Multi-paction effects can impact microwave components in high-power satellite communications (sat-com) systems. The nonlinear break-down-voltage phenomenon occurs in high-vacuum environments above a certain threshold voltage, and can degrade the performance of RF and microwave components or, in extreme cases, damage the components or the system. Although multi-paction effects are often difficult to predict and measure, properly equipped test systems with dedicated software can accurately identify microwave components that may multi-pact, effectively screening them to avoid damage in a deep-space application.
Performance degradation or damage due to multi-paction processes is associated with waveguide components used in deep-space applications, occurring at high power levels in a vacuum environment. Multi-paction phenomena can affect both waveguide and coaxial passive components in the transmit signal chain, such as diplexers, filters, multi-couplers, and antennas, as well as the coaxial cables and connectors used to link these components.
For a multi-paction discharge to take place, three ingredients are needed: an RF source, free electrons, and a vacuum. The process occurs when a charged particle (free electron) inside a gap (such as a waveguide component) in a vacuum environment, or one with relative low atmospheric pressure, oscillates under the influence of a large external electric field (the high-power RF source, such as a transmit amplifier). Every time these accelerated charged particles strike the gap wall, they release secondary charged particles. As the process is repeated millions of times per second, the number of electrons multiplies rapidly, leading to a multi-paction discharge. The discharge itself may absorb or release little power, but can cause increased outgassing in the component or system, which can lead to a gas discharge. A gas or corona discharge in a space system can release a large amount of power, thus causing component performance degradation or even system failure.1
Multi-paction occurs when the ambient pressure is sufficiently low that the electron mean free path is longer than the electron separation distance. Multi-paction effects have been observed at pressures less than 10-2 Torr. A vacuum environment is required for a multi-paction discharge since at atmospheric pressure, the charged particles are more likely to collide with air particles, reducing the velocity of the charged particles as well as their potential to release secondary charged particles. Waveguide components are often pressurized for this reason to avoid the effects of multi-paction at high RF power levels.
Mathematically, multi-paction events are understood as a function of the input power to a component, according to Larmor’s formula:
P = 2(e2a2)/[3(4πε0c3)]
where P = the power generated by an accelerated charge, e = the charge of an electron, ε0 = the dielectric constant, a = the acceleration of the charge, and c = the speed of light in a vacuum, or 3 × 108 m/s.
The role of a circuit’s dielectric constant, ε0, can also be seen in this formula, where higher values of dielectric constant will result in reduced power for the accelerated charge. Similarly, the multi-paction threshold voltage, V0, can be found from the equation:
V0 = (2πd/λ)2 [mec2/πe]
where V0 = the acceleration voltage between charged surfaces, me = the mass of an electron, λ = the wavelength, and d = the spacing between surfaces.
The dimensions of an RF circuit also play a considerable role in the occurrence of multi-paction effects. Since multi-paction effects are essentially arcing of electrons across a narrow gap, sharp or jagged edges within an RF/microwave component can provide the geometry for bunching of free electrons in high concentrations, leading to a discharge event.2 Conversely, rounded metal surfaces and thin dielectric coatings within a component can lessen the possibility of a multipaction discharge even within a high vacuum at high RF power levels.
Identifying Multi-Paction Events
Multi-paction effects can be detected by a number of different approaches, which make measurements globally (on the system) or locally (on a component). Global methods determine whether multi-paction is present somewhere in a satcom or other system, but don’t isolate which component is the source of the multi-paction effects. Local approaches are more useful during the system development stage to identify “weak links” in the system at the component level. Local evaluation methods can also be used to monitor part of a system under continuous operation in order to better understand long-term effects. Multi-paction effects can often be elusive to spot, and it can take years for a multi-paction discharge to occur. Over time, persistent multi-paction events can cause erosion of component surfaces, which degrade electrical performance and eventually lead to component and/or system failure.
Common global methods that analyze system performance under conditions conducive to multi-paction effects include detection of increased close-to-the-carrier noise, increased second and third harmonic levels, and changes in output power. Unfortunately, since the first two approaches rely on the generation of noise, they are prone to errors from noise caused by non-multi-paction causes. Multi-paction measurement systems typically suffer from a number of different sources of noise, including from microwave signal generators. Excessive noise from other sources can lead to establishing a too-low value for the multi-paction threshold of a system. Similarly, failing to detect a transient multi-paction event can lead to establishing an excessively high multi-paction threshold for a system.
Global techniques have also been developed based on the use of amplitude modulation (AM) detection, using high-speed sampling techniques to generate a Fast Fourier Transform (FFT) display of the test signal. This approach relies on the signal difference between the input test signal and the multi-paction threshold. When the signal carrier passes the multi-paction threshold, it produces a distinct modulated output signal. An FFT will reveal any periodicity in the measured signal.
Both global and local measurement methods have been developed for single-carrier and multiple-carrier test cases in order to closely emulate operating conditions of satcom and other high-power communications systems that employ complex modulation formats. In multicarrier systems, for example, the peak power levels of the system can vary widely, depending upon the relative signal phases of the carriers.
Traditional test systems have relied on a high-power RF/microwave source usually consisting of a frequency synthesizer and external pulse modulator boosted by means of a high-power traveling-wave-tube (TWT) amplifier (TWTA). Some means of power control is needed, usually in the form of a series of couplers and power dividers and precision variable attenuators. Phase shifters are typically used to control the phase of the test carrier signals, and often impedance tuners will be enlisted to ensure precise control of the impedance match to the device under test (DUT). The DUT itself, which may be a component, subsystem, or full system, is isolated in a vacuum chamber to emulate deep-space atmospheric conditions. Traditional monitoring equipment includes peak power meters, spectrum analyzers, and digital storage oscilloscopes (DSOs). Such a system (Figure 1) is designed to precisely compare signals entering and exiting the DUT for evidence of nonlinear behavior that might signal the onset of a multi-paction event. One of the goals of any multi-paction test system is not to instigate multi-paction events that lead to the destruction of a DUT, but to identify the threshold voltage and power levels for multi-paction events, below which the DUT can operate safely.
