Signal injection technique for measurement and control of source reflection coefficient of a device under test

Abstract

A method for measuring (and controlling) a characteristic performance parameter S of a device under test (DUT) 10 having an input port(at the minimum). The method comprises connecting the input port of the DUT 10 to a signal generator 1, subjecting the DUT 10 to a large signal input test signal, and executing a first measurement a1′, b1′ of the incident wave a1 and reflected wave b1 at a DUT input reference plane. The method further comprises subjecting the DUT 10 to a perturbation signal combined with the large signal input test signal, and executing a second measurement a1″, b1″ of the incident wave a1 and reflected wave b1 at the DUT input reference plane, and determining the characteristic performance parameter from the first measurement and the second measurement.

Description

FIELD OF THE INVENTION

The present invention relates to a method for measuring and controlling a characteristic performance parameter of a device under test (DUT) having an input port, such as a semiconductor active component in the form of a transistor, a FET, etc., operating at RF, microwave or (sub)millimetre-wave frequencies. In this document, characteristic performance parameters of the DUT are understood to be a source reflection coefficient Γ_S of the DUT.

BACKGROUND ART

U.S. Pat. No. 8,269,508 describes a method for calculating the performance of a transistor with any source impedance Γ_S. This method does not physically apply nor measure a different source impedance to the transistor, but it's a mathematical calculation done in a post-processing. Furthermore, the method as disclosed assumes that the transistor is perfectly unilateral (S₁₂=0), which is rarely true.

US patent publication US2002/0196033 describes a method to measure the “hot-S22” of a DUT wherein the DUT is stimulated by a one-tone excitation signal at the input port, and the S22 under large-signal is measured by applying a probe tone at an offset frequency at its output.

U.S. Pat. No. 7,038,468 describes a method to measure all the DUT hot S-parameters by using probe tones at an offset frequency.

The article by Jan Verspecht et al. ‘Characterizing Components Under Large Signal Excitation: Defining Sensible “Large Signal S-Parameters”; 19^th ARTFG, 1 Jun. 1997, pages 109-117, discloses a measurement technique enabling characterization of nonlinear microwave components under periodic large-signal excitation. A black-box model of the component is deduced from large-signal measurements.

US patent publication US2006/161409 discloses a behavioral model generation, wherein a device under test is stimulated with a large amplitude signal having a control frequency, and perturbed with a small amplitude signal tone, which is at a frequency offset slightly from a harmonic of the central frequency. The small amplitude signal tone at offset frequency is necessary to be able to extract a model of the device under test.

US patent publication US2009/184722 discloses a method of measurements including large-signal S scattering functions of a device under test having two distinct phases for small signals on a frequency grid established by intermodulation frequencies and harmonics of large tone signals.

International patent publication WO2007/081705 discloses a method of characterizing nonlinear behavior of amplifiers using load pull measurements, using a behavioral modelling technique.

SUMMARY OF THE INVENTION

A device under test (DUD can be represented in terms of travelling waves or in terms of its scattering matrix S. With respect to an active DUT, the scattering matrix S is sufficient enough to describe its linear behaviour under small signal conditions and in a 50Ω environment. However, when moving to a large-signal condition, the device performance changes non-linearly as a function of input power drive and as a function of the terminations provided at its input (source) and its output (load). Therefore, for an active DUT, it is important to measure certain characteristic performance parameters as a function of input power drive and of the source and load reflection coefficients Γ_S and Γ_L presented at the input and output of the active DUT. However, when stimulating a DUT from its input with a signal at a given frequency only, Γ_IN and Γ_L can be only measured as

${\Gamma }_{\mathrm{IN}}=\frac{b_1}{a_1}\mathrm{and}{\Gamma }_L=\frac{a_2}{b_2}.$

The present invention seeks to provide a method to measure characteristic performance parameters of a DUT, more in particular a method to control and measure the source reflection coefficient of a DUT in a reliable, consistent and robust manner, without the need for any additional hardware, such as sources or impedance tuners.

