be made and output filters can be added. Similarly, if a design uses a mixing subsystem rejecting the lower/upper input image, SideBand=LOWER IMAGE REJECTION or SideBand=UPPER IMAGE REJECTION will typically be used in the beginning stages of the design process. As the design is finalized and more physical realism is desired, a switch to SideBand=BOTH can be made and an actual image rejection mixing subsystem can be built from mixers with SideBand=BOTH. Such design processes take advantage of many features offered by Mixer2.
Linear Behavior
The linear behavior of Mixer2 is described by the conversion gain ConvGain, the reverse conversion gain RevConvGain, and the nine reflection/leakage/isolation parameters (Sparameters) SPij (i,j=1,2,3).
The ConvGain parameter is the conversion gain from RF to IF. It is applied to the lower sideband RFLO and the upper sideband RF+LO. Either of these sidebands can then be suppressed by OutputSidebandSuppression, as described in the "Sideband Suppression and Image Rejection" section.
If ConvGain is real, it will simply scale the output sidebands. To understand the behavior of Mixer2 when ConvGain is complex, consider a Mixer2 with SideBand=BOTH. Given an LO, we consider RF1 below LO and RF2 above LO. In a Harmonic Balance simulation, Mixer2 will downconvert to RF1LO=LORF1 and RF2LO=RF2LO and upconvert to RF1+LO and RF2+LO. If signals with nonzero phases are applied at RF1 and RF2, the phases will transfer directly to the output tones at RF2LO, RF1+LO and RF2+LO. However, the phase at LORF1 will be inverted. The reason is that RF1LO is a negative frequency and that the simulator therefore solves for the result at the corresponding positive frequency (RF1LO)=LORF1 and then conjugates the result. This conjugation is what manifests itself in the phase reversal for the LORF1 output tone. It has nothing to do with Mixer2. It is a Harmonic Balance concept which can also be seen with Mixer and VMult. For Mixer2, the phase of ConvGain is applied in an analogous manner. It is applied before the mixing of RF with LO takes place and before any compression from RF to LO is applied, not after. The phase of ConvGain is therefore subjected to the mixing process which means, per above, that the phase of ConvGain will be added to the output tones at RF2LO, RF1+LO and RF2+LO but subtracted from the output tone at LORF1. Mixer, on the other hand, simply applies the phase of ConvGain to all output tones regardless of any mixing taking place. The approach for Mixer2 is consistent with the underlying mixing process and appears much more appealing than that for Mixer.
The RevConvGain parameter, the reverse conversion gain from IF to RF, is similar to ConvGain, except no compression is associated with RevConvGain. Also, the phase of RevConvGain is applied after the mixing, not before. For most applications, RevConvGain will be zero.
The SPij (i,j=1,2,3) parameters describe the port reflection and porttoport leakage/isolation for the mixer. Mixer2 is a threeport device and in line with established theory for generalized Sparameters we denote the voltages and currents at port n by vn and in and define the input and output waves at each port as:
with Zn being the reference impedance for port n. This is illustrated in Definition of Voltages, Currents and Waves..
Definition of Voltages, Currents and Waves.
The nine Sparameters SPij (i,j=1,2,3) are then defined through
b1 = SP11*a1 + SP12*a2 + SP13*a3
b2 = SP21*a1_poly + SP22*a2 + SP23*a3
b3 = SP31*a1_poly + SP32*a2 + SP33*a3
where, relative to normal generalized Sparameters, two instances of a1 have been replaced by a1_poly which is the a1 wave with polynomial compression taken into account.
SP11, SP21 and SP31 operate on a1 or a1_poly and are therefore defined at the frequencies present at the RF port. SP11*a1, SP21*a1_poly and SP31*a1_poly are therefore contributions at the RF/IF/LO port at the frequencies present at the RF port. Similarly, SP12, SP22 and SP32 operate on a2 and are therefore defined at the frequencies present at the IF port. SP12*a2, SP22*a2 and SP32*a2 are therefore contributions at the RF/IF/LO port at the frequencies present at the IF port. Similarly, SP13, SP23 and SP33 operate on a3 and are therefore defined at the frequencies present at the LO port. SP13*a3, SP23*a3 and SP33*a3 are therefore contributions at the RF/IF/LO port at the frequencies present at the LO port. All three ports have contributions at all frequencies. If SPij were normal generalized Sparameters for a threeport, bi, aj and SPij would all be defined at the same frequency. The extension to the case of multiple frequencies appears natural so Sparameter terminology was chosen for Mixer2.
