Experimental Study of High Power Microwave Injection in Radar Receiving Front-end

In order to study the tolerance threshold and damage process of radar receiving front end against high power microwave pulse attack, the relationship between the damage of receiving front end and injected signal power, pulse width, and duration was obtained by injection test. The experimental results show that the damage of PIN limiter by high power microwave is a slow worsening process near the damage threshold. Based on the test results, the radar withstanding range of high-power microwave weapon attack is calculated theoretically, which can provide theoretical support for the radar receiving front-end design.


Introduction
High-power microwave wave weapon (HPMW) mainly refers to a new type of weapon that relies on the generated high-power microwave to suppress interference and damage the target.A high-power microwave is generally considered to be a strong electromagnetic pulse with a peak pulse power greater than 100 MW and a frequency range of 300 MHz to 300 GHz [1] .High-power microwave weapons have the characteristics of fast attack speed, high cost-effectiveness ratio, wide damage range, difficult protection, and diverse effects [2] .
Air surveillance radar can detect incoming air targets and is the preferred target for high-power microwave weapons in warfare.High-power microwaves can cause interference or damage to the radar receiving system through front door coupling, which is the main way of radar damage by high-power microwaves.
With the development of radar solid-state, most existing radars adopt PIN limiters for protection [3] .In the research on existing PIN limiters, the tolerance threshold and nonlinear performance of PIN limiter [4] are mainly studied from the aspects of PIN diode modeling and heating simulation [5][6][7] , PIN diode DC voltage injection [8] and PIN limiter RF injection [9,10] , etc., but few studies have been conducted on the radar receiving front end in combination with limiter and low noise amplifier.In this paper, the RF injection test is conducted on the radar receiving front-end system, the tolerance threshold is measured, the relationship between the damage of the receiving front-end and the injected signal power, pulse width, duration, and other factors is studied, and the influence of high-power microwave weapons on radar at different distances is analyzed, which can provide theoretical support for the RF front-end protection design of radar.

2.
Injection test design

Receiving front-end circuit analysis
The radar-receiving front end consists of a PIN limiter and a low noise amplifier (LNA).The PIN limiter limits the received signal and amplifies it by the LNA before sending it to the receiver.
As the first high-frequency amplifier of the receiving system, the low noise amplifier is mostly composed of field-effect amplification transistors to reduce the impact of noise on the receiving system, but its withstand power is small, and it is easy to be burned by strong echoes or the influence of the radar equipment in the same frequency band of other radars The role of the PIN limiter is to protect the LNA.When the input power is greater than the initial power limit, the input signal is deflated and reflected to prevent the later LNA from burning.The PIN limiter is mainly composed of a PIN diode, and the P region, I region, and N region of the PIN diode are respectively composed of P-type semiconductors, high resistivity intrinsic semiconductors, and Ntype semiconductors.A typical circuit for a PIN limiter is shown in Figure 1.When the input power is small, the PIN diode has a high resistance to the ground, and only weak energy flows through the PIN diode.When the incident signal gradually increases, the impedance of PIN to the ground will be lower and lower, close to the ground conduction, and part of the signal is reflected back due to the input impedance mismatch, at this time PIN plays a limiting role in high-power signal output.When the power is too large, resulting in part of the incident signal energy flowing through the PIN to the ground and heating, it may cause damage to the PIN limiter.It is assumed that the impedance of the front and rear stages of the limiter is 50 ohms matching load.When the limiter is connected to the circuit, the reflection coefficient (S11), insertion loss (S21), and absorption coefficient after the PIN diode connected to the circuit can be calculated because the limiter shows different ground resistance values affected by the input power, and the PIN diode is simplified into a variable resistance to the ground.Figure 2 shows the normalized reflected power, output power, and absorbed power of the theoretical PIN diode with different resistance values.
Resistance Of PIN diode \ Ω Normalized power ratio/ %

Figure 2 S-parameter of PIN limiter with different resistance values
Figure 2 can only briefly represent the input power distribution under different power sizes.In the real circuit, detailed calculations need to be made according to the thickness of the I layer and the change of the capacitance and reactance of the PIN diode.However, when analyzing the PIN limiter tolerance power threshold, the PIN diode tolerance threshold should not be considered only but should be considered in the circuit to consider its overall performance, such as considering peak leakage, tolerance threshold, etc.

Basic properties of receiving specimens
The tested PIN limiter and LNA operate in the L-band.The PIN limiter is a combination of multi-stage PIN diodes, and the LNA uses two-stage power amplification, and its basic performance is shown in Table 1.

