One of the most useful RF or radio frequency processes is that of mixing. Unlike an
audio mixer where signals are simply added together, when a radio or RF engineer
talks about mixing, he means a whole different process. Here signals are multiplied
together and signals an new frequencies are generated.
The process of RF or non-linear mixing or multiplication is used in virtually every
radio set these days and also in many other circuits beside. It enables signals to be
changed from one frequency to another so that signal processing for example can be
undertaken on a low frequency where it is easier to perform, but the signal can be
changed to a from a higher frequency where the signal is to be transmitted or
received.
In fact RF mixing is one of the key processes used in RF circuit design and mixers
are seen in very many circuits and items of equipment associated with radio
frequency technology.
What happens when signals are mixed
It is found that if two signals are passed through a non-linear circuit, then additional
signals on new frequencies are formed. These appear at frequencies equal to the
sum and difference frequencies of the original signals.
In other words if signals at frequencies of f1 and f2 enter the mixer, then additional
signals at frequencies of (f1+f2) and (f1-f2) will also be seen at the output.
To give an example if the two original signals are at frequencies of 1 MHz and 0.75
MHz, then the two resultant signals will appear at 1.75 MHz and 0.25 MHz.
Mixing two RF signals
Why RF mixing or multiplication works
To understand a little more about the RF mixing or multiplication process it is
necessary to look at exactly how the mixing process occurs. As mentioned before
the two signals are actually multiplied together, and this occurs as a result of a nonlinear element in the circuit. This may be a diode, or active devices such as
transistors or FETs that are suitably biased.
The two signals can be considered as sine waves. The instantaneous output level is
dependent upon the instantaneous level of signal A multiplied by the instantaneous
level of signal B. If points along the curve are multiplied, then the output waveform is
more complex as shown below.
Mixing or multiplying two signals together
The frequencies used to generate the example below for the frequencies mentioned
above, i.e. 0.75 MHz and 1.0 MHz. It can be seen that in the output there is a low
frequency component (the difference frequency at 0.25 MHz) and high frequency
component (the sum frequency at 1.75 MHz).
In operation, RF Mixers use one of two mechanisms for their operation:

Nonlinear transfer function: This approach uses device nonlinearities
creatively in a manner that intermodulation creates the desired frequency and
unwanted frequencies.

Switching or sampling This is a time-varying process in which elements
ofthe mixer are switched on and off by the local oscillator. This method is
preferred because creates fewer spurious signals and hence it provides
higher linearity for the required output signals.
RF / frequency mixer ports
Frequency mixers of RF mixers come in a variety of formats, but they all have the
same basic connections. There are three, and on many frequency mixer modules,
they are labelled as such.:

RF: This is the input used for the signal whose frequency is to be changed. It
is typically the incoming signal or equivalent and it is normally at a relatively
low level compered to the other input.

LO: This is for the local oscillator signal. The signal input level for this port is
generally much larger than that for the RF input.

IF: This is the output port for the mixer. It is the port where the "mixed" signal
appears.
In an RF design or system where the signal is being converted to a band where the
signals are lower in frequency than the incoming signal the circuit block can be
referred to as a downconverter, or a down-conversion process. This typically
happens in a receiver (although in some radios, signals can be converted up in
frequency before they are converted back down again).
Similarly, when the signals are being converted up in frequency, the process can be
referred to as up-conversion. This typically occurs in a transmitter and some other
RF systems.
Dependent upon the actual RF mixer and the application, the local oscillator signal is
typically quite large and may be a continuous sine wave, or a square wave. This
local oscillator signal often acts as a gate to the mixer, switching the mixer in line
with this signal.
The RF mixer can be considered ON when the LO voltage switches it on and OFF
when the local oscillator signal switches it off. This then acts upon the incoming
signal on the RF port to enable the two signals to mix and provide the two output
signals required.
Types of RF mixer
RF mixers or frequency mixers are available in many forms and there are several
types of terminology used to categorise them. Obviously there are the mixers based
upon different forms of semiconductor or other technology, but they are also
categorised in other ways.
One way in which RF mixers are described is in terms fo the type of device used in
them:

Passive mixers: Passive mixers typically use passive components in the
form of diodes as the switching element within the RF circuit. As a result they
cannot exhibit any gain, but many forms can provide excellent levels of
performance.
Passive mixers mainly use Schottky diodes because of their low turn-on voltage, but
they require the use of a balun / RF transformer if they are to be used in a balanced
or double balanced mixer. This can limit the frequency response.

