High Power Surface Mount Switch-Limiter Module
By Chandu Sirimalla, Aeroflex / Metelics, Inc.
This article describes the implementation of a limiter integrated switch module. The switch limiter module is designed for S- band, 2 to 4 GHz with a maximum incident CW power handling of 51 dBm in the antenna transmit path and 33 dBm CW power handling in the antenna receive path, the integrated limiter achieves flat leakage power of less than 15 dBm and spike leakage energy of less than 0.2 erg. It is a self-contained system with built-in bias tees and bias current limiting resistor on a compact 8 mm x 5 mm x 2.5 mm aluminum nitride (AlN) substrate for greater power dissipation and thermal stability. The inbuilt limiter suffice for the direct interface with the low noise amplifier (LNA) in the receive path. The module eliminates the need for external limiter circuits.
As radar system designers pack ever increasing numbers of radiating elements into ever smaller phased array antennae, the transceiver circuits connected to each of the antenna elements are required to be further functionally integrated. A typical phased array radar antenna has several radiating elements which are combined coherently to form a directive beam in the desired direction with minimal side lobes.
A high level of integration between these multiple antenna elements is required to enhance radiation and shape of the beam. The Aeroflex switch limiter module is a step in the direction of system integration at T/R module level, where a switch and limiter circuits are integrated onto a compact 8 mm x 5 mm x 2.5 mm aluminum nitride [AlN] substrate module. The module does not need external bias tees or current limiting resistors.
By digitally controlling the dedicated RF power source, receiver and phase shifter at each of the antenna elements, rapid beam steering, multiple independent beam generation, and optimal radar time usage and radar power management can be achieved. This kind of control provides ultimate flexibility for any radar and communication system.
PIN Diode – RF Control Element
A PIN diode is similar to a standard diode with an extra layer between its P and N layers, in which charge is stored when the diode is forward biased. The characteristics of the diode and the amount of charge stored in the I layer determine the diode’s impedance to RF and microwave signals. PIN diodes are efficient minority carrier devices that use DC current and voltage to control the resistance and capacitance, respectively, of the intrinsic I-region. This degree of freedom in the impedance control makes them essential building blocks in a wide variety of RF and microwave control applications such as switches, limiters, attenuators and phase shifters, etc.
PIN diode’s excellent adaptability to RF & microwave signals with good thermal properties, reliable power handling per unit silicon volume, and cost effectiveness make them fundamentally necessary in the RF & microwave applications.
PIN Diode Parameters
PIN diode parameters that are directly related to the RF & microwave circuit performance are (i) diode series resistance (Rs) in forward bias, (ii) diode parallel resistance or diode reverse bias resistance, (iii) junction capacitance (Cj), (iv) I-layer thickness, (v) the wire bond inductance (L). These parameters determine the circuit insertion loss, isolation, intermodulation products and operating frequency capability.
The minority carrier lifetime of the diode (τ) (which is related to the I-layer thickness) determines switching speed and operating frequency. The thermal resistance of the diode (θJC) determines RF & microwave operating power and power dissipation.
Series-Only Symmetrical Switch
Figure 2 shows the realization of a symmetrical single pole two throw switch using PIN diodes with external DC control.
When appropriate biasing conditions are applied to forward bias the diode D1 and simultaneously reverse biasing the diode D2 symmetrical switch, functionality can be achieved. The switch is ON (Low RF Attenuation) in the J0-J1 path when diode D1 is forward biased and diode D2 is reverse biased, and OFF (high RF impedance) in J0-J2 path when the bias states of the diodes are reversed. The ON and OFF paths can be transposed by applying complementary biasing conditions. The low RF impedance is also called a low loss or insertion loss state and the high impedance state is called the isolation state.
In this series-only switch configuration, the diode series resistance at a given DC current and junction capacitance determines the insertion loss and isolation specifications, respectively.
Series-only switch configurations have low insertion loss over a wide frequency range and generally perform well over multiple octaves. However such switches are limited in the power handling capability. Diode junction thermal management determines the power handling.
For the switch module to properly interface with the DC power source it will also need bias tees, which are low pass filter structures which connect DC bias sources to the signal path but block the RF & microwave signals from entering the bias sources.