SP11 is the RF reflection, SP12 is the IF to RF leakage/isolation, SP13 is the LO to RF leakage/isolation, SP21 is the RF to IF leakage/isolation, SP22 is the IF reflection, SP23 is the LO to IF leakage/isolation, SP31 is the RF to LO leakage/isolation, SP32 is the IF to LO leakage/isolation, and SP33 is the LO reflection.
ConvGain, RevConvGain and SPij (i,j=1,2,3) are specified as complex numbers. Use the functions polar(magnitude,angle), dbpolar(dB, angle), or VSWRpolar(VSWR, angle) to convert the ConvGain, RevConvGain and SPij specifications into complex numbers. For example, if a reflection/leakage is X dB, use dbpolar(X,0).
ConvGain, RevConvGain and SPij (i,j=1,2,3) are voltage gains and not power gains. For instance, a mixer with ConvGain=polar(10,0) will scale the voltage by a factor of 10 from input to output and will therefore result in a 20 dB increase in power. ConvGain=dbpolar(10,0), on the other hand, will result in a 10 dB increase in power.
As outlined above, SPij (i,j=1,2,3) are wavebased. The same is true for RevConvGain. ConvGain, on the other hand, is voltagebased, not wavebased. ConvGain applies to the input voltage, not the input wave. However, ConvGain is not blindly applied to the input voltage v1 at the RF port. Doing so would make the RF to IF conversion dependent on SP11 as a modified SP11 results in a different b1 and therefore changes v1. In keeping with the measurement standards used to define mixer conversion gain at the system level, the reflected wave due to SP11 is not included. This means that if we change SP11 from 0 to something finite, we will see no change in IF output power because the reflected wave is not taken into account when mixing from RF to IF. The same holds true for Sparameters. With a fixed SP21, the RF to IF leakage does not depend on SP11. This applies to other Sparameters in an analogous manner.
To illustrate the difference between waveand voltagebased parameters, consider a mixer with an RF source, LO source and IF termination. The RF input wave a1 is finite but the RF output wave b1 is zero. Similarly, the LO input wave a3 is finite but the LO output wave b3 is zero. The mixing results in a finite IF output wave b2 but the IF input wave a2 is zero. Now add LO to RF leakage through a finite LO_Rej1 (Mixer) or SP13 (Mixer2) parameter. This will result in a signal at the LO frequency at the RF port and generate a finite b1 wave. Assuming no reflection at the RF source, this will not change the a1 wave. The a1 wave will not have any content at the LO frequency.
Mixer has a wavebased ConvGain parameter. The RF to IF conversion is based on the a1 wave. Since the a1 wave does not change when a finite LO_Rej1 is added, neither will the RF to IF conversion. Specifically, no RF signal at the LO frequency will mix with LO to produce IF output at DC and twice the LO frequency. Thus, Mixer will produce zero DC output. Mixer2, on the other hand, has a voltagebased ConvGain parameter. The RF to IF conversion is based on the RF input voltage, not the a1 wave. Since the RF input voltage changes when a finite SP13 is added, so will the conversion. Specifically, the RF signal at the LO frequency will mix with LO to produce, absent of sideband suppression, IF output at DC and twice the LO frequency. Thus, Mixer2 will produce a finite DC output. However, note that that Mixer2 will not produce a direct feedthrough IF output at the LO frequency if SP21 is finite. The Sparameters, as stated above, are wavebased, not voltagebased. They operate on a1, not v1.
Nonlinear Behavior
The nonlinear behavior of Mixer2 is described by a number of different list and filebased options for specifying compression. A polynomial compression model is used. For detailed information about the polynomial modeling of compression, please see Modeling Basics and AM to PM Conversion in the documentation for Amplifier2. The compression options for Mixer2 and Amplifier2 are identical, as is the underlying modeling.
Amplifier2 applies compression from its input to its output. In an analogous manner, Mixer2 applies compression from RF to IF. In addition, Mixer2 also applies the same compression from RF to LO. There is no guarantee that the compression from RF to LO is the same as that from RF to IF for an actual mixer, but applying this compression from RF to LO seems more reasonable than to not apply compression from RF to LO.