Active mixers: As the name of the Active RF mixer contains active
electronic components like a bipolar transistor, FET or even a vacuum tube /
thermionic valve. These types of RF mixer are able to provide gain as well as
proving the multiplication or RF mixer capability.
Mixers are also looked at by whether they are balanced or not. Balancing them
requires baluns - balanced to unbalanced transformers - but this provides
improvements in performance.

Unbalanced mixer: An unbalanced RF mixer is one in which the mixer
simply mixes the two signals together and the output consists of the sum and
difference signals as well as significant levels of the original RF signal and
that of the local oscillator. In some instances this may not be an issue, but in
others it can really help to have these removed as part of the frequency
mixing process.

Single balanced mixer: A single balanced mixer has a single balun or
balancing circuit. Typically single-balanced mixers consist of two diodes along
with a hybrid which acts as a balun. Although 90° and 180° hybrids can both
be used to design single-balanced mixers, the majority of single-balanced
mixers incorporate a 180° hybrid.
The 180° input ports of the hybrid are mutually isolated and this enables the local
oscillator port to be isolated from the RF port which prevents the LO signal from
affecting the RF input circuits reducing the level of intermodulation products.
Balanced operation can also be achieved using balanced transistor or FET
configurations.
These are typically contained within integrated circuits where high levels of
performance can be achieved.

Double balanced mixer: The basic traditional double-balanced mixers
typically use four Schottky diodes in a quad ring configuration. The baluns or
hybrids are placed at both the RF and LO ports, while the IF signal is tapped
off from the RF balun.
In operation the double balanced mixer has a high level of LO-RF isolation and LOIF isolation and it provides a reasonable level of RF-IF isolation. The use of double
balanced mixers can reduce the level of intermodulation products by up to 75%
when compared to a single diode unbalanced RF mixer.
Like the single balanced mixer, the double balanced mixer can also be replicated
using balanced modes of operation within transistor or FET circuit designs. When
contained within integrated circuits, these circuit often utilise a double balanced
mixer configuration as the additional circuitry required can be incorporated into the
OC for negligible increase in cost.

Triple balanced mixer: To improve mixer performance still further it is
possible to use a triple balanced mixer.
A triple-balanced mixer is effectively made from two double balanced mixers and as
a result it is sometimes called a doubly double balanced mixer. It utilises many more
electronic components having two diode bridges or quads, with a total of eight
junctions. Power splitter at the RF and LO microwave baluns feed the mixer
structure, and this allows for both of the diode quads to be coupled. This allows for
the IF signal to be available at two separate isolated terminals, that typically exhibit
very large bandwidths compared to other mixer architectures.
The improved isolation provided by the triple balanced mixer provides for much
higher levels of spurious signal, intermodulation distortion suppression.
The improvement in performance needs to be offset against the fact that they need
higher levels of LO drive, and of course the increased complexity and electronic
component count result in increased cost.
RF mixer circuit symbol
The key RF mixer circuit symbol shows the two signals entering circuit block
consisting of a circle with a cross or "X" within it. This is widely used in the circuit
schematics for many RF circuit designs. It is typically used when an RF mixer
module is used.
This circuit symbol indicates the multiplication aspect of the mixer.
RF mixer circuit symbol showing the frequency translation scheme that is the
goal for good mixers
In some instances the different ports to the mixer will be suitably labelled: RF, LO, IF.
RF mixer circuits
RF mixers or frequency mixers can be realised using a variety of RF circuit designs.
Also different circuits have different levels of complexity and use different numbers
and types of electronic component. Accordingly the cost, specifications, operation
and other aspects mean that when undertaking any RF circuit design, the different
types of frequency mixer may be more applicable to one situation than another.
There is an enormous variety of different types of circuits including:

Single diode mixer: This form of RF mixer or frequency mixer is the
simplest form available, using very few electronic components. Accordingly
the level its performance far less than that of some more sophisticated
designs using additional and often more expensive electronic components.

