Mixer Workshop Instruction
Mixer Workshop Instruction
________________________________________________________________________
SpectreRF Workshop
Mixer Design Using SpectreRF
MMSIM6.0USR2
November 2005
Contents
The procedures described in this workshop are deliberately broad and generic. Your
specific design might require procedures that are slightly different from those described
here.
Purpose
This workshop describes how to use SpectreRF in the Analog Design Environment to
measure parameters which are important in design verification of mixers. New features of
MMSIM6.0USR2 are included.
Audience
Users of SpectreRF in the Analog Design Environment.
Overview
This application note describes a basic set of the most useful measurements for mixers.
Introduction to Mixers
Mixers are key components in both receivers and transmitters. Mixers translate signals
from one frequency band to another. The output of the mixer consists of multiple images
of the mixers input signal where each image is shifted up or down by multiples of the
local oscillator (LO) frequency. The most important mixer output signals are usually the
signals translated up and down by one LO frequency.
In an ideal situation, the mixer output would be an exact replica of the input signal. In
reality mixer output is distorted due to non-linearity in the mixer. In addition, the mixer
components and a non-ideal LO signal introduce more noise to the output. Bad design
might also cause leakage effects, complicating the design of the complete system.
Noise performance and rejection of out-of-band interferers are both critical to the receiver
system because they both limit the receiver system's sensitivity. Linearity is important to
transmitter performance, where you want an error-free output signal.
The example circuit, a single balanced differential down-converting mixer, runs with a
local oscillator at f(LO) = 5 GHz. The range of interest is baseband output noise from 1
kHz to 10 MHz. The RF signal frequency used for the simulation is around 5001 MHz.
■ To supply a LO input to the mixer, use a port (PORT1) with a matching resistor
and transfer the single-ended signal into the differential with an ideal passband
balun.
■ To represent the RF input to the mixer, use a port (PORT0) which is matched with
the mixer input.
■ To use the differential output for measurements, match the output port (PORT3)
to the output impedance of the mixer.
■ Set the LO bias voltage to 1.5 V and the mixer supply of 2.5 V.
■ Set the LO port to sinusoidal source for all the measurements described in this
workshop.
■ For both LO and RF ports, the amplitude and frequency of the signal are
parameterized as plo, prf and frf. You usually specify the amplitude in dBm. In
addition, for the RF port specify the small signal parameter PAC Magnitude. Use
pacmag or pacdbm, depending on the units you prefer.
We’ll begin our examination of the flow by bringing up the Cadence Design Framework
II environment and look at a full view of our reference design:
Lab1: Voltage Conversion Gain Versus LO Signal Power (Swept PSS with PAC)
When the mixer's input impedance and load impedance are both equal to the source
impedance, the power and voltage conversion gains, in decibels, are the same. Note that
when you load a mixer with a high impedance filter, this condition is not satisfied.
■ Using a small-signal analysis, like PSS with PAC or PXF. The PSS with PAC or
PXF analyses supply the small-signal gain information. You can use either PAC
or PXF analysis to compute the voltage gain.
This example measures the variation of conversion gain with the power of the LO signal.
Action 1-1: In the Library Manager window, open the schematic view of the
mixer_testbench design in the library RFWorkshop
Action 1-2: Select the PORT0 source by placing the mouse cursor over it and clicking
the left mouse button. Then in the Virtuoso Schematic Editor select
Edit—Properties—Objects… The Edit Object Properties window for the
port cell should come up…
Action1-6: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action1-9: In the Choosing Analyses window, select the pss button in the Analysis
field of the window. Set the fundamental frequency parameter, flo = 5
GHz. Set errpreset = moderate. Click Sweep and enter plo as the Variable
Name parameter to sweep LO power. Click Sweep Range and set Start =
-10 dBm and Stop = 20 dBm. This sweeps LO power from a small value
to a value above the expected gain saturation. The form should look this:
Action1-10: In the Choosing Analyses window, select the pac button in the Analysis
field of the window. Set fixed input frequency point to the RF signal
frequency, 5001 MHz. Select sidebands either by specifying Maximum
sidebands = 2 or using Select from range. Set the maximum sideband to 2
as you are only interested in the first harmonics of LO. The form should
look like:
Action1-11: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button. Now your Virtuoso Analog Design Environment will
look like:
Action1-14: In the Direct Plot Form, select the pac button, and configure the form
as follows:
Action1-15: Select port3 on schematic. You should get the following waveform:
The PAC analysis computes gain directly only when you set the pacmag parameter to 1
V. Otherwise, take a ratio of the output and input. The maximum conversion gain value is
reached somewhere above 15 dBm. Use this value for the plo parameter in the following
measurements.