LO Limiting
Under typical operating conditions, a circuitlevel mixer's LO is saturated. This means that the RF to IF mixing process is insensitive to the actual value of the LO signal. A small fluctuation in LO power will not change the RF to IF mixing. A mixer is not a voltage multiplier. To mimic this behavior for Mixer2, the LO is limited. Mixer limits the LO based on a filtered detector signal. Prior to Release 2004A, Mixer2 limited the LO based on the magnitude of the Hilbert transform of the LO signal. Starting in Release 2004A, Mixer2 allows limiting based on both a filtered detector signal and the magnitude of the Hilbert transform of the LO signal, with the latter being the default. Instead of the brick wall DConly filter used for Mixer, it is now a single pole rolloff with a variable bandwidth specified by DetBW. For extremely large bandwidths (1e12 Hz and above), this switches to Hilbert transform limiting. Hilbert transform limiting (large DetBW values) is best for bandpass LO waveforms with multiple signals but does rather poorly when an LO with significant harmonics is used, as analytic signal calculations such as instantaneous magnitude do not apply to broadband signals. For an LO with significant harmonics, filtered detector signal limiting (small DetBW values) should be used.
A circuitlevel mixer has an LO power level below which its behavior changes dramatically. To mimic this behavior for Mixer2, Mixer2 has a PminLO parameter which sets a limit for the effect of the LO power on the mixer's conversion gain. Mixer2 adds this power level to the limited LO power level (limited via a filtered detector signal for DetBW<=1e12 Hz and via the magnitude of the Hilbert transform for DetBW>1e12 Hz) in order to arrive at the total LO power level. Note that the model does not simply switch to using PminLO when the limited LO power level falls below PminLO. It is a gradual transition and not a threshold. When the limited LO power level is much larger than PminLO, the effect of PminLO on the conversion gain is negligible. As the limited LO power level approaches PminLO, the conversion gain will depend on the value of PminLO in a nonlinear manner. If this is undesirable, lower PminLO.
The LO power is not a fixed quantity but is computed dynamically from the signal at the LO port. This means that Mixer2's conversion gain will vary depending on the frequency content of the LO signal. Frequency content other than the primary LO signal could come from harmonics of the primary LO signal or from spurious LO signals. If the dependence of the LO power on the LO frequency content is undesired, a bandpass filter can be added to the LO path.
Filtered detector signal limiting and Hilbert transform limiting behave differently when an LO sideband enters Mixer2. Consider the case of RF=0.1 GHz, LO=1.0 GHz, LORF=0.1 GHz, and LO+RF=1.9 GHz. Assuming an RF power of 20 dBm, an LO power of 0 dBm and a conversion gain of 0 dB, we get 20 dBm IF power at both sidebands. Now add a 1.01 GHz tone at 30 dBm at LO. Filtered detector signal limiting gives zero power at 0.09 GHz and 1.89 GHz and 50 dBm power at 0.11 GHz and 1.91 GHz. Hilbert transform limiting, on the other hand, generates intermodulation products and gives 56 dBm power at 0.09 GHz, 0.11 GHz, 1.89 GHz, and 1.91 GHz. This supports the recommendation to use Hilbert transform limiting (high DetBW values, the default for Mixer2) unless significant harmonics are present at LO.
Mixer2 does not have an LO common mode leakage parameter which would allow the specification of the amount of LO common mode leakage to be present at the output. For a twotone LO, the LO common mode leakage is the level of the sum and difference intermodulation products of the two LO tones. A workaround to realize this is to add an Amplifier2 component in the LO path, set its SOI/TOI parameters to generate the appropriate sum/difference tones, and set SP23 on Mixer2 to leak these tones from LO to IF.
Noise
Given the minimum noise figure NFmin (real), the optimal reflection coefficient Sopt (complex), the noise resistance Rn (real), the noise reference impedance Rref (real), and the source admittance Ys (complex), the noise figure NF of an amplifier is determined by
Note that this is independent of the amplifier Sparameters.
