Using Generalized Multi-Dimensional Data Files to Model Amplifier Temperature and DC Operating Point Dependence within the Amplifier2 Behavioral Model in ADS®

Jun 25, 2026 | Amplifiers, Engineering Resources

This application note serves as a companion to our previous application note “Using S-Parameter Files to Model Small-and Large-Signal Amplifier Performance in ADS®”.

Mini-Circuits characterizes its amplifiers across multiple operating conditions, including temperature and DC bias, and publishes the resulting data in separate sets.

  • Amplifiers are typically characterized at a minimum of three operating temperatures and several DC operating points
  • Each operating condition is provided as its own data set
  • Discrete S-parameter files are available on our website model pages under S-PARAMETERS, while cascade parameters such as OIP3, P1dB, Psat, and GCS are available under View Data.
  • Manually downloading these files, building individual MDF files for each condition, and creating the corresponding ADS schematic can be time-consuming and cumbersome.

To simplify this workflow, Mini-Circuits now provides a single MDF file for each amplifier, which consolidates all the data available for that model on our website. 

The remainder of this application note outlines the modified MDF file format and provides step-by-step guidance for building the required ADS schematic.

MDF File Structure

Within a single MDF file, each operating condition is separated by variable (VAR) statements. Each variable must be arranged in either ascending or descending monotonic order. The variables used are VAR Vd(1), which defines the operating device voltage, and VAR Temp(1), which defines the operating temperature. In most cases, !VAR Id(1) is also included, but it is commented out to avoid the monotonicity requirement. 

The variable data must appear immediately before each BEGIN Block statement. The example below illustrates a MDF file structure that includes three operating temperatures (85, 25, and -45 C) and two DC operating voltages (5.0 and 6.0).

!
!Case 1: Vd = 5.0 V
VAR Vd(1) = 5.0
!VAR Id(1) = 54.81
VAR Temp(1) = 85
BEGIN Block1_ACDATA
% freqW(real) S11dB(real) S11Ang(real) S21dB(real) S21Ang(real) S12dB(real) S12Ang(real) S22dB(real) S22Ang(real)
10000000 -0.1460908 173.584 -50.38005 -2.944099 -80.23989 -133.4602 -0.0036072 -35.22616
20000000 -0.2380212 171.9628 -28.65564 25.11547 -82.61297 -114.3838 -0.1619668 -66.75084
30000000 -0.2691055 169.1232 -19.28501 14.1805 -70.39982 -129.4078 -0.378299 -92.1173
.
.
19800000000 -1.792018 -3.852982 -24.32718 130.2681 -32.91138 -43.58435 -1.904642 -30.99292
19900000000 -1.968028 -7.09544 -21.63564 125.3852 -33.03466 -54.324 -1.990327 -35.31568
20000000000 -2.167639 -9.791246 -20.02058 118.9317 -31.7876 -73.4836 -2.036779 -39.22533
END
!
VAR Vd(1)=5.0
!VAR Id(1)=54.81
VAR Temp(1)=85
BEGIN Block2_NF
% freqW(real) NF(real)
200000000 4.96
300000000 2.83
400000000 2.14
.
.
9600000000 4.18
9800000000 4.46
10000000000 4.77
END
!
VAR Vd(1) = 5.0
!VAR Id(1) = 54.81
VAR Temp(1) = 85
BEGIN Block3_GCOMP6
% freqW(real) IP3dBm(real) P1dB(real) PsatdBm(real) GCSdB(real)
200000000 28.15 15.47 17.47 3.0
300000000 27.23 15.44 17.44 3.0
400000000 28.37 16.19 18.19 3.0
.
.
9600000000 21.32 14.22 16.22 3.0
9800000000 20.73 13.29 15.29 3.0
10000000000 19.14 11.28 13.28 3.0
END
!

The remaining temperature data appears before each Block as follows.

VAR Vd(1) = 5.0
!VAR Id(1) = 64.76
VAR Temp(1) = 25

Followed by:

VAR Vd(1) = 5.0
!VAR Id(1) = 67.90
VAR Temp(1) = -45

Note that in the example above, Temp(1) has monotonically decreased from +85 to -45 C while Vd(1) is fixed at 5.0 V.

The next set of variable data is for the Vd(1)=6.0 V operating point, which must also appear before each block as follows.

!
!Case 2: Vd = 6.0 V
VAR Vd(1) = 6.0
!VAR Id(1) = 71.82
VAR Temp(1) = 85

Followed by:

VAR Vd(1) = 6.0
!VAR Id(1) = 81.87
VAR Temp(1) = 25

And lastly:

VAR Vd(1) = 6.0
!VAR Id(1) = 90.29
VAR Temp(1) = -45

Note that for this case, Temp(1) must also monotonically decrease from +85 to -45 C and Vd(1) must monotonically increase from 5.0 to 6.0.

Single Tone ADS Schematic using Amplifier2 Behavioral Model

The ADS schematic for the Amplifier2 behavioral model of the Mini-Circuits PMA3-83LN+ MMIC amplifier is shown in Figure 1.

Figure 1: Single Tone Amplifier2 Behavioral Model for PMA3-83LN+

All data required by the Amplifier2 model is extracted from a single MDF file using three individual Data Access Components (DAC). 

  • DAC1 contains the measured s-parameter data
  • DAC2 contains the measured compression data
  • DAC3 contains the measured noise figure data

Parameters and variables for each DAC are as shown in Figures 2 through 4 below.