Action1-16: After viewing the waveforms, click Cancel in the Direct Plot form. Close
the waveform window.
This example measures how conversion gain varies with the frequency of the stimuli.
Action2-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action2-3: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action2-6: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as follows:
Action2-8: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
Action2-11: In the Direct Plot Form, select the pac button, and configure the form
as follows:
Since the sweep type in the analysis was linear by default, uniform frequency points
display along the X-axis in the above plot. For a large frequency range, set the sweep
type to logarithmic as it is by default.
The same PAC analysis generates results you can use to measure RF to LO isolation and
will be used in measurements that follow.
Action2-13: Close the waveform window and click Cancel in the Direct Plot form.
This example measures the small-signal voltage conversion gain using the PXF analysis.
Action3-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action3-3: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action3-6: In the Choosing Analyses window, select the pss button in the Analysis
field of the window. Set the fundamental frequency parameter, flo = 5
GHz or use Auto Calculate button. Set errpreset = moderate. The form
will the same as in Action2-6.
Action3-7: In the Choosing Analyses window, select the pxf button in the Analysis
field of the window. Sweep the output frequency from 1 kHz to 10 MHz.
Select sidebands either by specifying Maximum sidebands = 2 or using
Select from range. The form looks like:
Note that Maximum sidebands = 2 is good for this example, other circuits might require
a different value. Set the Output by specifying output as Voltage with Positive Output
Node being the Ifp net and Negative Output Node being the Ifn net.
Action3-11: In the Direct Plot Form, select the pss button, and configure the form
as follows:
Action3-13: After viewing the waveforms, click Cancel in the Direct Plot form
The sweep type in the analysis is logarithmic by default. The large frequency range in
above figure requires a logarithmic X-axis.
Yet another way to measure small-signal gain is to use the PSS and PSP analyses to get
the gain and noise parameters with one simulation. You may want to refer to SpectreRF
user guide Appendix L (using psp and pnoise analysis) for more details.
You can also set up an appropriate QPSS analysis to measure large-signal gain. Set LO as
a large tone on the Plo port. Use a sinusoidal voltage source for the Prf port. This analysis
models the signal at a particular frequency going through the mixer. In the Direct Plot
form for QPSS, the Voltage and Power Gain provide all the needed information.
You can measure the Power Conversion Gain and Power Dissipation for an unmatched
source and load using a QPSS analysis. If the effect of the RF tone is small, you might
use a PSS analysis instead, as mentioned in previous sections.
The QPSS and PSS analyses provide only spectrum data, not a scalar value, of the total
power. To get a scalar value for total power, work through the summation over the
harmonics and sidebands. In general, most of the power is in the main output harmonics.
Note: Some users might not be fully aware of the notion of the port and they might get
wrong results by relying on a 50 Ohm resistance for the port. To get the correct results for
unmatched ports, you need to save the currents on the power supply terminals.
Action4-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action4-2: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action4-3: Select the PORT0 source by placing the mouse cursor over it and clicking
the left mouse button. Then in the Virtuoso Schematic Editor select
Edit—Properties—Objects… The Edit Object Properties window for the
port cell should come up. Make sure the properties are set as follows:
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type sine
Frequency name 1 RF1
Frequency 1 frf1
Amplitude 1 (dBm) prf
Action4-7: In the Choosing Analyses window, select the qpss button in the Analysis
Action4-10: In the schematic, select the positive terminals of port0, port3 and V0.
The terminals are circled in the schematic window after you select them,
indicate that you will save the currents at these nodes.
Action4-11: press Esc with your cursor in the schematic window to end the selections.