The noise behavior of Amplifier2 is characterized by the four noise parameters NF, NFmin, Sopt and Rn and the reference impedance Z1 for port 1. Amplifier2 is implemented as a Noisy2Port cascaded with an SDD. The abovementioned five parameters control the parameters for the Noisy2Port, the Noisy2Port generates a noise voltage on its output and this noise voltage is passed through the SDD in the same manner as the signal.
The noise behavior of Mixer2 mimics that of Amplifier2. Similar to Amplifier2, Mixer2 is implemented as a Noisy2Port cascaded with an SDD. The parameters NF, NFmin, Sopt, Rn and Z1 are passed to the Noisy2Port and noise is generated. The signal is not affected by the generation of noise but goes unaltered through the Noisy2Port. All parameters except NF, NFmin, Sopt and Rn are then passed to the SDD and the signal and noise goes through the SDD. The Noisy2Port part of Mixer2 is identical to that of Amplifier2 and the noise at the output of the Noisy2Port for Amplifier2 and Mixer2 will consequently be the same if the noise parameters and reference impedances are the same. The difference between Amplifier2 and Mixer2 is the SDD that follows the Noisy2Port.
NFonly mode is used for NFmin=0. This is a special case where only one noise parameter must be specified. In this case, the Noisy2Port has the parameters NFmin=NF, Sopt=0, and Rn=Z1/4 × (10^{NF/10}1). The reference impedance for the noise calculation (not available on the Noisy2Port user interface) is Z1. The NFmin=0, Sopt and Rn parameters are ignored.
(NFmin, Sopt, Rn) mode is used for NFmin>0. This is a more general case than NFonly mode. In this case, the Noisy2Port has the parameters NFmin=NFmin, Sopt=Sopt, and Rn=Rn. The reference impedance for the noise calculation (not available on the Noisy2Port user interface) is Z1. The NF parameter is ignored.
Given an output noise voltage vn , the single sideband noise figure NFssb and the double sideband noise figure NFdsb are given by
NFssb=10*log(( vn^{2} / R + k × T 0 × ( G 1 + G 2 +...)) / ( k × T 0 × G 1))
NFdsb=10*log(( vn^{2} / R + k × T 0 × ( G 1+ G 2 + ...)) / ( k × T 0 × ( G 1 + G 2 + ...)))
where R is the output resistance, k =1.380658e23 J/K is Boltzmann's constant, T 0=290 K is the standard noise temperature, G 1 is the primary power gain from the input noise frequency to the output noise frequency, and G 2+... is the sum of all higher order mixing gains which mix from some input frequency to the output noise frequency. For an amplifier, G 2+... is zero under smallsignal operation. The same holds true for a mixer with input image rejection. For a mixer without input image rejection, G2 is significant and often equal to G1 under smallsignal operation while G3+... is zero under smallsignal conditions. vn^{2} / R represents the noise added by the component(s) while k × T 0 represents the noise power available from the input termination.
The documentation for Amplifier2 outlines how Amplifier2 calculates noise voltages and noise figures in various cases and compares this to the behavior of Amplifier. Due to the similarities between Amplifier2 and Mixer2, much of this material is also relevant for Mixer2. As mentioned above, the difference between Amplifier2 and Mixer2 is not the generation of noise by the Noisy2Port but how this noise is translated through the SDD that follows. For more information about noise voltages, please see the documentation for Amplifier2.
The NFonly and (NFmin,Sopt,Rn) noise modes are covered above and many details about the computation of noise voltages for Noisy2Port are given in the Amplifier2 documentation. The last remaining aspect to be understood is the influence of SideBand, OutputSidebandSuppression and InputImageRejection on the noise voltages and noise figures.
To this end, consider a Mixer2 with a given LO. We choose RF<LO. The lower output sideband is LSB=LORF and the upper output sideband is USB=RF+LO. The input image for downconversion is LSB_Img=2*LORF (LSB_Img mixes down with LO to give LSB) and the input image for upconversion is USB_Img=2*LO+RF (USB_Img mixes up with LO to give USB). This is illustrated in Frequency plan for noise computation..
Frequency plan for noise computation.