Figure 2: DAC1 S-Parameter Data

Figure 3: DAC2 Compression and 3rd Order Intercept Point Data

Figure 4: DAC3 Noise Figure Data

Each DAC references a specific Block within the MDF file.

  • Block1_ACDATA
  • Block2_NF
  • Block3_GCOMP6

Each Block must also reference a specific set of independent variables:

  • iVar1=”freqW”.  Simulation frequency
  • iVar2=”Temp”. Operating Temperature
  • iVar3=”Vd”.  Operating Voltage
  • iVar4=”Id” Device Operating Current (listed but not used)

The variables are then initialized and activated using parameters defined in the Var Eqn Data Item blocks shown in Figure 5.

Figure 5: Variable Initialization and Activation

VAR20, shown in Figure 5, instructs the simulator to select the Temp = 85 and Vd = 5.0 data from the MDF file.

The simulation setup uses the following sweep definitions:

  • VAR1 initially sets the carrier frequency, freqW, to 6 GHz and then sweeps it from 1 GHz to 10 GHz using PARAMETER SWEEP Sweep1.
  • Input power, Pin, is initially set to -5 dBm and then swept from -15 dBm to +5 dBm using SWEEP PLAN SwpPlan1.

The required ADS simulation elements are shown in Figure 6.

Figure 6: Simulation Controller and Parameter Sweep Definitions

Single Tone ADS Simulation Results

Single Tone simulation results for Gain and Output Power versus Input Power at +85 C with Vd = +5 V, are presented in Figure 7.

Figure 7: Output Power (y-axis) and Gain (right y-axis) vs Input Power

From the simulation results in Figure 7, the device performance at a 1 GHz carrier frequency, Vd = +5 V, and Temp = +85 C can be estimated as follows:

  • Small-signal gain (m6): ~21.8 dB (21.8 dBm)
  • Output power at 1 dB compression (m5): ~+18.9 dBm (+19.1 dBm)
  • Output power at 3 dB compression (m7): ~+20.5 dBm (+21.1 dBm)

The values in parentheses are taken directly from the MDF file and show close agreement with the simulation results. Small signal S-Parameters, obtained using the S-PARAMETER simulation controller, are presented in Figure 8 and agree exactly with the measured S-Parameter data contained within the MDF file.

Figure 8: Small-Signal S-Parameter Simulation Results

Two Tone ADS Schematic using Amplifier2 Behavioral Model

The ADS schematic for the Amplifier2 behavioral model of the Mini-Circuits PMA3-83LN+ MMIC amplifier is shown in Figure 9.  DAC elements are identical to those of Figures 2 through 4.

Figure 9: Two Tone Amplifier2 Behavioral Model Schematic for PMA3-83LN+

The schematic is identical to the one shown in Figure 1, except that the input source now includes two tones, P[1] and P[2]. The power level of each tone is controlled by the sweep variable SweepVar = “Pin”, which is swept according to SwpPlan1 from -13 dBm to +3 dBm, each.  The required ADS simulation elements are shown in Figure 10.

Figure 10: Simulation Controller and Parameter Sweep Definitions

Two Tone ADS Simulation Results

Figure 11 presents two-tone simulation results for third-order output distortion in a spectrum analyzer format. The simulation was run at 6 GHz ± 1 MHz with Vd = +5 V and Temp = +85 C, and the results are plotted as a function of input power.

Figure 11: Simulated 2-Tone 3rd Order Distortion as a function of Input Power

The output third-order intercept point (IP3out) can be calculated from Figure 11 data using Equation 1.

Eqn1: IP3out(dBm) = Pout/Tone(dBm) + delta(dB)/2

with delta = Pout/Tone(dBm) – PIP3/Tone(dBm)

In this expression, PIP3/Tone is the power level of the third-order intermodulation product.

As an example, Equation 1 can be applied to the trace markers in Figure 11 at two output power levels:

  • Example 1: Pout/Tone (m11) = 15.962 dBm, PIP3/Tone (m9) = 3.694 dBm, and delta = m11 − m9 = 12.268 dB. Therefore, IP3out = 15.962 dBm + 12.268 dB/2 = 22.097 dBm.
  • Example 2: Pout/Tone (m12) = 10.357 dBm, PIP3/Tone (m10) = -26.966 dBm, and delta = m12 − m10 = 37.323 dB. Therefore, IP3out = 10.357 dBm + 37.323 dB/2 = 29.018 dBm.

These values are identical to those calculated in the ADS Data Display and shown in Figure 11.

Because applying Equation 1 manually can be cumbersome, ADS also provides a built-in IP3out simulation component and ipo1 function, shown in Figure 12.

Figure 12: ADS Built-In Simulation Component for 3rd Order Output Intercept Point

Figure 13 shows the results produced by the built-in IP3out simulation component, which agrees with the manual calculations.

Figure 13: IP3out Built-In Simulation Component Results

Bottom Line

Mini-Circuits characterizes its amplifiers across a range of operating conditions, including temperature and DC bias, and publishes the resulting data in separate files. Managing these files manually—building individual MDF files for each condition and creating the corresponding ADS schematics—can be time-consuming and cumbersome.

To simplify this workflow, Mini-Circuits will now provides a single MDF file for each amplifier that consolidates all of the data available on the website.

This note highlighted the key elements of the new MDF workflow:

  • It explained the modified MDF file format.
  • It provided step-by-step guidance for building the required ADS schematic.
  • It included single-tone and two-tone simulation examples that demonstrate the versatility and effectiveness of the new MDF format.

For those who may be unfamiliar with the full utility of these resources, this note may provide valuable tools to your toolkit.