Action4-14: In the Direct Plot Form, select the qpss button, and configure the form
as follows:
Action4-16: Close the waveform window, click Cancel in the Direct Plot form
The receiver amplifies the small input signals to the point where they can be processed by
the baseband section. You develop a gain budget where every stage in the receiver is
assigned the gain it is expected to provide. Therefore, the signal gain or loss provided by
the mixer must be known.
There are various ways of characterizing gain and all are derived from the mixer's S-
parameters. As such, it must be easy to calculate the various S-parameters of the circuit
and apply the various gain metrics.
Action5-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action5-2: Select the PORT0 source by placing the mouse cursor over it and clicking
the left mouse button. Then in the Virtuoso Schematic Editor select
Edit—Properties—Objects… The Edit Object Properties window for the
port cell should come up. Change the source type to dc. Blank out all the
Second Sinusoid fields before setting the Source type to dc.
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type dc
PAC Magnitude ( blank )
Action5-5: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action5-8: In the Choosing Analyses window, select the pss button in the Analysis
field of the window. Set the form as you did in Action2-6.
Action5-9: In the Choosing Analyses window, select the psp button in the Analysis
field of the window and set the form as follows:
Action5-10: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
Action5-13: In the Direct Plot Form, select the psp button, and configure the form
as follows:
Action5-16: In the Direct Plot form, choose sp function, set plot type to Rectangular
and Modifier to dB20. Plot S21.
Action5-17: In the Direct Plot form, choose GAIN and click OK button to plot the
input to output gain;
Action5-18: Close waveform window, click Cancel in the Direct Plot form. Close the
waveform window.
PSP GAIN differs from the PSP S21 gain and the Pnoise gain because PSP GAIN is
independent of input match (determined by the impedance of the RF port). PSP S21and
Pnoise gains will vary depending on the input match. PSP Gain is the voltage gain from
the internal port voltage source to the output. You can refer to SpectreRF user guide
Appendix L (using psp and pnoise analysis) for more details.
Because the noise from the mixer is moderated by the LNA’s gain, it places a limit on
how small a signal can be resolved. The sensitivity of the receiver is then adversely
affected. Noise is measured using the noise figure (NF), which is a measure of how much
noise the mixer adds to the signal relative to the noise that is already present in the signal.
An NF of 0 dB is ideal, meaning that the mixer adds no noise. An NF of 3 dB implies that
the mixer adds an amount of noise equal to that already present in the signal. For a mixer
alone, an NF of 15 dB is typical.
Running the PSS and PNoise analyses produces all the needed information which
includes total output noise and noise figure.
Action6-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action6-2: Select the PORT0 source, set the source type to dc. Select the PORT3,
set the source type to dc and set PORT1 to sine.
Action6-4: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action6-5: You can choose Session—Load State, load state “Lab6_PNOISE” and
skip to Action6-10 or …
Action6-7: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as you did in Action2-6.
Action6-8: In the Choosing Analyses window, select the pnoise button in the
Analysis field of the window and set up the form as follows:
Action6-9: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
Action6-12: In the Direct Plot Form, select the pnoise button, and configure the form
as follows:
Action6-14: The waveform window will show up. In waveform window, click on New
Subwindow button.
Action6-15: In the Direct Plot form, set the Function to Output Noise and configure
the form as follows:
Action6-16: Click the Plot button. The waveform window should look like this:
Action6-17: After viewing the waveforms, click Cancel in the Direct Plot form
It is Valuable to know the main contributors of noise in a system, once the noise
performance of the circuit is calculated. This information is readily available from a
Pnoise simulation.
The isolation required between a mixer's ports depends on the circuit and the architecture
of the product. Isolation is critical for the mixer to function properly.
You can combine PAC and PXF analyses to produce transfer functions from different
ports to each other. One suggested configuration might be to set up a PAC analysis with
nonzero pacmag parameter at the signal input (the RF port) A PXF analysis with the IF
port as the output probe This example uses pacmag = 1 for simplicity.