We will use Mixer2 in NFonly mode. The results carry over directly to (NFmin,Sopt,Rn) mode. We operate the mixer at low RF powers. We let NF=5.00 dB. Since NF is the double sideband noise figure, we expect Mixer2 to give NFdsb=NF in all cases. This is exactly what Mixer2 gives, see Noise Results. NFssb, on the other hand, depends on SideBand, OutputSidebandSuppression and InputImageRejection. From the expressions for NFssb and NFdsb, we see that
NFssb=NFdsb+10*log( (G1 + G 2 +...) / G 1)
Letting all gain terms above G2 to zero and substituting NF for NFdsb, we expect
NFssb=NF+10*log(1 + G 2 / G 1)
In the following, we will explain the influence of SideBand, OutputSidebandSuppression and InputImageRejection on NFssb.
Noise Results
LSB  USB  
SideBand  Rejection/dB  Suppression/dB  _NF/ dB_  _v.noise/ pV_  _NFdsb/ dB_  _NFssb/ dB_  _NFdsb/ dB_  _NFssb/ dB_ 
BOTH  N/A  N/A  5.00  930.5  5.00  8.01  5.00  N/A 
LOWER IM REJ  0  N/A  5.00  930.5  5.00  8.01  N/A  N/A 
LOWER IM REJ  10  N/A  5.00  690.0  5.00  5.41  N/A  N/A 
LOWER IM REJ  200  N/A  5.00  657.9  5.00  5.00  N/A  N/A 
UPPER IM REJ  0  N/A  5.00  930.5  5.00  8.01  N/A  N/A 
UPPER IM REJ  10  N/A  5.00  690.0  5.00  5.41  N/A  N/A 
UPPER IM REJ  200  N/A  5.00  657.9  5.00  5.00  N/A  N/A 
LOWER  N/A  0  5.00  930.5  5.00  8.01  N/A  N/A 
LOWER  N/A  10  5.00  930.5  5.00  8.01  N/A  N/A 
LOWER  N/A  200  5.00  930.5  5.00  8.01  N/A  N/A 
UPPER  N/A  0  5.00  930.5  N/A  N/A  5.00  8.01 
UPPER  N/A  10  5.00  690.0  N/A  N/A  5.00  5.41 
UPPER  N/A  200  5.00  657.9  N/A  N/A  5.00  5.00 
For SideBand=BOTH, we get NFssb=8.01 dB=NF+3.01 dB for both the lower and the upper sideband. OutputSidebandSuppression and InputImageRejection are both ignored in this case. Since this case offers no image rejection, we have G2=G1 and the expression for NFssb confirms the result by Mixer2.
For SideBand=LOWER IMAGE REJECTION and SideBand=UPPER IMAGE REJECTION, we get an NFssb value that varies from NF+3.01 dB to NF as InputImageRejection is changed from 0 dB (no rejection, same as SideBand=BOTH) to 200 dB. Since these SideBand modes are intended for downconversion applications only, the noise figure for the upper sideband does not matter. OutputSidebandSuppression is ignored in this case. With no rejection, we have G2=G1 and the expression for NFssb confirms the result by Mixer2. With full rejection, we have G2=0 and the expression for NFssb confirms the result by Mixer2. In between, we have G2=10^(InputImageRejection/10)*G1 and expect to get NFssb=NF+10*log(1+10^(InputImageRejection/10)). For InputImageRejection=10, this evaluates to NFssb=NF+0.41 dB and confirms the result by Mixer2.
For SideBand=LOWER, we get NFssb=NF+3.01 dB as OutputSidebandSuppression is changed from 0 dB (no suppression, same as SideBand=BOTH) to 200 dB. Since this SideBand mode is intended for downconversion applications only, the noise figure for the upper sideband does not matter. InputImageRejection is ignored in this case. With no suppression, we have G2=G1 and the expression for NFssb confirms the result by Mixer2. As we change OutputSidebandSuppression, all we do is change the conversion gain to the upper sideband. This should have no effect on the noise in the lower sideband, our sideband of interest. Indeed, we keep having G2=G1 and the expression for NFssb confirms the result by Mixer2.
