Content uploaded by Hao Yan
Author content
All content in this area was uploaded by Hao Yan on May 06, 2018
Content may be subject to copyright.
Abstract—An integrated microwave detector at X-band
(8-12 GHz), based on a Micro-Electromechanical Systems
(MEMS) approach, is proposed in this letter. A six-port
MEMS coupler is used for coupling two parts of the input
signal from the main line to the two branches. The MEMS
coupler is designed to optimize frequency and phase
measurements, and this is achieved via several steps. First, the
primary signal travels from the main line for amplitude or
power detection. Then, a MEMS-based integrated detection
approach is used to apply vector composition and the Seebeck
effect. Next, the device is fabricated using a variation of the
GaAs monolithic microwave integrated circuit (MMIC)
process. Experimental results show that the MEMS detector
has return loss of greater than 10 dB and isolation loss of
more than 24 dB for the frequency band of 8-12 GHz. The
measuring device can detect frequency, phase, and amplitude
or power on a single chip.
Index Terms—Coupler, integrated detection, MIM
capacitor, Micro-electromechanical systems (MEMS).
I. INTRODUCTION
ODERN communication systems require low-power,
miniaturization, and higher scale integration with
various electronic components. Micro-Electromechanical
Systems (MEMS) techniques are expected to achieve these
goals [1]. Researchers have reported novel designs for
frequency sensors [2-4], phase detectors [5-7], microwave
power sensors [8-11], and amplitude demodulators [12-13].
The most common design approach used is based on
micromachining technology. MEMS devices using these
approaches could have simple structures, compact size, and
efficient microwave performance. However, single
modules cannot achieve measurement of the frequency,
phase, and amplitude or power on a single chip.
Thus, an integrated microwave detector based on MEMS
technology for X-band applications is proposed in this
letter. This integrated detector couples a part of the input
signal from the main line towards the two branches.
1This work was supported in part by the National Natural Science
Foundation of China under Grant NSFC 61674031 and NSFC 61076108,
in part by the Fundamental Research Funds for the Central Universities
under Grant KYLX16_0217, and in part by the Scientific Research
Foundation of Graduate School of Southeast University under Grant
YBJJ1614. (Corresponding author: Xiaoping Liao)
The authors are with Key Laboratory of MEMS of the Ministry of
Education, Southeast University, Nanjing 210096, and China (E-mail:
xpliao@seu.edu.cn and yanhao@seu.edu.cn).
Frequency and phase are detected via a single six-port
MEMS coupler. The input signal travels through the main
line, which is used for amplitude or power detection. A
MEMS-based RF receiver architecture (see Figure 1) is
incorporated into the structure as well. As Figure 1 shows,
the system is composed of an antenna, a band-pass filter
(BPF), a low noise amplifier (LNA), a local oscillator (LO),
a MEMS integrated detector, and a post-processing circuit.
This proposed MEMS detector not only detects multiple
parameters on a single chip, but can also serve as a passive
sensing device with zero-DC power consumption.
Fig. 1. The MEMS integrated detector is a part of a radio receiver for
coarse signal measurements.
II. DESIGN AND FABRICATION
Fig. 2. Six-port MEMS coupler: (a) Schematic overview and (b)
Equivalent circuit
An Integrated Microwave Detector based on
MEMS Technology for X-band Application
Hao Yan, Xiaoping Liao*,1 Member IEEE, ChenChen, and Chenglin Li
M
A schematic overview of the six-port MEMS coupler is
shown in Figure 2(a). The MEMS coupler consists of two
metal-insulator-metal (MIM) capacitors (Cs), coplanar
waveguides (CPWs), a coplanar waveguide with defected
ground structure (CPW-DGS), and air bridges. The
CPW-DGS and two MIM capacitors (Cs) constitute a
π-type network that achieves good impedance matching
and bandwidth expansion. Air bridges are applied to
suppress the coupled slot-line mode. Figure 2(b) shows that
equating the actual six-port MEMS structure to an
equivalent circuit is possible.
Our proposed technology can be used to detect amplitude
or power using integrated and networked components. First,
the primary signal travels from the main line via a
conducting circuit. Two pairs of capacitors (Cs/2) couple a
part of the RF signal from the main line towards the two
bifurcated paths for frequency and phase detection. Each
coupling capacitor (Cs) has a capacitance of 288 fF to
ensure return loss of greater than 10 dB at 8-12 GHz
frequency band.
In the branch paths, the phase and frequency of the
microwave signal is transformed into a power measurement.
The thermoelectric power sensors are applied to each port
to measure the output power. Therefore, this MEMS device
can realize the frequency (V3), amplitude (V4), power, and
phase detection (V5, and V6) on a single chip in the form of
the output thermal voltage.
Fig. 3. (a) Microscopic image of the MEMS integrated microwave
detector, (b) MEMS coupler, (c) air bridges and MIM capacitor, and (d)
thermopile.
Figure 3(a) shows a microscopic image of the MEMS
integrated detector. The separate insets show the MEMS
coupler, the air bridges, the MIM capacitor, and thermopile.