Action7-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action7-2: Select the PORT0 source by placing the mouse cursor over it and clicking
the left mouse button. Then in the Virtuoso Schematic Editor select
Edit—Properties—Objects… The Edit Object Properties window for the
port cell should come up. Change the port properties as follows:
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type dc
PAC Magnitude pacmag
Action7-4: Set the source type of PORT1 to sine, and PORT3 to dc.
Action7-6: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action7-9: In the Choosing Analyses window, select the pss button in the Analysis
field of the window. Set up the form as follows:
PSS simulation is set up here to check the LO feed through. PXF is a small signal
analysis. LO feedthrough is a large-signal effect and shouldn't be measured using a
small-signal analysis.
Action7-11: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
After the simulation runs, use the Direct Plot feature to plot the results.
To avoid desensitizing the stage following the mixer with high-level LO signal
feedthrough to the output, measure LO-to-IF isolation. Use the results of the PSS analysis
with the LO port as input and IF port as output to measure the level of isolation.
Action7-14: In the Direct Plot Form, select the pss button, and configure the form
as follows:
Action7-15: In the schematic, select the net IFp as the output and net VLO as input.
LO-to-RF feedthrough affects the functionality of LNAs and antennas and causes the dc
offset because of self-mixing of Mixer. Use the results of the PSS analysis with the LO
port as input and RF port as output to measure the LO-to-RF feedthrough.
Action7-17: In the Direct Plot Form, change the Plotting Mode to Replace.
Action7-18: In the schematic, select the net RF as the output and net VLO as input.
RF-to-LO feedthrough affects the local oscillator by letting strong interferers at the input
pass to the LO. Measure RF-to-LO feedthrough using the PAC analysis results.
Action7-20: In the Direct Plot Form, select the pac button, and configure the form
as follows:
Action7-22: In the Direct Plot Form, select the pac button, and configure the form
as follows:
Action7-23: Select 0 in Output Harmonics and select the IF output PORT3. The
following waveform shows the RF to IF feedthrough:
Action7-24: Click Cancel in the Direct Plot form. Close the waveform window.
Large interfering signals are called blockers. Blocking signals reduce the mixer's gain and
deteriorate the mixer's noise performance. As such, you need to measure the gain and
noise of a mixer in the presence of a blocking signal. All major communication standards
include blocking requirements for both mobile and base stations. The requirements use
several in-band and multiple out-of-band blocking signals.
Because a mixer has both signal and LO inputs, you should use the multi-tone large
signal QPSS analysis for these measurements. Follow the QPSS analysis with QPAC and
QPNoise analyses to measure gain and NF variations versus the level of the interfering
signal. In the QPSS analysis, model the blocker as a moderate tone.
Action8-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action8-2: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type sine
Frequency name 1 RF1
Frequency 1 5.003G
Amplitude 1 (dBm) prf
PAC magnitude (dBm) -30
Action8-4: Make sure the source type of PORT1 is set to sine and PORT3 dc type.
Action8-10: In the Choosing Analyses window, select the qpac button in the Analysis
field of the window. input frequency = 5.001 GHz and Sweep type =
absolute. The form should look like this:
Action8-12: In the Choosing Analyses window, select the qpnoise button in the
Analysis field of the window. Use a 1 MHz frequency point and
Maximum clock order = 10. Set Output probe as PORT3 and Input source
as PORT0. Use the Reference sideband as (1 0) to represent a
downconverted RF signal relative to the IF output signal, 1 MHz + 1 * f
(LO) = f (RF). The form should look like this:
When the simulation ends, we can check how the blocking signal affect the performance
of the Mixer.
Action8-16: In the Direct Plot Form, select the qpac button, and configure the form
as follows:
Action8-17: Select PORT3. Here shows the Blocker Effect on Voltage Gain
Action8-19: In the Direct Plot Form, select the qpnoise button, and configure the form
as follows:
Action8-21: Close the waveform window. Click Cancel in the Direct Plot form.
Mixer distortion limits the sensitivity of a receiver if there is a large interfering signal
present that is within the bandwidth of the RF input filter (a characteristic known as
selectivity). There are two aspects of distortion that are of concern:
■ Compression
■ Intermodulation Distortion
The 1 dB compression point (CP1) is the point where the output power of the
fundamental crosses the line that represents the output power extrapolated from small-
signal conditions minus 1 dB. The 3rd order intercept point (IP3) is the point where the
third-order term as extrapolated from small-signal conditions crosses the extrapolated
power of the fundamental.