According to the present invention, a method is provided for measuring characteristic performance parameters Γ_S of an active device under test (DUD having an input port (or multiple input ports). The DUT operational behaviour is described by a scattering matrix S relating an input incident wave a₁ and an input reflected wave b₁ at the input port of the DUT, and/or an output incident wave a₂ and an output reflected wave b₂ at the output port of the DUT. The method comprises connecting the input port of the DUT to a signal generator, subjecting the DUT to an input test signal, and executing a first measurement a₁′, b₁′ of the incident wave a₁ and reflected wave b₁ at a DUT input reference plane. The method further comprises subjecting the DUT to a perturbation signal combined with the input test signal (or equivalently changing the input power to the DUD, and executing a second measurement a₁″, b₁″ of the incident wave a₁ and reflected wave b₁ at the DUT input reference planes, and determining the characteristic performance parameter from the first measurement and the second measurement, wherein the characteristic performance parameter is a source reflection coefficient Γ_S.

The present invention provides method embodiments that allow to measure and control the source reflection coefficient Γ_S presented to a DUT as a function of the input power drive and/or of load reflection coefficient Γ_L, while the device is stimulated from its input port. Thus, the present invention provides a method to perform source pull in any test system having a signal source and an input coupler section, without the need for any additional hardware. Also, the method does not use excitations at offset frequencies. Further embodiments are described by the dependent claims. An advantageous aspect of the present invention is that the control of the source reflection coefficient is achieved by physically using a single signal source (the same source used for generating the input test signal to the DUD, and without the need for impedance tuners (e.g. passive mechanical or electronic tuners).

SHORT DESCRIPTION OF DRAWINGS

The present invention will be discussed in more detail below, with reference to the attached drawings, in which:

FIG. 1 shows a schematic representation of a DUT having an input and output port in a measurement set-up according to an embodiment of the present invention;

FIG. 2 shows a schematic representation of a measurement system according to a further embodiment of the present invention;

FIG. 3A-C show graphical representations relating to exemplary measurements executed according to an embodiment of the present invention.

DESCRIPTION OF EMBODIMENTS

The present invention embodiments relate to a method for measuring a characteristic performance parameter Γ_S of an active device under test (DUT) 10 having an input port. The DUT 10 can be a semiconductor component such as a transistor or a Field Effect Transistor (FED. FIG. 1 show schematic representations of such a DUT having an input and output port, and FIG. 2 shows a typical measurement set-up according to an embodiment of the present invention.

The present invention method embodiments can be applied on any measurement system comprised of a signal generator 1 and one or more receivers 4 to measure a₁, b₁ such as a Vector Network Analyzer (VNA) or a load-pull system. FIG. 2 shows a schematic representation of a measurement system, according to one embodiment of the present invention, which depicts a load-pull measurement system based on a four-channel receiver 4 (e.g. a VNA) and a generic load tuner 5. A signal generator 1 is connected to the DUT 10. The DUT 10 can be stimulated at any given frequency f₁ by the signal generator 1. The incident, reflected and scattered waves (a₁, b₁, a₂, b₂) are coupled with an input coupler 2 and an output coupler 3 and measured by a VNA 4. The output load impedance Z_L can be tuned by use of any generic tuner 5, which can be either active or passive. Multiple tuners on the source or on the load can be used for controlling the reflection coefficient at one or more frequencies. Further, bias tees can be used for providing biasing voltages to the DUT 10. Various additional instruments, e.g. spectrum analysers, vector signal analysers or power meters can be used to perform additional measurements. In an alternative embodiment of the present invention, the signal generator 1, the input coupler 2, output coupler 3, and the receiver 4 can be integrated into a single measurement instrument like a VNA. The number of channels in the receiver 4 can vary as the measurements of a₁, b₁, can be made in sequence. In such a case, a reference channel might be needed so that the waves a₁, b₁, a₂ and b₂ can be acquired in amplitude and/or phase.