Basic transistor RF mixer:
Read more about . . . . Bipolar Transistor RF Mixer.

Basic FET mixer:
FETs are ideal electronic components to used for mixing. Having a good switching
capability and the ability to use two gates if a dual gate MOSFET is used, these
devices provide excellent performance.
There are many different circuits for FET mixers, each one having its own
advantages and disadvantages.
Read more about . . . . FET RF Mixer.

Single-balanced diode mixer: The single balanced diode mixer provides
isolation of the local oscillator from one of the other ports. It is straightforward
and works well, although because of the limited isolation between ports, it will
give rise to higher levels of intermodulation distortion.

Double-balanced diode mixer: The double balanced diode mixer provides
increased isolation - isolating the the LO-RF and LO-IF ports. It requires two
baluns and four diodes. The diodes used are normally Schottky diodes
because of their low turn on voltage. In view of the increased isolation
capabilities, the levels of intermodulation distortion are lower than those of the
single balanced mixer.
Read more about . . . . Double Balanced Diode RF Mixer.

Gilbert cell mixer: The Gilbert cell mixer is often used within integrated
circuits that are used for radio receiver and other RF design applications. In
view of the number of electronic components needed, they are not seen so
often being built from discrete electronic components. The Gilbert cel mixer
performs particularly well, being able to offer double balanced operation using
the differential inputs, etc of long tail pair transistor or FET circuits.
Read more about . . . . Gilbert Cell Mixer.
RF Mixer applications
RF mixers or frequency mixers are used in all areas of RF design, and development.
They are used in circuits from radio receivers and transmitters to radar systems, and
in fact anywhere that radio frequency signals are used.
These mixers can be used in a variety of different ways:

Frequency translation: The most obvious application for RF mixers is for
frequency translation. This technique is used in many areas and in particular
in receivers and transmitters to move the frequency of a signal from one band
to another. Using the fact that the two input frequencies generate sum and
difference frequencies, it is possible to change the signal input to another
frequency by taking either the sum or the difference signal. One of the first
major applications of this was in the superheterodyne radio receiver.

Phase comparison: Using a mixer it is possible to detect the phase
difference between two signals. This RF mixer application can be used in
many areas, one of which is within phase locked loops.
RF mixers or frequency mixers as they are often called, are one of the main building
blocks for RF circuit design. Frequency translation is a major capability, used in a
host of different applications, and it is a key element of radio communications
equipment technology: both for transmitters and receivers. IN addition to this, mixers
can be used as phase detectors for many applications including many phase locked
look and synthesiser RF designs.
Accordingly understanding RF mixer operation and the different types and their use,
is essential for anyone involved in RF design, systems development, or the operation
of any RF or radio communications equipment.
The performance of an RF mixer can be a pivotal element in the overall operation of an RF circuit
design or system: selection of the correct mixer is key to the design.
Although many designs use small active mixers within the overall circuit made from
discrete electronic components, for many other designs high performance mixer modules or
integrated circuits are the answer.
Whether a mixer designed from discrete components, or a module or integrated circuit that is
bought in, the specification and performance of the mixer is key.
Under specify the mixer and the performance of the whole RF circuit design may be compromised.
Over-specify and costs are increased. Select the wrong type and even though it is high
performance electronic component it may not work in the way that is intended, or some element
of its performance may be wrong.
Selecting the right RF mixer is a key stage in the overall RF circuit design. With many hundreds or
thousands to choose from, and from a variety of manufacturers an ordered selection process is
essential.
RF Mixer basics
When looking at RF mixers for any RF circuit design there are a number of definitions that will be
of interest.
Aspects like the ports, form of mixer and the like all play an important part of the specification.
There are three ports to any RF mixer / frequency mixer:



RF: This is the input used for the signal whose frequency is to be changed. It is
typically a low level signal.
LO: This is the local oscillator signal and is at the specified level, higher than that of
the RF input.
IF: This is the output port for the mixer.
RF mixer circuit symbol showing the different port names
There are also various forms of RF mixer which need to be understood. One of the first relates to
the type of electronic components or devices within the mixer:

Passive mixers: Passive mixers typically use passive electronic components in the
form of diodes as the switching element. As a result they cannot exhibit any gain, but
many forms can provide excellent levels of performance.
Passive mixers mainly use Schottky diodes because of their low turn-on voltage, but
they need balanced transformers for most high performance designs and this can
limited the frequency band over which they can operate.
One key aspect of passive mixers is that they introduce what is called a conversion
loss, explained later, and this can have an impact on the RF circuit design.

Active mixers: As the name of the active RF mixer implies, it contains active
electronic components like a bipolar transistor, FET or even a vacuum tube / thermionic
valve. These types of RF mixer are able to provide gain as well as proving the
multiplication or RF mixer capability.
Unlike passive mixers, active mixers can actually have a conversion gain, and this will
have an affect on the RF design for the item.
RF mixers or frequency mixers can also be categorised according to whether they are balanced
or unbalanced. This is an important decision to make.

Unbalanced: An unbalanced RF mixer is a basic form of RF mixer and one in which
it simply mixes the two signals together and the output consists of the sum and
difference signals as well as significant levels of the original RF signal and that of the
local oscillator. As there is little isolation between the ports this can lead to increased
levels of intermodulation distortion as well as the local oscillator and RF signals being
present on the output.
 Balanced: A balanced mixer is one in which the ports have a balanced or differential
structure. Dependent upon the actual type there can be isolation between the different
ports, and the LO and RF can be suppressed at the IF port. There are different types
of balanced mixer: single balanced; double balanced and triple balanced (more
correctly termed a doubly double balanced mixer).
Selecting the right type of RF mixer to meet the requirements of the circuit design is one of the key
choices to be made.
Mixer package type specification
This decision is one of the first that can be made. The connection technology and requirements
will be known early in the design. There are generally three types of package type:




Surface mount technology: RF mixers using surface mount technology are probably
the smallest types in terms of area and can be mounted directly onto a printed circuit
board. These are ideal where the whole circuit or system is printed circuit board based.
However it is necessary to be aware of any special soldering restrictions, especially in
terms of the solder reflow temperature, etc..
Leaded component: Some RF mixers will be available in the traditional leaded styles.
These are normally used for low volume, through hole mounting on printed circuit
boards.
Connector: In some instances a connector-ised RF mixer will be required, Often
these come with either BNC or SMA connections, but other connectors may be
requested including N or TNC types, but these tend to be less common or they may
need to be requested as special items.
These mixers tend to be used in larger rack based systems. Consideration of the size
and connector type is necessary when choosing these options. Consider also the way
these mixers will be mechanically mounted because many mixer manufacturers offer
various options for this.
Plug-in: These mixers are through-hole mounted units. They have at least four pins
and this enables them to be securely connected both electrically and mechanically.
These may be used on through hole printed circuit boards. Typically these mixers have
at least four pins, one each for the three signal lines and one for earth, although many
may provide an earth or ground connection with each signal port.
Mixer local oscillator level
The local oscillator or LO input level is another key parameter to be considered. It may be a key
factor in determining which set of mixers, or the mixer itself.
The higher the local oscillator input level, the higher the RF level that can be accommodated
without running into issues with distortion, etc. Typically the local oscillator input should be 10dB
above the highest anticipated RF signal. This keeps the mixer running within its linear operating
range.
Mixer modules tend to be specified at various common levels, e.g. 7dBm, 10 dBM, 17 dBm, etc.
These are sometimes referred to as level 7, level 10 or level 17 mixers. Other values are available
for these mixers dependent upon the application, but these levels possibly form the most widely
used values.
Unfortunately the higher power mixers tend to be more expensive, and amplifying the LO to the
higher level so there is often a trade-off between performance and cost. Keeping the lowest LO
level will not only keep the cost down, but also result in lower LO leakage within the system as
well.
It is best to drive these mixers at levels approximately equal to the required drive input. Higher than
this will particularly result in greater levels of LO leakage and other performance parameters may
fall off.
Lower than the required level, then the performance again falls, typically providing an increase in
conversion loss. Running a mixer with the local oscillator at around -3dB of the required level may
increase the conversion loss by 0.5dB or so. Also the third order intermodulation performance may
be degraded slightly - which is hardly surprising since the diodes will not be switching as hard.
Mixer 1dB compression point specification
The 1dB compression point of a mixer is very important specification where spurious signals are
concerned.
An ideal mixer would operate linearly, i.e. for every 1 dB increase in the RF input level, the output
from the IF port would also increase. However a point is reached where the output cannot handle
the signal, and it starts to level out. The 1 dB compression point, is the point at which the output
deviates from the linear curve by 1 dB, i.e. it is 1 dB less than the plotted linear line. The
specification normally refers to the RF input power level at which this compression occurs.
The 1 dB compression point is easy to measure and it provides a useful comparison between mixer
to see what their high level performance is like. Obviously for high level signals the higher the 1
dB compression point the better.
The 1 dB compression point is linked in to other mixer parameters as well.
Maximum RF port power specification
In any RF circuit design, a power budget may be prepared showing the power levels at different
stages of the circuit. Knowing how the power level varies, it is often possible to accurately
determine the maximum power level entering the RF port of the mixer.
With a knowledge of this figure, selecting the required mixer is simply a case of choosing the mixer
whose 1 dB compression point exceeds this value.
In terms of inputs where the signal levels varies over a very wide range, it is very important to
ensure that the level does not exceed a safe value. This can be exemplified in that one of the major
problem areas on some older spectrum analysers with no automatic input protection was the
destruction of the input mixer when high level signals were applied when the engineer forgot to put
an attenuator in circuit.
Conversion gain or loss
The conversion gain for a mixer is very important when undertaking the RF circuit design for a
project as it will determine the signal levels after the RF mixer.
The conversion gain or loss is defined as the ratio of the desired output level to the RF input signal
level.
It can be seen from this that the local oscillator level does not feature in this figure - the conversion
gain is only interested in the levels of the wanted input and output signals.
As might be expected, passive mixers feature a conversion loss. Dependent upon the mixer in
question this might be around 7dB or so, but it is very dependent upon the actual mixer itself.
Active RF mixers normally feature a conversion gain, and again the level is very dependent upon
the actual mixer itself.
The main issue is that the level of the conversion gain, or loss is known so that the appropriate
action can be taken in the earliest stages of the RF circuit design.
Frequency range
Although RF mixers tend to support wide-band operation, the actual frequency range to be used
must obviously be covered by the mixer. Again if the mixer is over-specified in terms of either /
both the bandwidth and top frequency, then costs may be more than they need to be.
Typically it is good practice for any RF circuit design is to select a mixer where the mid-band
frequency range covers the intended operating range.
That said, the performance of many mixers extends outside their specified ranges, although with
some increasing degree of degradation the further outside the operating range the frequency is.
Isolation
The isolation level between the ports is often important and it states the level of what may be
termed the leakage between the different ports. The RF and local oscillator is not normally needed
at the IF, and if, for example the local oscillator leaks through to the RF port, it could give rise to
intermodulation distortion.
As might be expected the isolation is measured in terms of dB, comparing the signal entering one
port, to the same signal level at the other port where it is not required.
It is found that mixer isolation tends to deteriorate with increasing frequency as the reactance of
stray capacitance falls, and also the circuit imbalances become more apparent.
Third order intercept point, IP3 & third order
intermodulation
One major issue with any RF mixer is the level of unwanted signals that are generated within the
mixing process. Non-linearities within the mixer give rise to additional signals and these can cause
issues in many ways dependent upon the circuit design or system in which they are used.
The third order intercept point of a mixer (or amplifier) is a hypothetical point where the power of
the third order products will have the same power level as the fundamental.
The third order intercept point of a mixer of any other device is theoretical because it lies well
beyond the saturation level of the device, and it many cases it would be well beyond the point at
which damage occurred, especially in the case of a mixer.
The reason that the IP3 figure is useful is that it provides a very good guide or figure of merit for
the distortion generated by the device as the power levels rise.
The IP3 point can be defined for either the input or output ports.
There are two main ways of defining the intercept points:

Based upon the intermodulation products: The most commonly used approach for
determining the IP3 of the RF mixer. For this, the mixer is given two sine wave signals
that have a small frequency difference.
The spectrum of intermodulation products from two signals
The intermodulation products then appear at spacing equal to the input tones, and the
levels can be measured. The third order products appear at three times the frequency
spacing of the two signals either side of them.

Based upon harmonics: An alternative method is to use a single signal, and then the
products appear at multiples of the input tone. The third order product is at tree times
the fundamental.
The input third order intercept point is often designated as IIP3 and the one of the output is
designated OIP3. These intercept points differ in level by an amount equal to the small signal gain
(or loss) of the mixer.
The selection of an RF mixer for any RF circuit design can have a major impact on the
performance. Accordingly it is important to ensure that its performance meets the needs for the
particular RF design in terms of its electrical performance, environmental specification, mechanical
and connector parameters, as well as aspects such as whether it in a surface mount technology
format for printed circuit board mounting and large scale production, etc.
Normally selecting the right mixer is a balance between aspects like electrical performance,
mechanical aspects, and cost. This is not always easy, but by understanding the specifications
and the impact they have on the overall performance of the RF circuit design, then the best
compromise can be selected.
Double balanced mixers are able to provide very high levels of performance in RF or frequency
mixing applications.
The action of the double balanced mixer means that the input RF and local oscillator signals are
“balanced out” and their level is considerably reduced at the output by having differential circuits
on their inputs.
This reduces the need to remove the often unwanted RF and local oscillator signals at the output
and reduces the effect of these input signals causing intermodulation distortion.
Double balanced mixers can either be made from the basic electronic components, or they may
also be bought as modules for inclusion in a circuit - the latter approach is often adopted because
the purchased modules will have been designed, optimised and manufactured by specialist
manufacturers ensuing the highest performance.
In view of the level of their usage, double balanced mixers are widely available from a number of
specialist RF component suppliers. These suppliers have a wide range of double balanced mixers
both as hybrid diode and FET based mixers as well as fully integrated MMIC based devices that
should meet the requirements for the majority of RF circuit design applications.
Need for balanced mixers
Many forms of mixer are not balanced and as a result they allow through considerable levels of
the local oscillator and RF signals. These are normally not wanted and normally they would have
to be removed by filtering which is often inconvenient and expensive. The solution is to balance
the mixer to remove the input signals.
There are two types of RF mixer that are balanced:

Single balanced mixer: Often called just a balanced mixer, this type of mixer will
suppress either the LO or RF signal but not both
 Double balanced mixer Unlike the single balanced mixer, the double balanced mixer
suppresses both of the input signals.
While single balanced mixers offer many advantages over simpler designs, the double balanced
mixer is more widely used. However there are a number of advantages and disadvantages over a
single balanced mixer to consider:
Double balanced mixer advantages.

Increased linearity.

Better suppression of spurious products - all even order products of the LO and RF
inputs are suppressed.

Isolation between all ports.
Double balanced mixer disadvantages.

Higher level LO drive level is often required.

At least two baluns are required within the design - these add cost and complexity
Despite the increased complexity, double balanced mixers are more widely used for applications
where high performance is paramount.
RF / frequency mixer ports
Like all other RF mixers, double balanced mixers have the same three ports or connections.