In previous version of SpectreRF, user can use either qpss-based or qpac-based method to
calculation IP3 in a system contains mixer and LO. In qpss-based method, three-tone
qpss analysis with LO, RF1 and RF2 frequencies ωLO, ωRF1 and ωRF2 is run at a given RF
power level. IM3 of harmonic 2ωRF1-ωRF2–ωLO is obtain from the solution. Assuming RF
power is low enough and IM3 is dominated by leading order VRF3 terms, log(VIM3) is
expected to be linear function of log(VRF) with a slope of 3. IP3 is then extrapolated from
VIM3. Here VIM3 and VRF are amplitudes of IM3 and RF signals, respectively. This method
requires very high accuracy to accommodate large dynamic range between RF and LO
signals because they are mixed in the same solution vector. For large circuit, it also relies
on speed and convergence of multi- tone qpss.
Compared to qpss-based approach, the qpac approach reduces computation from three-
tone qpss to two-tone qpss plus a qpac by applying first order perturbation to RF2 signal.
The amount of computation can be further reduced if we treat both RF signals as
perturbation to the steady-state operating point at LO frequency with zero RF input. In
this way, leading order intermodulation between RF1 and RF2 in IM3 can be computed
directly from third order perturbation.
L ⋅ v + FNL (v ) = ε ⋅ s
Here the first term is the linear part, the second one is the nonlinear part, and s is RF
input source. Parameter ε is introduced to keep track of order of perturbation expansion.
Under weakly nonlinear condition, nonlinear part is small compared to the linear part, so
the above equation can be solved by using Born approximation iteratively:
(
u ( n ) = v (1) − L−1 ⋅ FNL u ( n −1) )
where u(n) is the approximation of v and it accurate to the order or O(εn).
Since the evaluation of FNL takes full nonlinear device evaluation of F and its first
derivative, no higher order derivative is needed. This allows us to carry out higher order
perturbations without modifications in current device models. Also, the dynamic range of
perturbation calculations covers only RF signals. It gives the perturbative method
advantages in terms of accuracy.
Action9-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type sine
Frequency name 1 RF1
Frequency 1 frf1
Amplitude 1 (dBm) prf
PAC magnitude (dBm) prf
Action9-5: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action9-8: In the Choosing Analyses window, select the qpss button in the Analysis
field of the window and set the form as follows:
Action9-10: In the Choosing Analyses window, select the qpac button in the Analysis
field of the window. Set the frequency of the small signal very close to
f(RF), for example 5.0011 GHz. In the Select from range option of the
Sidebands section, highlight the harmonics of interest. Limit the
harmonics to second order in the large tone (Set Clock order = 2), from 0
Hz to 6 GHz. The example does not use the 3rd harmonic of the moderate
tone, so remove them from the list.
Action9-11: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
When the simulation completes, use the Direct Plot feature to view the results.
Action9-14: To plot the 1 dB compression point, click qpss analysis in Direct Plot
form. Select Compression Point. Select a point in the linear region for an
extrapolation or leave it blank to use default value. The output harmonic is
(-1 1) or 1 MHz.
Action9-15: Select output Port3 on schematic. You then see the value of P1dB value
as shown in follows:
Action9-17: To plot IP3, select qpac analysis in the Direct Plot Form, select IPN
Curves button, select Variable sweep and choose -40 dB for the prf
extrapolation. If the first extrapolation point you select is not in the linear
range of the IM1 and IM3 curves, you might want to reset the
extrapolation point later. To plot the third order input referred intercept
point, set the first order harmonic to (-1 0) or 1.1 MHz, and the third order
harmonic to (1 -2), or 0.9 MHz. Since the mixer is down-converting to the
baseband, the first harmonic is calculated as:
The third harmonic is at 0.9 MHz or -0.9 MHz depending on the freqaxis
you selected in the QPAC Options form.. The form should look like this:
Action9-18: Select output Port3 on the schematic. The third order input referred
intercept point is calculated and curves of harmonics versus prf are
presented as shown in below:
Action9-19: After viewing the waveforms, click Cancel in the Direct Plot form
Note: For more accurate results, you may want to set errpreset = conservative when
setting up QPSS analysis. Initially, when you do not know the exact location of
the linear region for IM3 and IM1, you may use errpreset = moderate to get a
better understanding of your design. When the linear region is known, defining
a single point simulation with errpreset = conservative is typically more
accurate and less time-consuming.