The DUT 10 operational behaviour is described by a scattering matrix S relating an input incident wave a₁ and an input reflected wave b₁ at the input port of the DUT 10, and an output incident wave a₂ and an output reflected wave b₂ at the output port of the DUT 10. The method comprises connecting the input port of the DUT 10 to a signal generator 1 (see FIG. 2), and the output port to an output load impedance Z_L (in FIG. 2 embodiment shown as a tuner 5). The method embodiments further comprise subjecting the DUT 10 to an input test signal, and executing a first measurement a₁′,b₁′ of the incident wave a₁ and reflected wave b₁ at a respective DUT input reference plane 10a. The input test signal is generated by a signal generator 1 and can have a fixed input power signal (e.g. an input test signal referred to in the art as a large signal input test signal). The method further comprises subjecting the DUT 10 to a perturbation signal combined with the input test signal, and executing a second measurement a₁″, b₁″ of the incident wave a₁ and reflected wave b₁ at the respective DUT input reference plane 10a. The method further comprises determining the characteristic performance parameter Γ_S from the first measurement a₁′, b₁′ and the second measurement a₁″,b₁″, wherein the characteristic performance parameter is a source reflection coefficient Γ_S. Thus, the present invention embodiments allow to for e.g. to control the source impedance (or source pull) on a DUT at any given frequency. As shown in FIG. 1, the signal generator 1 is equivalent to a voltage source Es having a source impedance Z_S. The current at the input side is represented by and at the output side is represented by I₂. The voltage across the input terminals is represented by V₁ and between output terminals is represented by V₂. Γ_S is the source impedance provided at the DUT input reference plane 10a. The performance parameters of the DUT such as input and output power, efficiency, etc. can be measured from the measurement of a1″, b1″; a2″, b2″, as a function of the newly applied Γ_S.

In some embodiments of the present invention, the large signal input test signal is one of the group comprising: a continuous wave (CW) signal, a pulsed CW signal at a test frequency f1, a multi-tone signal, a modulated signal. The multi-tone signal can be comprised of multiple frequency components. Alternatively, the method described here can be applied at any or all the frequency components of the input signal. In a further embodiment of the present invention, the perturbation signal has a frequency corresponding to one or more frequency components of the large signal input test signal.

To control the source reflection coefficient to the DUT 10, the input test signal is perturbed during a test by adding a signal which can be controlled in amplitude and/or phase. In another embodiment of the present invention the perturbation signal comprises a perturbation of the amplitude of the (large signal) input test signal. In a further embodiment of the present invention, the perturbation signal comprises a perturbation of the phase of the (large signal) input test signal. By changing the amplitude and/or phase of the perturbation, the source reflection coefficient Γ_S can be controlled, while Γ_S can be measured at each perturbation by measuring the incident wave a₁ and the reflected wave b₁ with two subsequent measurements. The two subsequent measurements comprises the first measurement with the original test signal without applying any perturbation and the second measurement with applying the perturbation. Yet another embodiment of the present invention relates to subjecting the DUT 10 to a perturbation signal combined with the (large signal) input test signal that is obtained by controlling the signal generator 1.

In one embodiment of the present invention, the characteristic performance parameter is a source reflection coefficient Γ_S, and the DUT reference plane is at the input port of the DUT 10. The input and reflected waves a₁ and b₁ are linked to the wave b_S generated by the signal generator 1 and the source reflection coefficient Γ_S can be related by the equation:

a ₁ =bs+Γ _S ·b ₁

According to this embodiment, the following steps can be used to measure and control Γ_S:

A first reference measurement of a₁ and b₁ is performed at a fixed input power, providing the following equation:

α′₁=bs₀+Γ_S0·b′₁

where Γ_S0 is the (passive) source reflection coefficient given by the source at the DUT input reference plane. Γ_S0 is known a-priori either with a separate VNA measurement, or through a calibration measurement. The waves a₁′ and b₁′ are measured by the receiver 4. The parameter bs₀ is a source wave, which is not required to be known as demonstrated below.

In a second step, the source signal generated by signal generator 1 can be varied in a controlled, but arbitrary way, in amplitude and/or phase by adding an arbitrary wave b_s′ to the original source signal b_s0 so that bs₁=bs₀+bs′. This can be achieved by simply varying the power generated by the signal generator 1. Further, a new measurement of a₁ and b₁ is taken with the receiver 4 providing the following equations:

a″ ₁ =bs ₁+Γ_S0·b″₁

a″ ₁ =bs ₀ +bs′+Γ _S0 ·b″ ₁

The wave bs′ is effectively changing only the source impedance, instead of changing the source power. As a result, the following equation applies:

Γ_S1·b″₁=bs′+Γ_S0·b″₁

where Γ_S1 is the effective new source impedance provided to the DUT.

Thus, the method of determining the characteristic performance parameter from the first measurement and the second measurement comprises measuring and controlling an effective source reflection coefficient Γ_S1 in accordance with the formula:

${\Gamma }_{S1}=\frac{a_1^{″}-a_1^{\prime }+{\Gamma }_{S0}·b_{1^{\prime }}}{b_1^{″}}$

wherein Γ_S0 is a predetermined passive source reflection coefficient given by the source at the DUT input reference plane. To further control Γ_S, the wave bs′ can be modified and a new equivalent Γ_S1 can be calculated.