RF input: This port on the mixer is connected to the incoming signal that is to have its
frequency converted.
Local Oscillator or LO input: This port takes in the internal local oscillator signal that
is used to convert the RF signal to the new frequency.
IF output: The third port of the double balanced mixer is normally referred to as the
IF or intermediate frequency output. The signal on the output of an ideal RF mixer
should contain only the mixer products, i.e. the sum and difference frequencies of the
two input signals.
RF mixer circuit symbol
Types of double balanced mixer
Double balanced mixers come in a variety of forms using different types of electronic
component and having some slightly different formats.
 Hybrid based diode double balanced mixer: This type of double balanced mixer is
the stereotypes double balanced mixer. It uses two main electronic components namely
a ring of four diodes, which are normally Schottky diodes and baluns on two or more
ports. It is primarily these baluns that add the cost and also limit the frequency response
of the RF mixer.
 Hybrid based active double balanced mixer: This form of double balanced mixer
replaces the diodes with an active device to act as the switching elements within the
ring circuit. Again hybrid baluns are used on both input circuits, but the format is again
the same. The electronic components include the expensive wound baluns as well as
the active devices, e.g. FETs which have a good switching performance.
 Active double balanced mixer: Using active differential amplifier circuits, it is
possible to achieve the balanced operation for the RF mixer. This enables the complete
mixer circuit to be fabricated on a single semiconductor chip. Most high performance
RF mixers, as well as many lower performance ones use this technology. The cost of
the high performance active double balanced mixers is much less than those using
wound components, and they offer a much wider bandwidth.
In view of the better RF performance and lower cost of active double balanced mixers using no
hybrids, this type of mixer is used in most RF circuit design applications these days. Mixers are
available in standard surface mount technology packages for many applications and for more
exacting ones, special monolithic microwave ICs, MMICs are available for RF mixers.
Reversing switch mixers
Double balanced mixers are a form of what is termed a "reversing switch mixer." Reversing switch
mixers operate by using electronic switches in a bridge formation to reverse the input RF signal
under the action of the local oscillator used as a square wave switching signal. They normally offer
significant advantages over analogue mixers for radio communications and general RF design
applications as they are able to offer better levels of dynamic range and noise. In view of this fact,
they are normally used in high performance applications where noise and dynamic range are of
importance - e.g. in the front end of a radio receiver or spectrum analyzer.
Radio & RF equipment
Double balanced mixer basics
The most common form of double balanced mixer is the diode double balanced mixer. Essentially
a double balanced mixer has both its inputs applied to differential circuits, so that neither of the
input signals and only the product signal appears at the output.
As there are two balanced ports it is called a double balanced mixer.
In its simplest form it consists of two unbalanced to balanced transformers and a diode ring
consisting of four diodes as shown, although other circuits using active devices etc can be devised.
Double balanced mixer circuit
Although the design of the RF mixer looks straightforward, high performance mixers are designed
and built to exacting standards to achieve the high levels of performance needed.
One of the key specifications for a double balanced mixer is whether any of the LO or RF signals
appear at the IF port. This depends upon the diode and transformer uniformity. In addition to this
the circuit offers high isolation between the RF and IF ports because the balanced diode switching
precludes direct connection between T1 and T2.
Diode double balanced mixer components
Although there are comparatively few components in a double balanced mixer, their individual
performance is crucial to the performance of the RF mixer as a whole.
Normally Schottky barrier diodes are used for the diode ring. They offer a low on resistance and
they also have a good high frequency response. Ordinary signal diodes may be used for low
performance applications, although the cost difference is small. It is found that the diode forward
voltage drop for the diodes determines the optimum local oscillator drive level. RF mixers requiring
to handle a high RF input level will need a correspondingly high LO input level. As a rule of thumb
the LO signal level should be a minimum of 20dB higher than either the RF or IF signals. This
ensures that the LO signal rather than the RF or IF signals switch the RF mixer, and this is a key
element in reducing intermodulation distortion, IMD, and also maximising the dynamic range.
To increase the required drive level, it is possible to place multiple diodes in each leg. The most
common LO drive level for a double balanced mixer is probably +7dBm. However they can be
obtained with a variety of drive levels. Values of 0, +3, +7, +10, +13, +17, +23, and +27 dBm are
normally available.
In order to provide the required level of performance, the quad diodes used win these mixers are
generally fabricated monolithically. By doing this they will have very closely matched performance
parameters, and in particular the level of forward voltage will be virtually identical in all the diodes.
The transformers are also critical to the performance of the RF mixer. Creating a wideband balun
for the mixer is one of the key elements within the overall mixer design and achieving the required
bandwidth and performance can be difficult to achieve.
The matching of the transformers and the individual legs are important in determining the balance
of the RF mixer. The transformer also plays an important role in determining the conversion loss
and drive level of the RF mixer. As the transformers are wound on a ferrite core, the core loss,
copper loss and impedance mismatch all contribute to the transformer losses.
Double balanced mixer operation
The operation of the double balanced mixer is relatively easy to understand. The local oscillator,
LO, signal turns on first one arm (D3, D4), and then the other (D1, D2) within the diode ring.
As the points where the LO signal enters the diode ring at the junction of D1 and D4 appear as a
virtual earth to the RF signal, this means that the points where the RF signal enters are alternatively
connected to ground as the diodes turn on and off.
The operation of the mixer means that the RF signal with alternating inverse phases is routed to
the IF port according to the switching action of the local oscillator - in other words the signal at the
IF port has been multiplied by the local oscillator waveform.
Active double balanced FET mixer
While diode mixers are able to offer excellent performance, the increase in use of wireless and
general radio communications systems means that receivers need to be able to accommodate a
larger number of local strong signals than may have been the case previously. Better low end
noise performance along with higher third order intercept points are required. The performance
figures required by double balanced diode mixers cannot always meet the requirements for some
designs, unless significant tailoring is undertaken and this increases the costs beyond economic
viability. Conventional double balanced diode mixers can offer a third order intercept performance
up to figures of between about +25 and +30 dBm.
Radio & RF equipment
To offer an alternative to the diode mixer, it is possible to use a double balanced FET mixer. Welldesigned FET mixers are able to offer extremely linear performance along with high third order
intercept points - some as high as +38dBm.
Double balanced FET mixer
The diagram shows the basic concept of a double balanced FET mixer. However some mixers
require the application of a DC bias to ensure the correct switching of the diodes, and some mixers
show a high conversion loss or noise figure. Double balanced FET mixers using discrete
components can sometimes be optimised to provide better performance figures, and newer
commercially available items are also offering better performance.
Practical aspects for double balanced
mixers
Using double balanced mixers enables very high levels of performance to be achieved. However
a few useful hints and tips ensure that the mots can be made from using them.