For SideBand=UPPER, we get an NFssb value that varies from NF+3.01 dB to NF as OutputSidebandSuppression is changed from 0 dB (no suppression, same as SideBand=BOTH) to 200 dB. Since this SideBand mode is intended for upconversion applications only, the noise figure for the lower sideband does not matter. InputImageRejection is ignored in this case. With no suppression, we have G2=G1 and the expression for NFssb confirms the result by Mixer2. As we change OutputSidebandSuppression, we change the conversion gain to the lower sideband. As for SideBand=LOWER, we might expect this to have no effect on the noise in the upper sideband, our sideband of interest. However, the input image USB_Img is above USB in this case and therefore is subjected to the conversion gain to the lower sideband which is being changed by OutputSidebandSuppression. An upper sideband mixer is, in effect, an image rejection mixer. With full suppression, we have G2=0 and the expression for NFssb confirms the result by Mixer2. In between, we have G2=10^(OutputSidebandSuppression/10)*G1 and expect to get NFssb=NF+10*log(1+10^(OutputSidebandSuppression/10)). For OutputSidebandSuppression=10, this evaluates to NFssb=NF+0.41 dB and confirms the result by Mixer2.
Noise can be contributed and lowered in several ways. To see this, consider a downconverting mixer with a certain RF and LO. IF is RFLO and RFimg is 2*LORF. Output noise at IF can be a result of noise at RF and RFimg mixing with LO to contribute noise at IF but can also be a result of noise at IF leaking directly through the mixer from RF or LO if it has a finite RF/LO to IF leakage. This output noise at IF can be lowered if the mixer is an image rejection mixer that rejects the signal/noise contribution at RFimg or if the mixer has no RF/LO to IF leakage. In the absence of a mixer with such properties, output noise at IF can be lowered external to the mixer by adding filters at the RF and LO ports which eliminate the input noise at RFimg and IF (RF port) and IF (LO port). Eliminating the noise at RF is much harder since adding a filter at RF will also eliminate the desired signal at RF.
Notes/Equations
 Mixer2 is introduced as a replacement for Mixer for anything but frequency conversion AC analysis (see note 3). To change an existing Mixer component to a Mixer2 component, change the name from Mixer to Mixer2 on the schematic. Many parameters for the two models are the same and Mixer2 will adopt the values for Mixer, making parameter reentry unnecessary. The only exception is that the parameters SideBand, GainCompType and ReferToInput will take their default values BOTH, LIST and OUTPUT, respectively, regardless of the values these parameters had for Mixer. The Mixer parameters ImageRej, LO_Rej1, LO_Rej2, RF_Rej, S11, S22 and S33 correspond to the Mixer2 parameters OutputSidebandSuppression, SP13, SP23, SP21, SP11, SP22 and SP33 and will have to be reentered. In addition, Mixer2 has the new parameters InputImageRejection, SP12, SP31, SP32, DetBW, AM2PM, PAM2PM and ClipDataFile whose values cannot be deduced from Mixer and will therefore take their default values. Also, the Mixer display settings will be ignored. Mixer2 will adopt its default settings, displaying SideBand and ConvGain. Major Differences between Mixer and Mixer2. summarizes the major differences between Mixer and Mixer2
Major Differences between Mixer and Mixer2.
Mixer Mixer2 Image rejection not supported Image rejection supported Reverse conversion gain not supported Reverse conversion gain supported Some leakage terms supported All leakage terms supported Real/imaginary polynomial fit Magnitude/phase polynomial fit AM to PM not supported AM to PM supported for all magnitude modes One type of LO limiting More flexible LO limiting Complex ConvGain leads to nonphysical behaviors Complex ConvGain leads to physical behaviors FCAC analysis supported FCAC analysis not supported For large harmonic balance and circuit envelope simulations Mixer2 may be slower than Mixer.  For a tutorial example of how to use Mixer2 in various cases, see examples/Tutorial/Mixer2_Example_prj. For other examples of how to use Mixer2, search the ADS examples for the Mixer2 component.
 Mixer supports frequency conversion AC (FCAC) analysis for smallsignal AC or Sparameter analysis, while Mixer2 does not. This capability allows smallsignal frequency analysis traditionally done at only one frequency to be somewhat extended to deal with more than one frequency. It is not as accurate as harmonic balance analysis. If FCAC analysis is needed, use Mixer. If not, use Mixer2.
 Mixer2 does not support complex reference impedances.
 Mixer2 passes dc.
 Mixer2 is not recommended for baseband envelope applications.
 Mixer2 may be slower than Mixer for large harmonic balance and circuit envelope simulations.