Shunt-Only Symmetrical Switch
The shunt-only symmetrical switch has higher isolation and low loss across a wide frequency range. Shunt switch modules can handle higher power levels as the diodes are mounted directly to the module housing, which produces lower thermal resistance between the diodes and the system heat sink.
Unlike the series switch module, in the shunt-only switch module a given RF path is ON (high shunt RF impedance) when the diode is reverse biased and the RF path is OFF (low shunt RF impedance) when the diode is forward biased.
The ON path insertion loss is dependent on the junction capacitance (Cj) and the OFF path isolation is dependent on the diode series resistance (Rs).
Series Shunt Switch
As summarized in the topology based RF parameter comparison table, the series switch can offer only a low level of isolation and insertion loss is very dependent on the diode Rs.
There are drawbacks in just series-only and shunt-only switch modules. A combination of series and shunt topology will provide optimum performance for various RF specifications over a wide frequency band like high switching speed, low insertion loss, high isolation and moderate power handling.
These two parameters, isolation and insertion loss, can be greatly improved by adding a shunt diode (D3) after the series diode (D2) to form a series-shunt switch module (Figure 6). The addition of diode D3 makes the series-only switch module into a TR switch (Transmit Receive switch).
The primary advantages of this topology are low insertion loss in the antenna–transmit path and high isolation in the antenna-receive path. The additional shunt diode in the receive path provides additional isolation which is dependent on the diode Rs and can be controlled by the external DC source. This additional isolation is very helpful in protecting sensitive LNAs in the receive path. The additional shunt diode D3 in the path Rx – B1 will increase isolation typically by 15 dB or better, but true isolation value depends on the diode parameters of diode D3 and the DC conditions applied.
More complex configuration switch modules can be designed as per the desired requirements.
The two diodes in parallel in series path will provide the low insertion for the ON arm and the two diodes in parallel in shunt configuration will provide higher order of isolation. The design can be tweaked to design a multi-throw switch module.
PIN Diode – A Limiter
Limiters are necessary to shield sensitive front end semiconductor components such as LNA and mixer components from ubiquitous high-power RF signals, especially from co-located transmitters which may deliver large signals in the order of kilowatts or megawatts. A receiver must reliably detect and process weak incoming signals. With a sensitive low noise amplifier present at the input of the receiver even a fraction of a high power transmit signal will likely damage these sensitive front end components. A quasi-active limiter is self-actuated and performs receiver protection virtually instantaneously, does not require any external driver and is much less likely to fail than the receiver components. This wide-band self-biased limiter protects receivers from large signals from adjacent radars, other high power transmitters and reflected signals with no foreknowledge of time of arrival or incident power of the large RF signal.
Single Stage Limiter
For RF power limiting, PIN diodes are normally used in shunt configuration. This is the chosen method as it provides excellent isolation and power handling capability, since the shunt configuration provides a direct conduction path for heat extraction from the limiter diode junction during large input signal events.
In the absence of a large RF incident signal the impedance of the limiter diode is at its maximum, providing a typical insertion of 0.5 dB. When a large RF signal is incident on the limiter diode the impedance changes to a low value and creates an impedance mismatch which reflects the incident signal back to the signal source. Figure 11 shows a single stage limiter circuit. Its performance specifications directly depend on the PIN diode electrical parameters- diode series resistance (Rs) and the junction capacitance (Cj).
The isolation of large incident RF power by the limiter diode can be optimized by employing a cascade of multiple limiter diodes, separated from each other by one quarter wave length (λ/4). An RF choke inductor acts as a DC return for the current generated in the limiter diode under large incident signal conditions.
The design employs a cascade of limiter diodes with the thicker I-layer limiter diodes (also known as coarse limiter diodes) at the input end of the circuit and the thinner I-layer limiter diodes (also known as cleanup limiter diodes) placed at the output end. The number of stages and the primary criterion for limiter diode selection (I-layer region thickness) depend on the required isolation, incident power handling capability, and threshold level or the flat leakage level requirements of the receiver circuit to be protected. The placement of these stages is also a key factor in the design. The limiter stages are spaced by one quarter wave length (λ/4) for the given frequency of interest.