To calculate IP3, the other way is to apply the LO and two moderate RF input tones in a
single QPSS analyses.
Action10-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type sine
Frequency name 1 RF1
Frequency 1 frf1
Amplitude 1 (dBm) prf
Frequency name 2 RF2
Frequency 2 frf1+0.1M
Amplitude 2 (dBm) prf
Action10-5: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action10-8: In the Choosing Analyses window, select the qpss button in the Analysis
field of the window and set the form as follows:
Action10-9: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
After the simulations finish, plot the IP3 and compare it with the results from Lab9
(QPSS plus QPAC simulations)
Action10-12: In the Direct Plot Form, select the qpss button, and configure the form
as follows:
Action10-13: Select output Port3 on schematic. The IP3 calculation results should look
like this:
Action10-14: After viewing the waveforms, click Cancel in the Direct Plot form
Action10-17: In the Choosing Analyses window, select the qpss button in the Analysis
field of the window.
Action10-19: In field of Function Tones, choose FLO. Change the Maxharms to 15,
because Flexible balance algorithmic need more harmonics to calculate.
Click Update.
1. Flexible Balance Flag: Harmonic balance engine shares the same PSS/QPSS
statement with time-domain engine and is invoked by setting flag flexbalance=yes in the
analysis statement. A toggle button is available for user to switch between time domain
shooting and HB in ADE PSS and QPSS set up form.
A proper choice of the maximum harmonic depends on signal waveform and circuit
nonlinearity. The faster the signal varies with time, or the more nonlinear the circuit is,
the more harmonics are needed to represent the solution accurately. In multi-tone mixer
cases, since the large LO tone has higher power level than moderate RF tones and causes
more nonlinear effects, usually more harmonics are used for large tone than moderate
tones.
Maximum harmonic also depends on the order of nonlinear effect user wants to study.
For example, for mixer IP3 measurement, maxhamrs of moderate tones must be at least
set to 2 in order to capture the IM3 mode at frequency 2ω1 − ω2 − ω LO .
3. tstab: Similar to time domain shooting method, tstab is a valid parameter for initial
transient analysis in HB. The default tstab for both PSS and QPSS is one cycle of signal
period. For QPSS analysis, user can choose the specific tone during tstab period and only
one tone is allowed. One additional cycle is run for FFT. If tstab is set to 0, dc results
will be used as initial condition for HB.
It should be pointed out that even with oversampling, user must make sure maximum
harmonic meets the accuracy requirement on results he is interested in.
As the simulation progresses, note messages in the simulation output log window that are
different from time domain qpss:
Action10-24: In the Direct Plot Form, select the qpss button, and configure the form the
same as for shooting engine.
Note: The Flexible Balance engine provides same results as with the shooting engine but
in this case in a much reduced simulation time. For multi-tone cases such as mixers, HB
is significantly more efficient than the shooting QPSS method due to its natural
representation of circuit equations.
For circuits driven by multi-tone stimulus, it is better to use HB QPSS other than HB
PSS.
As HB multi-tone simulation does not have the convergence and speed issues
encountered in shooting QPSS, HB QPSS should always be used to simulate multi-tone
circuits. HB PSS analysis using beat frequency as fundamental is very inefficient in
handling multi-tone cases. When source frequencies are closely spaced, their common
frequency is so low that hundreds or even thousands of harmonics must be used.
Action10-26: Close the waveform window and click Cancel on the Direct Plot form.
Rapid Ip2/Ip3 based on perturbation technology extends both shooting and flexible
balance. It is new feature of MMSIM6.0USR2. Rapid IM (IP2, IP3) calculations are an
order of magnitude faster than using flexible balance or shooting alone.