The integrated detector is fabricated via a GaAs MMIC
process. The structure’s schematic is shown in Figure 2(a),
and the corresponding structural parameters are also listed
(in μm). The thermopiles consist of an AuGeNi/Au and a
N+ GaAs layers. The AuGeNi/Au layer is used as the
positive electrode, while the N+ GaAs layer is used as the
negative electrode. The load resistor has a 50-ohm
resistance, which is made by depositing of 2-μm thick TaN
layer to the existing layers. The CPW-DGS is fabricated by
evaporating a 2.4-μm thick Au layer, and its final
dimensions (G×S×G) are 20 μm, 100 μm, and 20 μm,
respectively. The dielectric layer Si3N4 is deposited with
0.23-μm thickness via a plasma-enhanced chemical vapor
deposition (PECVD) process. The air bridges are formed
by sputtering a 2-μm thick Au layer that has a
56-μm×20-μm surface area.
III. MEASUREMENT AND DISCUSSIONS
A. Microwave Performance
To evaluate the input mismatch and isolation
characteristics of the MEMS detector, we measured the
X-band microwave performance of this device using a
vector network analyzer (VNA). As shown in Figure 3(a),
ports 1 and 2 are input ports for the test signal and reference
signal, respectively. In Figure 4, test results show the return
losses of port 1 and 2 (S11 and S22) are greater than 10 dB,
while the isolation loss (S21) between port 2 and port 1 is
larger than 24 dB at X-band. These results indicate that the
impedance of this device is well matched. Thus, our
MEMS detector has a relatively wide bandwidth and better
isolation in comparison with previous devices [2-7]. The
considerable isolation prevents signal crosstalk between
the test signal and reference signal.
(a) (b)
Fig. 4. Microwave performance of the MEMS detector: (a) Return losses
(S11 and S22), and (b) Insertion loss (S21).
B. Microwave Frequency, Amplitude and Phase Detection
An RF signal generator, a power splitter, and a tunable
analog phase shifter generate test and reference signals.
The phase shifter is mechanical and tuned by a knob. Each
phase-shift step is 9 degrees for a 10 GHz center frequency.
A DC voltage meter is applied to measure the voltage of
each output port, and an oscilloscope is used to observe the
demodulated signal.
To determine the capability of frequency detection (8-12
GHz), measurements are made at 20, 22, 23 dBm power
levels. In microwave networks, the relationship between
the phase and the frequency is linear. The principle of
frequency detection is based on vector composition, i.e.,
the law of cosines [2-4]. Therefore, the relationship
between the normalized output voltage and frequency can
be represented by a cosine function. Figure 5(a) shows a
plot of the normalized output voltage versus the frequency
of the test signal at 8-12 GHz frequency band [2-4]. Notice
that the normalized output voltage and frequency follow a
cosine curve trend and can accordingly be fitted to a cosine
function. By using an arccosine function-based
change-of-variables (Δφ=arccos(||V3||)), the relationship
between the phase and frequency is shown in Figure 5(b).
Notice that the phase versus frequency has a linear trend
and can accordingly be fitted to a line.
(a) (b)
Fig. 5. (a) The normalized output voltage versus the frequency (8-12 GHz),
and (b) The phase versus the frequency (8-12 GHz).
(a) (b)
(c) (d)
Fig. 6. Measured voltage versus the input power: (a) from 5 mW to 200
mW (7-23 dBm) @8 GHz, 10 GHz and 12 GHz, (b) from 0.1 mW to 5
mW (-10-7 dBm) @8 GHz, 10 GHz and 12 GHz; (c) Amplitude versus the
modulation wave frequency fm (0.001-10 kHz), and (d) Output waveform
frequency versus frequency fm (0.001-10 kHz), and relative errors.
Figures 6(a) and (b) show that the power (Port 4) can be
measured using an input power between 0.1 mW and 200
mW (between -10 dBm and 23 dBm) at 8-12 GHz. This
result suggests that there is good linearity between the input
power and output voltage, within the desired input power
range. The power detection performance can achieve
sensitivities of about 34.53, 27.4, 23.1 mV/W at 8, 10, and
12 GHz, respectively, with a resolution of 0.5 mW (-3
dBm). In practice, this is a relatively conservative estimate.
On the one hand, this value is accurate enough for coarse
phase and frequency detection. On the other hand, a
resolution of 0.5 mW is limited by the voltage accuracy of
the DC voltage meter. Therefore, adding a simple
amplification circuit can improve the measurement
performance [10]. Moreover, the optimization of power
sensors can also improve measurements [11].
Furthermore, the amplitude demodulation performance
is investigated for modulation frequencies (fm) between
0.001 kHz and 10 kHz, with a 20% modulation depth (ma),
at a 20 dBm input power. The relationship between
frequency and amplitude for the output signal is shown in
Figure 6(c). The resulting curve shows a low-pass filter
representation of the thermoelectric conversion [13-14].
The relative data errors are less than 1.5%, as shown in
Figure 6(d). Experiments also show that port 4 can achieve
relatively good power measurement and direct amplitude
demodulation performance.