Injection test design
For broadband signals, due to the filtering effect of the antenna, only a small part of each pulse enters the receiving front end, and because the pulse width is narrow, the energy of the monopulse signal is weak, unless the signal power is very large (100 GW level), it is difficult to form effective damage, which will not be discussed here.For the narrow-band signal in the band, the antenna feeder system can pass smoothly without the deformation of the signal, and the rectangular pulse signal in the working band of the LNA and PIN limiter can be directly used as the injection signal, which can simulate the damaging effect of the front-door coupling on the receiving front end.According to the CHAMP missile using a magnetron transmitter [11] , the transmitted signal is an L-band narrow pulse signal, and the injected signal is a narrowband pulse signal.This test is divided into two parts: the nonlinear effect test and the damage effect test.The nonlinear effect test is mainly to study the nonlinear change of the peak leakage, isolation degree, and LNA gain of the PIN limiter under the condition that the damage is not achieved, so as to test the nonlinear suppression effect of high-power microwave on the radar front door coupling.In the damage test, the power, pulse width, and pulse number are gradually increased until the device is damaged, so as to obtain the effect phenomenon and damage threshold of the component to be measured.The test data were injected into multiple devices, the average values were calculated and the representative data were analyzed.
For the high-power damage injection test, a test platform as shown in Figure 3 is established, which can simultaneously observe the incident, reflected, and output power of the part under test through an oscilloscope.The high-speed oscilloscope can also observe the change of the intra-pulse signal, especially the power and duration of the peak leakage signal after the high-power microwave enters the PIN limiter.Test process and result analysis

3.1PIN injection test and result analysis
Since the limiter has a large tolerance power and a high threshold level, the PIN limiter can be injected directly by using the platform shown in Figure 3.The injection signal pulse repetition frequency is 100 Hz to prevent energy accumulation between pulses [12] .
First, the threshold level of the limiter was tested.When the input power reached about 10 dBm, the insertion loss became larger, and the peak leakage phenomenon was observed at the injection output signal.The threshold level of the limiter is about 10 dBm, and the statistical results are shown in Table 2.
Table 2 Starting Limit Level Statistics Second, the peak leakage effect of the PIN limiter was tested. 1 us pulse width signal was used for injection to prevent the PIN limiter from burning due to excessive power.When the PIN limiter starts to limit, a certain response time is required, and there is a spike leak of the signal, and then the output signal is limited.When the injection power is 11 dBm, the output signal is shown in Figure 4.
Figure 4 Output signal waveform after limiter limiter Taking a limiter as an example, the relationship among its injection power, peak leakage power, and flat leakage power is shown in Figure 5.As can be seen from Figure 5(a), the peak leakage power is positively correlated with the injection power, and the relationship is basically linear.Combined with the injection power, the leakage energy is very small, and the calculated peak leakage energy under different injection power is only about 1 nJ, which is difficult to produce a damaging effect on the post-LNA.The relationship between injection power and flat leakage is shown in Figure 5(b), and the maximum flat leakage power of the PIN limiter is only about 20 dBm.
The power-bearing capacity of the PIN limiter was tested with a wide pulse.At this time, the heat balance was formed inside the PIN limiter, and the injection power was slowly increased so that the maximum internal temperature reached the melting point.In the test, a pulse with a pulse width of 50 us was used for injection, and 1 dB was used as the step to adjust the injection signal pulse power.Each injection time was 30 seconds.After each injection, the device was cooled for 1 minute until the insertion loss or reflection coefficient of the limiter increased by more than 1 dB than the original parameter, which is the power capacity of the PIN limiter.
After testing, the PIN limiter works normally when the injection power was less than 53 dBm.When the injection power reached 53 dBm, the reflected signal suddenly increased, the reflected power of the directional coupler reached more than 48 dBm, and the output signal became smaller.After removing the PIN from the platform and standing for 1 minute, the S-parameter of the limiter was measured, and the insertion loss reached 10.68 dB, which proved that the PIN limiter has been damaged.The damage threshold in this configuration is 53 dBm.
On this basis, the damage effects of different pulse width lower limits were measured.The pulse width was gradually increased from 2 us, and the injection power was gradually increased from 53 dBm to change the pulse energy.Each injection time was 10 s.After injection, the insertion loss and reflection coefficient of the limiter were measured.If the insertion loss increases, it is judged that the injection signal has damaged the limiter, and it is recorded as the tolerance threshold of the current pulse width lower limit.The limiter was replaced, the injection power was adjusted, and the pulse width was adjusted from the beginning of the injection.Through curve fitting of pulse width and limiter tolerance threshold results, the expressions of pulse width τ and damage power threshold P can be obtained, as shown in Equation ( 1), and the fitting results are shown in Figure 6.The limiter was replaced for a continuous damage test, and the changes in insertion damage under different duration and pulse widths were counted.For example, when the injection power is 55 dBm, the results are shown in Table 3.The cumulative energy injected at 55 dBm pulse power and the damaging effect on the insertion loss of the limiter was calculated, as shown in Figure 7.
Cumulative injected energy / J Cumulative damage effect /dB

Figure 7 Relationship between insertion loss change and energy accumulation
Since each injection is cooled, it is not the accumulation of damage caused by the heat accumulation effect, but the accumulation of the damage effects each time.That is, each injection pulse causes certain damage to the device, the damage is irreversible, and the damaging effect is expanded in the next pulse, resulting in further insertion loss.The fitting expression of cumulative injection energy and insertion loss is shown in Equation (2).
where E is the accumulated injected energy.It can be seen that with the increase of the damaging effect, the effect caused by the same energy on the limiter will decrease.Because when a certain damage effect is caused, the resistance value of the limiter will decrease, the reflection coefficient will increase, resulting in a decrease in absorption power, and greater energy will be required to continue the damage.Taking limiter No. 3 as an example, when the final insertion loss of the limiter reaches 8.036 dB, the measured reflection coefficient has reached 1.937 dB, and the reflected energy is already large.It was found that the first stage had been short-circuited and the resistance was only 2.4 Ω.