Use the right drive level: In order to ensure the correct operation of the mixer it is
necessary to ensure that the correct specified drive level is used. In this way the diodes
in the RF mixer will switch correctly.
Choose the right level mixer for the RF design: In a similar vein to using the
specified drive level, the particular RF mixer should be chosen so that the drive level is
sufficiently high for the particular RF design. Normally the LO drive should be at least
20 dB higher than the highest expected RF or IF signal anticipated. This will ensure the
optimum IMD and dynamic range.
Ensure the ports are accurately matched: Diode double balanced mixers are
termination impedance sensitive. They must be terminated with the correct resistive
load or source impedance (normally 50 ohms). A wideband resistive output is
particularly important if it is to achieve the highest dynamic range. This can be achieved
by using an attenuator pad in the line. Although this can be used for the LO port, this
approach is not normally suitable for the RF and IF ports as it would impair the noise
figure. Instead accurate matching of the amplifier stages preceding and following the
mixer is one solution.
Tap off the IF from the RF balun: By tapping of the IF output from the RF input, it is
possible to achieve a far greater level of LO rejection - typically 20dB.
Double balanced mixers are widely used in RF applications where performance is paramount.
Although they are more costly than many other forms of frequency mixer, they offer the
performance that is often required Typically these mixers are bought from specialist manufacturers
because they have the expertise and the manufacturing capability for the transformers and other
components that are required. Additionally the development and optimisation cost is spread over
a large number of units.