Action11-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action11-2: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type dc
PAC magnitude (dBm) pacm
Action11-8: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as shown in Action2-6
Action11-9: In the Choosing Analyses window, select the pac button in the Analysis
field of the window. Choose Rapid IP3 as Specialized Analyses. Set the
Input Source 1 to /PORT0 by select PORT0 on the schematic. Push the
ESC key on your keyboard to terminate the selection process. Set the Freq
of source 1 to 5001M and Freq of Source 2 to 5001.1M. Set the Frequency
of IM Output Signal as 0.9M and the frequency of Linear Output Signal as
1.1M. The form should look like:
Action11-10: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
As the simulation progresses, note messages in the simulation output log window that are
similar to the following:
Action11-13: Click on Plot Button. The calculated IP3 appears in the waveform
window:
Action11-14: Close the waveform window and click cancel on the Direct Plot form..
Action12-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type dc
PAC Magnitude (dBm) pacm
Action12-5: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action12-8: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as shown in Action2-6.
Action12-9: In the Choosing Analyses window, select the pac button in the Analysis
field of the window and set the form as follows:
In the above form, the Maximum Non-linear Harmonics is not specified, the default value
4 is taken.
Action12-10: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
As the simulation progresses, note messages in the simulation output log window that are
similar to the following:
Action13-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Action13-2: Select the PORT0 source by placing the mouse cursor over it and clicking
the left mouse button. Then in the Virtuoso Schematic Editor select
Edit—Properties—Objects… The Edit Object Properties window for the
port cell should come up. Set the source type to sine.
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source type DC
Action13-5: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action13-8: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as follows:
Action13-9: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the apply button.
Action13-10: In the Choosing Analyses window, select the pac button in the Analysis
field of the window. In the field of Specialized Analyses, choose rapid
IP2. Set the Input Source 1 to /PORT0 by select PORT0 on the schematic.
Push the ESC key on your keyboard to terminate the selection process. Set
the Freq of source 1 to 5001M and Freq of Source 2 to 5001.1M. Set the
Frequency of IM Output Signal as 0.1M and the frequency of Linear
Output Signal as 1.1M. The form should look like:
Action13-11: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
As the simulation progresses, note messages in the simulation output log window that are
similar to the following:
Action13-16: Close the waveforms window, click Cancel in the Direct Plot form.
Action14-1: If not already open, open the schematic view of the mixer_testbench
design in the library RFWorkshop
Parameter Value
Resistance 50 ohm
Port Number 1
DC voltage 500 mV
Source Type DC
Action14-4: From the Mixer_testbench schematic, start the Virtuoso Analog Design
Environment with the Tools—Analog Environment command.
Action14-7: In the Choosing Analyses window, select the pss button in the Analysis
field of the window and set the form as you did in Rapid IP2 simulation.
Action14-8: In the Choosing Analyses window, select the pac button in the Analysis
field of the window. In the filed of Specialized Analyses, choose IM2
Distortion Summary. Set the Input Source 1 to /PORT0 by select PORT0
on the schematic. Push the ESC key on your keyboard to terminate the
selection process. Set the Freq of source 1 to 5001M and Freq of Source 2
to 5001.1M. Set the Frequency of IM Output Signal as 0.1M. The form
should look like:
Action14-9: Make sure the Enabled button is on. In the Choosing Analyses window,
hit the OK button.
As the simulation progresses, note messages in the simulation output log window that are
similar to the following:
The Results Display Window shows the PAC IM2 Distortion Summary.
The distortion is listed dB for each instance. Due to very low RF input power, the
distortion is very small.
Conclusion
This workshop highlights how to use SpectreRF to simulate a Mixer efficiently and
extract design parameters such as IP3, 1dB compression point or port-to-port isolation.
Various techniques using PSS, Pnoise, PAC and QPSS analyses are being presented.
SpectreRF Flexible Balance and Time domain algorithms accuracies are also
demonstrated and compared.
.
Reference
[1] "The Designer's Guide to Spice & Spectre", Kenneth S. Kundert, Kluwer
Academic Publishers, 1995.
[4] "The Design of CMOS Radio Frequency Integrated Circuits", Thomas H. Lee.
Cambridge University Press, 1998.