 When using Mixer2 in transient simulations, Mixer2 gives a warning about noncausality of the H(2), H(15) and H(18) weighting functions. Please ignore this warning.
 When setting up simulations with Mixer2, make sure to avoid colliding tone issues where multiple mixing products map to the same frequency. Setups with colliding tones should be changed so colliding tones are eliminated.
 This note describes Mixer2 operation with the SOI and TOI parameters set. SOI and TOI are used for specifying the secondorder and thirdorder intercept points IP2 and IP3.
The general equation for the nth order intercept point IPn is IPn=(n*P1Pn)/(n1) where P1 is the power level of the firstorder tone and Pn is the power level of the nthorder tone. The latter power level, however, is not unique. It can be based on either harmonics or intermods, and the two will generally not be the same. If the formula IPn= (n*P1Pn)/(n1) works for one type of second and thirdorder tone, it will not work for the other. The industry standard for IPn is based on intermods, and the same goes for the SOI and TOI parameters for Mixer2. IPn must therefore be validated using a twotone setup, not a onetone setup. Note that IPn is defined at a low power level. If IPn is computed at a power level where either P1 or Pn deviate from their lowpower values, the results will be in error. Note that it is not enough that the fundamental tone varies linearly. IPn is computed based on the fundamental tone and a higherorder intermod so one must ensure that the higherorder intermod is also linear or IPn will change. Also, note that IPn is defined under outputmatched conditions.
While the basic definition of the power levels used for computing IPn are the same for an amplifier and a mixer, this is not the case for the frequencies at which these power levels are evaluated.
For an amplifier with two input tones at f1 and f2 (assume f1<f2) at the same power level, a twotone harmonic balance simulation will result in firstorder tones at f1 and f2, secondorder intermod products at f2f1 and at f1+f2, and thirdorder intermod products at 2 × f1f2 (will be smaller than f1 by f2 f1) and 2 × f2f1 (will be greater than f2 by f2f1). For computing IP2=SOI, we can use the power levels at f1 and f2f1 or we can use the power levels at f2 and f1+f2. For computing IP3=TOI, we can use the power levels at f1 and 2 × f1f2 or we can use the power levels at f2 and 2 × f2f1.
For a mixer, everything is the same except that all frequencies are being shifted down (downconverting mixer) or up (upconverting mixer) by the LO frequency LO due to mixing. Therefore, all these frequencies will either have LO subtracted or added. Note that this means that from a harmonic balance point of view, IP2=SOI is calculated off of second and thirdorder intermod products and IP3=TOI is calculated off of second and fourthorder intermod products. For computing IP2=SOI, we can use the power levels at f1+/LO (secondorder as far as the harmonic balance simulation is concerned) and at f2f1+/LO (thirdorder as far as the harmonic balance simulation is concerned) or we can use the power levels at f2+/LO (secondorder as far as the harmonic balance simulation is concerned) and at f1+f2+/LO (thirdorder as far as the harmonic balance simulation is concerned). For computing IP3=TOI, we can use the power levels at f1+/LO (secondorder as far as the harmonic balance simulation is concerned) and at 2 × f1f2+/LO (fourthorder as far as the harmonic balance simulation is concerned) or we can use the power levels at f2+/LO (secondorder as far as the harmonic balance simulation is concerned) and at 2 × f2f1+/LO (fourthorder as far as the harmonic balance simulation is concerned).  When the RF port is excited with two tones, the down/upconverted sum/difference frequencies are present at the IF port. What about the sum/difference frequencies themselves? The difference frequency can be very important for a zeroIF downconverting mixer. The answer is that the sum/difference frequency level at the IF port depends on SP21 and SOI. For SP21=0, the sum/difference frequency level is zero. For finite SP21 values, the sum/difference frequency level depends on SOI. The sum/difference frequency level will be zero at zero input power and will rise linearly with the expected 2:1 slope as the input power level is increased. Consequently, the sum/difference frequency level gets more and more significant as the input power gets closer and closer to SOI. At a high enough input power level, the sum/difference frequency level compresses. If SOI is excessively high, the sum/difference frequency level remains low for all realistic input powers. The shape of the curve is determined by SOI and can be scaled by varying SP21. The sum/difference frequency level will experience a sharp increase when the mixer starts hard limiting, regardless of the SP21 and SOI values. For an explanation of why this happens, see the "Modeling Basics" section of the Amplifier2 documentation. All this is also true at the LO port, except that SP31 controls this level.