(a) (b)
Fig. 7. (a) The normalized output voltage versus the phase shift, and (b) A
Lissajous Figure for phase calibration.
Figure 7(a) shows the normalized output voltage (for
Port 5 and Port 6) versus phase shifts measured at center
frequency of 10 GHz with an input power under 100 mW
(20 dBm). In Figure 7(b), a Lissajous pattern is formed by
using port 5 as the ordinate and port 6 as the abscissa.
According to the fitting analysis, the initial phase
difference between ports 5 and 6 is approximately 111.5
degrees. Results show that the normalized phase detection
validates the phase detection principle for the entire cycle.
IV. CONCLUSIONS
To summarize, this letter has introduced an integrated
microwave detector based on MEMS technology for
X-band applications. A six-port MEMS coupler is
integrated into this device to couple parts of the signal from
the main line towards the two branches, for frequency and
phase detection. The input signal travels through the central
line for amplitude or power detection. Experimental results
show that the proposed design can realize frequency, phase,
and power or amplitude detection on a single chip. This
MEMS device has zero DC power consumption and can be
used to detect coarse measurements of received signals
from low-power microwave and wireless systems.
REFERENCES
[1] S. Lucyszyn, Advanced RF MEMS, London, U.K.: Cambridge
University Press, Cambridge, Aug. 2010.
[2] Z.X. Yi, X.P. Liao, “An 8–12 GHz microwave frequency detector
based on MEMS power sensors” J. Micromech. Microeng., vol. 22,
no. 3, p 035005, 2012, DOI:10.1109/ICSENS.2011.6127229.
[3] Z.X. Yi, X.P. Liao, “Packaging test-fixture research for 10–12 GHz
frequency detector based on thermoelectric microwave power
sensor” Electronics letters. vol. 48, no. 2, pp. 103-104, 2012,
DOI10.1049/el.2011.3094.
[4] Z.X. Yi, X.P. Liao, “Design of the microwave frequency sensor for
power-unknown signal based on MEMS technology", In
Proceedings. Int. IEEE 29th MEMS Conference. 2016.
[5] D. Hua, X.P. Liao, and Y. Jiao, “X-band microwave phase detector
manufactured using GaAs micromachining technologies” J.
Micromech. Microeng., vol. 21, no. 3, p. 035019, 2011, DOI:
10.1088/0960-1317/21/3/035019.
[6] J.Z. Han, and X.P. Liao, “A Compact Broadband Microwave Phase
Detector Based on MEMS Technology” IEEE Sensors J. vol.16, no.
10, pp. 3480-3481, 2016, DOI: 10.1109/JSEN.2016.2540636.
[7] H. Yan, X.P. Liao, and D. Hua, “A Four-Port Microwave Phase
Detector at X-Band Based on MEMS Power Sensors” IEEE
Sensors J., vol. 17, no. 7, pp. 2029-2035, 2017, DOI
10.1109/JSEN.2017.2659802.
[8] V. Milanovic, M. Gaitan, E.D. Bowen, N. H. Tea, and M. E.
Zaghloul, “Thermoelectric power sensor for microwave
applications by commercial CMOS fabrication,” IEEE Electron
Device Lett., vol.18, pp.450-452, 1997. DOI: 10.1109/55.622527
[9] D.B. Wang, X.P. Liao and T. Liu, “A novel thermoelectric and
capacitive power sensor with improved dynamic range based on
GaAs MMIC technology”, IEEE Electron Device Lett., vol. 33, pp.
269-271, 2012. DOI: 10.1109/LED.2011.2174610
[10] D.B. Wang, X.P. Liao. “A terminating-type MEMS microwave
power sensor and its amplification system” J. Micromech.
Microeng., vol. 20, no. 7, p. 075021, 2010. DOI:
10.1088/0960-1317/20/7/075021
[11] D.B. Wang, X.P. Liao and T. Liu, “Optimization of
indirectly-heated type microwave power sensors based on GaAs
micromachining” IEEE Sensors J. vol. 12, no. 5, pp. 1349-1355,
2012. DOI: 10.1109/JSEN.2011.2170677
[12] R. B. Reichenbach, M. Zalalutdinov, J. M. Parpia, and H. G.
Craighead, “RF MEMS oscillator with integrated resistive
transduction,” IEEE Electron Device Lett., vol. 27, no. 10, pp.
805-807, 2006. DOI: 10.1109/LED.2006.882526
[13] H. Yan, X.P. Liao, and C.L. Li, “A Dual-Channel MEMS
Amplitude Demodulator for On-Line Detection in Radio Relay
Station,” IEEE Electron Device Lett., vol.38, no.8, pp. 1121-1124,
2017, DOI: 10.1109/LED.2017.2718005
[14] A. Graf, M. Arndt, M. Sauer, and G. Gerlach, “Review of
micromachined thermopiles for infrared detection,” Measurement
Science and Technology, vol. 18, no. 7, R59–R75, 2007.
DOI:10.1088/0957-0233/18/7/R01