In another embodiment of the present invention, the first measurement is executed at a fixed input power of the (large signal) input test signal. In a further embodiment of the present invention, the second measurement is executed with a perturbation of the fixed input power of the (large signal) input test signal. This can be obtained by simply varying the power of the signal generator 1.

It is noted that the input test signal may be a large signal input test signal, but the method may also be seen as subjecting the DUT 10 to an input signal or an input wave b_s0.

The perturbation can be implemented as modifying the power of the input test signal or modifying the power of the input wave b_s0. The modification of power of the input test signal is equivalent to mathematically adding a new wave b_s′ to the original wave b_s0. By controlling the amplitude and phase of b_s′ it is possible to change the input power. In the method embodiments as disclosed above it is mathematically imposed that the change in input power does not result in a physical change of input power, but rather in a physical change of Γ_S.

FIG. 3A-C show graphical representations related to measurements executed according to an embodiment of the present invention. In this embodiment, the method has been applied to measure and control the source reflection coefficient Γ_S of a DUT 10, in this particular case an RF transistor. The setup used for the measurement and control of the source reflection coefficient is similar to the one depicted in FIG. 2, where a VNA 4 is used for measuring the a₁, b₁, a₂ and b₂ waves, and an arbitrary waveform generator 1 is used as signal source. By varying the power generated by the signal source (i.e. generator 1) and applying the method, a source pull sweep is performed and the transistor (DUT 10) performance is evaluated for several different source reflection coefficients Γ_S, FIG. 3A shows the user-defined source reflection coefficient that is applied to the DUT for three specific points on the Smith chart. FIG. 3B shows the source reflection coefficient Γ_S measured with the present invention method at the input of the DUT 10. FIG. 3C shows a measurement of gain vs. output power for all the three different source reflection coefficients Γ_S presented to the RF transistor as DUT 10. From this plot it can be seen how by changing the source reflection coefficient Γ_S the Gain of the active device (DUT 10) is affected. The method can therefore be used to find the optimum source impedance matching to maximize some specific performance parameters of the active devices (DUT 10) such as transducer gain or third order intermodulation distortion.

The present invention has been described above with reference to a number of exemplary embodiments as shown in the drawings. Modifications and alternative implementations of some parts or elements are possible, and are included in the scope of protection as defined in the appended claims.

Claims

1. A method for measuring and controlling a characteristic performance parameter of a device under test, DUT, having an input port, wherein DUT operational behaviour is described by a scattering matrix S relating an input incident wave a1 and an input reflected wave b1 at the input port of the DUT, the method comprising connecting the input port of the DUT to a signal generator,subjecting the DUT to an input test signal, and executing a first measurement of the incident wave and reflected wave at a DUT input reference plane,subjecting the DUT to a perturbation signal combined with the input test signal, andexecuting a second measurement of the incident wave and reflected wave at the DUT input reference plane, anddetermining the characteristic performance parameter from the first measurement and the second measurement, wherein the characteristic performance parameter is a source reflection coefficient ΓS.

2. The method according to claim 1, wherein the input test signal is one of the group comprising: a CW signal, a pulsed CW signal at a test frequency f1, a multi-tone signal, a modulated signal.

3. The method according to claim 1, wherein the perturbation signal has a frequency corresponding to one or more frequency components of the large signal input test signal.

4. The method according to claim 1, wherein the perturbation signal comprises a perturbation of the amplitude of the input test signal.

5. The method according to claim 1, wherein the perturbation signal comprises a perturbation of the phase of the input test signal.

6. The method according to claim 1, wherein subjecting the DUT to a perturbation signal combined with the input test signal is obtained by controlling the signal generator.

7. The method according to claim 1, wherein the DUT input reference plane is at the input port of the DUT, and wherein determining the characteristic performance parameter from the first measurement and the second measurement comprises measuring and controlling an effective source reflection coefficient ΓS1 in accordance with the formula:

8. The method according to claim 7, wherein the first measurement is executed at a fixed input power of the input test signal.

9. The method according to claim 7, wherein the second measurement is executed with a perturbation of the fixed input power of the input test signal.