 An S2D file typically consists of an ACDATA block containing smallsignal information and a GCOMPi block (i=1,...,7) containing compression information. For Mixer2, the ACDATA block is ignored and the Sparameters specified on the Mixer2 component are used. Similarly, any NDATA blocks containing noise data are ignored by Mixer2.
 When an S2D file contains gain compression data at more than one frequency, the GainGompFreq must be set to one of the frequencies in the S2D file to identify the data to be used. It is imperative that GainCompFreq be set to one of the frequencies in the S2D file as no interpolation or extrapolation between gain compression data at different frequencies can be performed.
 When an S2D file has a power range that exceeds that of a simulation, a choice must be made for the power range used for fitting. Assume an S2D file covers 30 dBm to 30 dBm but that a simulation is carried out from 10 dBm to 10 dBm. In this case, a choice must be made as to whether the polynomial fitting of S2D data is done over the power range 30 dBm to 30 dBm or 10 dBm to 10 dBm. In the former case, the fitting may be inaccurate as the polynomial must cover a large power range that could hold a lot of variations. This is undesirable. However, the advantage of this approach is that the results we get when simulating from 10 dBm to 10 dBm are a subset of what we would have gotten in that interval had we simulated from 30 dBm to 30 dBm. In the latter case, the fitting is much more accurate as the fitting is done over a much smaller power range which presumably holds a lot less variation. This is desirable. However, the problem with this approach is that the results between 10 dBm and 10 dBm will be different for a simulation done from 30 dBm and 30 dBm rather than from 10 dBm and 10 dBm since the polynomial coefficients change as we change the power range of the simulation. ADS does the former. It fits a polynomial to the whole S2D file, not just the subset for which the simulation is carried out. To change the fitting in a power range, it is not enough to change the power range of the simulation. To change the fitting, one must modify the S2D file. The S2D file power range, not the simulation power range, dictates the fitting power range. This is relevant in the following where we shall discuss different fittings in different power ranges.
 A typical Pout (output power) vs. Pin (input power) curve consists of a linearly increasing region, a transition region and a saturation region. Another way of thinking of this is that typical PoutPin vs. Pin curve consists of a flat region, a transition region and a linearly decreasing region.
When the saturation region is made larger and larger, the fitting approach adopted by Mixer2 (polynomial fitting, odd order terms, order dependent upon the number of data points in the S2D file, max order 9) will tend to produce fitted curves which ring/oscillate more and more at higher powers. Mild ringing is often tolerable and might not even be noticed but if the transition region becomes too large it can make the results useless. To alleviate the problem, reduce the size of the saturation region to the minimum needed and leave no extra points in the S2D file. If the results are still not satisfactory, make sure ClipDataFile is set to yes and reduce the saturation region even more, relying on Mixer2 extrapolation. If the results are still not satisfactory, try breaking the S2D file into two files and simulate the problem in two steps.
If fitted results do not accurately match the data in the S2D file and it is uncertain if this ringing problem is the cause, the problem is very easy to exaggerate. Simply extend the GCOMP7 block of the S2D file with a large flat region (more input powers, saturated output power, saturated output phase) and resimulate. If the ringing problem is the cause, the results should get worse.  The S2D file capability is a legacy from OmniSys. OmniSys used GComp1GComp7 data items for specifying gain compression. Gain Compression Data for OmniSys and ADS summarizes the gain compression data for OmniSys and ADS. Refer to OmniSys Parameter Information for OmniSys parameter information. GComp1GComp6 can be specified by using the corresponding ADS gain compression parameters and setting GainCompType=LIST or they can be contained in an S2D format setting GainCompType=FILE.
OmniSys
ADS
GComp1: IP3
TOI
GComp2: 1dBc
GainComp=1dB
GComp3: IP3, 1dBc
TOI
GComp4: IP3, Ps, GCS
TOI
GComp5: 1dBc, Ps, GCS
GainComp=1dB
GComp6: IP3, 1dBc, Ps, GCS
TOI
GComp7
GainCompType=FILE
OmniSys Parameter Information