3.2LNA injection test and result analysis
Because of the small input power of LNA in normal operation, a good nonlinear effect can be obtained by using a vector net for power scanning.The output power of the vector network used in the experiment ranges from -25 dBm to 10 dBm, and the linearity is good near 0 dBm.The 20 dB attenuator connected to the port of the vector network 1 can ensure good measurement accuracy in the power range of concern.The LNA gain is about 30 dB, and the 40 dB attenuator is connected to the vector network port 2, which can ensure the linearity of the vector network measurement port and prevent the LNA output power from burning the vector network.
Taking a test piece as an example, the power-gain test results are shown in Figure 8.

Input power / dBm
Gain / dB As can be seen from Figure 8, the normal gain of LNA for small signals is about 35.6 dB.When the power is increased to -22.62 dBm, its gain decreases by 1 dB to 34.6 dB, indicating that LNA has entered the nonlinear amplification interval.If the input power continues to increase, LNA will lose its amplification ability due to output limiting.Generally, the LNA gain is considered invalid when it drops by 10 dB.
High power microwave injection damage test was carried out on LNA by using the platform shown in Figure 3.The injection power was gradually increased, when the injection power reached 26~27 dBm (about 446 mW), the LNA operating current suddenly decreased, and the peak power of the output signal decreased, which indicated that the LNA had been damaged.The tolerance power threshold is about 26.5 dBm.The damage suppression of LNA is less than 1 dB, which indicates that the tolerance threshold of the LNA amplifier is consistent.The damage threshold is shown in Table 4.In the table, f0 indicates the LNA center operating frequency.The damaged LNA was scanned by Sparameter with vector net, as shown in Figure 9.In Figure 9, B is the working bandwidth of the LNA.As can be seen from the figure, the reflection coefficient of LNA is large, and the forward gain basically does not exist.Although there is an output signal, the output is unstable, and the output is judged to be noise.

Range / km
Coupled power/ dBm As can be seen from Figure 10, assuming that the pulse width of CHAMP is 5 us, the tolerated power of the radar receiving front-end is 58 dBm, that is, CHAMP can only cause nonlinear compression of PIN limiters and LNA at a distance of 32.66 km, but cannot produce damage effect on the radar.In this case, only false target interference and suppression interference are generated in the receiving channel.When the distance is less than 32.66 km, a certain damage effect can be produced, but it needs to be continuously irradiated for a few seconds before the radar can be completely damaged.After damage, only the limiter of the receiving front will be damaged, while the rear LNA will not be damaged, but the detection distance will be reduced and the signal will be weakened.If the distance is closer, such as less than 25 km, it can directly and quickly cause large damage, resulting in larger limiter insertion loss, weaker signal, or direct disappearance.

Conclusion
In this paper, through the injection test of a radar receiving front end, the damaging effect of HPM on the radar front end under multi-pulse is studied, and the following conclusions are drawn: 1) The PIN limiter of the radar has a good protective effect to protect the LNA from being destroyed.
2) The destruction effect of high-power microwave on PIN limiter is the comprehensive effect result of cumulative peak power, pulse width, and pulse frequency.When the power capacity of the PIN limiter is reached, the damaging effect of the high-power microwave on the limiter can be increased by these three factors.
3) When the pulse power exceeds the power capacity of the PIN limiter, the multi-pulse high-power microwave damage to the PIN limiter is a gradual process, and the final performance is insertion loss and reflection coefficient increase.
4) In the process of destroying the PIN limiter with multiple consecutive pulses of the same energy, the reflection coefficient of the limiter increases during the damage process, resulting in a decrease in the incoming energy and a gradual decline in the damage efficiency of the single pulse.
This study can provide a basis for the evaluation of radar front door coupling protection, and provide theoretical support for radar front door protection.In the next part, more types of receiving front ends need to be studied, and suggestions for improvement are proposed for different radars.

Figure 1
Figure 1 Typical PIN limiter circuit

Figure 3
Figure 3 Injection test platform

Figure 5
Figure 5 Relationship between injection power, peak leakage power, and peak leakage time

Figure 6
Figure 6 Fitting curve of pulse width and limiter tolerance threshold

Figure 8
Figure 8 LNA power-gain scanning curve

Figure 10
Figure 10 Relation between radar received power P r and range R

Table 3
Damage Effect of Continuous Injection of Limiter

Table 4 LNA
Damage Threshold Device No. Damage Threshold Gain After Damage at f0