The first stage of the two stage limiter design is the coarse limiter stage, implemented using a pair of limiter diodes (doublet die comprising a common cathode pair of diodes) in shunt configuration; the second stage (clean-up stage) is spaced by a quarter wavelength from the coarse limiter stage. The second stage comprises a cleanup diode (thinner I-layer limiter diode) and a RF choke inductor which acts as a DC return for the cleanup diode. DC blocking capacitors are placed at the limiter’s input, output and between the coarse and cleanup stages. The cleanup stage determines the threshold level and the spike leakage of the entire limiter module.
Working of Two Stage Limiter
When a large signal is incident upon the input of the two stage limiter, both the stages diodes are in their high impedance state, the entire large signal amplitude traverse the limiter module.
As the cleanup stage diode has thinner I-layer and has shorter carrier transit time, the impedance of the cleanup diode changes first. This change establishes a standing wave on the transmission line with a voltage minimum at the clean-up stage. As the coarse limiter stage is spaced by λ/4, a voltage maximum occurs at the across the coarse stage diodes. This voltage maximum forces the coarse limiter stage into conduction by reducing its impedance and ultimately provides the overall limiting.
RF Control Element – Applications
The PIN diode control component usage is frequency and application dependent.
A fundamental building block of a radar system is the transmit/ receive (T/R) switch module. The T/R module allows a single antenna element to be alternately driven by a high power amplifier in transmit mode and drive a sensitive LNA in receive mode.
Phased Array Radar – Switch Limiter Module Application
As radar systems are designed with increasing numbers of radiating elements into smaller form factor antenna arrays, the transceiver circuits connected to individual elements must also be functionally integrated.
Radar systems realized with high integration like the active phased array antenna offer several advantages like electronic steering, rapid repositioning, high gain in beam directivity, and multiple independently-steered antenna beams from a common aperture.
Figure 10 shows the T/R module configuration with a power amplifier on the transmit and an LNA on the receiver.
This kind of distributive architecture of the active aperture can smooth the effect of pulse-to-pulse amplitude and phase variations introduced in the power source and therefore increase the MTI improvement and further obtain better detection of moving targets in clutter.
Switch Limiter Module
The SP2T switch-limiter module incorporates a high-power single-pole two-throw (SP2T) switch with a passive receiver-protector limiter and a fully integrated DC bias network. It is designed to be used in high power T/R switch applications in S-band (2 GHz to 4 GHz). The SP2T switch-limiter module is located in the front end of the receive path as shown in the Figure 10.
Three complementary control signals are required for proper operation. Bias voltages are applied to the Tx (B1) bias port, Rx bias port, and the Bias (B2) bias port to control each state of the switch-limiter module.
In the Tx state, the two parallel PIN diodes between the Ant and Tx ports in series configuration are forward biased by applying 30 V to the Tx bias input port. The magnitude of the resultant bias current through the diode is primarily determined by the magnitude of the voltage across the PIN diodes (applied voltage), the 220 Ω current limiting resistor. (The bias tee components and the 220 Ω current limiting resistor are integrated into the module AlN substrate). The bias current in this state is nominally 130 mA. At the same time, the two PIN diodes connected between Bias (B2) and the GND in shunt configuration are forward biased by applying a negative voltage source of 5 V (nominal) with a current sink limit of 25 mA.
Simultaneously, the Rx series PIN diode, which is connected between the Ant and the Rx port, is reverse biased during the transmit state. Typical isolation of 40 dB from Ant to RX path is achieved from the single non-conducting series diode the two conducting shunt diodes in the Ant to Rx path.
The minimum voltage required to maintain the series diode in the Ant to Rx path of the switch out of conduction is a function of the magnitude of the RF voltage present, the frequency of the RF signal, and the series diode’s parameters, among other factors.
In the Tx state, a typical insertion loss of 0.3 dB is achieved, and the module can handle RF CW power of 51 dBm from Ant to Tx.
