Abstract: This paper introduces a method for implementing an ultra-wideband MIR motion sensor , and describes in detail the implementation of the low-power ultra-wideband pulse generation circuit and the ultra-wideband correlation detection circuit. The sensor was used to conduct detection sensitivity and penetration tests on targets of different materials, achieving the expected results.
Keywords: Ultra-wideband miniature power pulse radar, fast pulse sensor, ultra-wideband (UWB)
Ultra-wideband (UWB) is defined as a waveform with a bandwidth that is a high proportion of its center frequency. In other words, any waveform with a bandwidth greater than 25% of its center frequency can be considered UWB. UWB uses baseband pulses with very narrow pulse widths, typically on the order of nanoseconds. In the frequency domain, its energy is sparsely distributed throughout the entire bandwidth used. UWB technology has wide applications in many military fields, such as UWB covert communication, UWB sensors, through-wall surveillance, non-metallic mine detection, and guidance systems that penetrate vegetation.
Motion sensors based on micropower pulse radar can be used as burglar alarms, concealed in the ceiling of a building. They can detect human movement within a designated area by penetrating the ceiling. The low-power functionality and low cost of micropower pulse radar necessitate that its specific circuit design have unique characteristics.
1. Implementation principle of MIR motion sensor
The MIR motion sensor consists of three parts: a transmitting unit, a receiving and display unit, and an antenna, as shown in Figure 1. The transmitting unit generates pulses from pulse oscillation (1.8MHz, 50% duty cycle), shapes them into narrow pulses (tens of nanoseconds in pulse width), and finally shapes them into fast pulses with a pulse width of about 400ps using a step diode, which are then transmitted by the dipole ultra-wideband antenna. The receiving unit uses phase sampling to extract the ultra-wideband signal reflected from the target, and consists of sampling delay, sampling gate signal generation, integration sampling, wideband signal amplification, bandpass filtering, and a display unit.
Sampling delay involves delaying the narrow pulse signal transmitted from the antenna, shaping it into a fast pulse, and then using it as a reference signal for the relevant receiver. The pulse width of the sampling gate determines the spatial resolution of the sensor, and can be selected from 100ps to 400ps depending on specific needs. Integrating sampling integrates the weak, ultra-wideband signal received by the antenna and reflected by the target, thereby improving the signal-to-noise ratio. For a transmitted pulse with a repetition frequency of 1MHz, if the integration time constant is 1ms, 10,000 pulses will be averaged in the integrator. The averaged received signal is amplified 1000 times and then passed through a bandpass filter to detect the signal. For human motion detection, the upper and lower cutoff frequencies can be set to 0.1Hz and 10Hz, respectively.
The circuit is shown in Figure 2. The most distinctive features of Figure 2 are the "inductor-driven fast pulse shaping" and "correlation detection circuit" at the transmitter. In the printed circuit board design of this circuit, distributed parameters must be considered, and microstrip operation must be employed.
1.1 Inductively Excited Fast Pulse Shaping
A step recovery diode (PDD) is a special type of varactor diode, also known as a step diode. Under sudden voltage excitation, it can generate extremely short-duration fast pulses, typically used for frequency multiplication of ten times or more. D1 and D2 are step diodes, and L1 and L2 are microstrip inductors (nH level). This circuit can excite a narrow-pulse square wave as input and shape it into a picosecond-level fast pulse. One path is used for transmitting the pulse, and the other path, after delaying the fast shaping, is used for the sampling gate signal. In the specific parameter selection of this circuit, pulse width and pulse amplitude are contradictory; that is, to pursue a narrower pulse width, the output pulse amplitude must be sacrificed, reducing the average output power.
After multiple circuit experiments, a pulse source with a rise time of less than 350ps was obtained (see Figure 3), and a pulse width of less than 800ps was obtained. In the selection of actual motion sensor parameters, the main considerations are matching the signal spectrum to the antenna bandwidth (500MHz~1.4GHz), while also considering the largest possible output amplitude.
1.2 Correlation Detection Circuit for Low-Power Pulse Radar Signals
The ability to detect the weak reflected signal from the object under test at microwatt-level transmit power is crucial for the receiving system. The advantage of correlation detection is that it accumulates the received signal relative to a reference signal, generating an output signal higher than the noise level.
In Figure 2, when there is no antenna signal, the voltage level at point A is the bias reference level, ranging from 50mV to 150mV. When the receiving antenna receives the reflected signal from the object, the signal charges on the reference voltage of C1. At this time, the negative signal of the sampling gate turns on D through Cp. The reflected signal from the object is charged and discharged multiple times according to the repetition period of the signal and the pulse width of the sampling pulse signal. It is averaged at point A and can effectively suppress external noise. This voltage is output through operational amplifier A1, and its waveform time relationship is shown in Figure 4.
2. Experiment on motion sensor detection using low-power pulse radar
Two main tests were conducted on the completed motion sensor system: (1) detection and resolution tests of targets made of different materials; and (2) penetration tests. Experiment (1) used a variety of materials (metal, wood, human body, etc.). The experiment showed that its detection sensitivity is proportional to the dielectric constant ε of the material and proportional to the volume of the target.
The penetration test aims to detect the penetration capability of the sensor system, which is also the most significant feature of ultra-wideband detection. Using a TEK602 storage oscilloscope connected to the output of an MIR receiver, the sensor was able to clearly detect the waveform acquired by the sensor when a human body moves through the detection area in a corridor, penetrating a 30cm brick wall in the laboratory.
Figure 5 shows a person walking down a corridor. The MIR antenna is attached to the wall in the laboratory, facing the corridor. Figure 5 shows two peaks as the person enters and exits the detection area.