In the Rx state, the single series diode between the Ant to Rx port is forward biased by applying 5 V to the Bias (B2) port. The magnitude of the resultant bias current through the diode is primarily determined by the magnitude of the voltage across the PIN diodes and the 220 Ω current limiting resistor. This current is nominally 25 mA. This biasing condition at Bias port B2 reverse biases the two shunt diodes, which reduces the insertion loss parameter degradation. Further down the Ant to Rx path there is a combination of thin limiter PIN diode and a zero bias Schottky diode. The DC blocking capacitors eliminate the influence of the switching diode bias voltages on the limiter circuit.
Simultaneously, the PIN diode connected between the Ant and Tx path is reverse biased by applying a bias voltage, nominally negative 5 V with 0 mA. The diodes are reversing biased, thus isolating the Tx port from the Rx signal path.
These biasing conditions produce low insertion loss in the ANT-Rx path. The additional PIN diode and the Schottky diode in the shunt configuration work as a self-actuated limiter.
During this state, if the RF voltage incident on the Ant port of the module is not greater than the limiter diode junction potential, the limiter diode is in non-conducting state and RF power at the Ant port reaches the Rx port. If the available RF voltage at the Ant port increases and is above the diode forward junction potential, the diode starts conducting. This allows the limiter PIN diode to store charge in its I layer, which reduces the diode’s series resistance accordingly.
This reduction in shunt impedance across the transmission line causes some portion of the incident signal to be reflected. As the diode’s impedance decreases, the current maximum at the location of the diode increases the charge in the diode, further reducing the diode’s resistance.
The zero bias Schottky diode (ZBD) has a typical junction potential of only 200 mV. The ZBD also acts as a DC return to the limiter PIN diode in the circuit in the ANT to Rx path. The ZBD is the key contributor for the module’s ultra-low flat leakage power and spike leakage energy.
In the Rx state, a typical insertion loss of 1.0 dB is achieved, and the module can handle RF CW power of 33 dBm from ANT to Rx with a spike leakage power of 0.1 ergs typical and flat leakage power of 13 dBm typical.
Switch Limiter Application
The switch limiter performs active and passive receiver protection. Active receiver protection is effected when the ANT-Tx path is in its low insertion loss state. Large signals incident upon the antenna terminal are directed to an external high power 50 Ω termination connected to the Tx port. When the device is biased to produce low insertion loss in the ANT-Rx path, any large signal incident upon the ANT input port will activate the passive limiter circuit comprising of the limiter PIN diode and the Schottky diode located at the RX output port. This passive limiter circuit produces an impedance mismatch in the ANT-Rx signal path which reflects the large incident signal back towards its source, thereby protecting the sensitive radar receiver system.
The limiter integrated switch module is designed with a bias tee and current limiting resistor for optimal biasing built into the module. Additionally the Ant and the Rx port of the module have DC blocking capacitors making them ready for direct interface with the external radiating/receiving elements and the low noise amplifier for the received signal processing for intelligence. The Tx port can be directly connected to the 50 Ohm termination to protect against high power reflections from the radiating element during the high power beam transmission.
The limiter integrated switch module is manufactured using Aeroflex / Metelics proven hybrid manufacturing process. The limiter section of the module is a self actuated circuit with no external driver needed. This section provides protection against signals up to RF CW incident power levels of 33 dBm with flat leakage power of less than 15 dBm and spike leakage energy of less than 0.2 ergs.
About the Author
Chandu Sirimalla is the senior applications engineer for switch and control products at Aeroflex / Metelics. He designs and supports high power RF/microwave switch and limiter products. He has prior experience in Transient and ESD protection applications.
J. White, Microwave Semiconductor Engineering , Artech House, 1977
IEEE Std, 521-2002 , Standard radar band nomenclature
Bruce A Kopp , Craig R Moore and Robert V Coffman “Transmit/Receive Module Packing : Electrical Design Issues” Johns Hopkins APL Technical digest, vol 20 number 1 (1999)
C. Sirimalla , R. Cory, “Wide Band Quasi-Active Limiter Design”, Proceedings EWIC 2012
Skolnik, M., Introduction to Radar Systems, McGraw-Hill, New York, 3rd Ed., 2001
Datasheet, P/N- MSWLM2420-242, Aeroflex / Metelics.
this article to a friend!