One of the problems with the traditional multi-paction test setup is its reliance on spectrum analyzers to identify spectral phenomena, such as increased levels of third-harmonic signals, that may be extremely short lived and difficult to capture, even by a high-speed DSO. Analog spectrum analyzers operate with swept analog resolution-bandwidth and video filters. Resolution increases as sweep speed decreases, and often an analog spectrum analyzer’s sweep speeds are incapable of capturing the short-term increases in third-harmonic distortion that may be the signals for a multi-paction event. Newer, real-time spectrum analyzers, which rely on digital capture of a fairly wide instantaneous input bandwidth, can improve the chances of capturing the transient events related to multi-paction effects, but these instruments still resolve signal information by means of swept filters, albeit using digital signal processing (DSP) for those filters.
Test System Options
The block diagram for an improved multi-paction test system is shown in Figure 2; a photograph of an actual system is shown in Figure 3. This system relies on multiple-channel, high-speed, analog-to-digital converters (ADCs) to capture test signals entering and exiting the DUT. As with a traditional test system, the DUT remains in a vacuum chamber, with high-power coaxial transmission lines to route test signals to it and from it. The transmission lines are housed within a high-power cart, which provides mobility for ease of connection to different DUTs.
The test system uses phase-nulling techniques to study changes in the linearity of the input signal versus the output signal of a DUT. In a conventional system, incident and reflected test signals are summed and results displayed on a spectrum analyzer or peak power meter. But unlike a traditional system, in which the combination of a variable attenuator and a phase shifter and manual tuning are used to null the phase of the incident and reflected test signals, this type of system digitizes the incident and reflected signals by means of a dual-channel, high-speed, analog-to-digital converter (ADC) and uses digital signal processing (DSP) techniques to manipulate and analyze the data. By analyzing the digitized signals by means of dedicated software, any adjustments needed between the incident and reflected signals in order to detect multi-paction events can be done almost instantaneously while still controlling the time constant to simulate what were formerly manual adjustments of variable attenuators and phase shifters to account for changing conditions at the DUT.
This technology differs from traditional test systems in that it can acquire data continuously, with zero dead time in the measurements. The ADCs digitize and send all test signals to the system controller, which runs the specialized measurement and analysis software. The software operates according to predefined parameters, such as threshold voltages, phase nulls, amplitude changes, increases in third-harmonic levels, etc., to acquire test data of interest. The hardware, in the form of the computer-controlled digitizer, continuously monitors the test signals for any kind of multi-paction-related event, and the software triggers on various event thresholds to either store or disregard the captured signal data.
The combination of hardware and software ensures that measurement channels are phase matched and phase coherent, even accounting for manufacturing tolerances in passive components, such as couplers and attenuators. Each measurement channel is sampled and measured for any deviations in phase, and corrections are made automatically under software control.
This approach uses various techniques to identify possible multi-paction events. For example, it measures changes in system and DUT return loss under different signal conditions, since a progressive or dramatic change in return loss can indicate that a DUT is defective or likely to suffer multi-paction discharges at high power levels. The system is designed to evaluate a device under swept power levels, beginning with relative low voltage and current levels and ramping to levels as high as multiple kilowatts of power, monitoring DUT performance for changes during the increases in power levels.
An ADC-based system configured in this manner can test as many as four DUTs sequentially, with only about five seconds required to switch a test sequence from one DUT to another. The system can pro-
vide test results in a variety of formats. In one approach, the degradation of pulsed waveforms under changing signal conditions can help identify impending failure of a DUT due to multi-paction effects (Figure 4). Once a test routine is begun, it is completely under software and computer control, with no human intervention needed. It can run continuously, acquiring test results continuously according to preset conditions stored in the software prior to running the test.
A typical system could be configured to perform testing in bands from 1 to 18 GHz at power levels as high as 10 kW at 2 GHz. It provides extensive monitoring and display capabilities, allowing an operator close scrutiny of real-time events over a wide range of frequencies and power levels. The system is also calibrated to account for variations in performance as a function of temperature and power. For example, the insertion loss of the system test cables will increase as temperature rises due to high power levels. The initial calibration accounts for these temperature-dependent system performance variations in order to provide measurement results that are accurate in amplitude within ±0.1 dB and in phase within ±0.1 deg.
For those interested in predicting multi-paction effects on their own DUTs, the European Space Agency (ESA) offers a free online multi-paction calculator at http://multipactor.esa.int/. The calculator includes a variety of RF software tools, such as a mismatch calculator, skin depth calculator, waveguide matching transformer, rectangular and circular waveguide insertion loss calculator, venting and outgassing calculator, and shielding effectiveness (SE) calculator. It can be used for analyzing microstrip, stripline, coaxial connectors, and standard waveguide gaps.
This article was written by Howard Salvesen, Vice President, Engineering at In-Phase Technologies, Inc., Clarksburg, NJ. For more information, Click Here
- S. J. Feltham and S. Johns, “The ESTEC Multipaction Support Facility,” Preparing for the Future, ESA Newsletter, European Space Agency, Vol. 7, No. 4, 1997, p. 2.
- Harlan S. Hurley and Dennis J. Kozakoff, “RF Multicoupler Design Techniques to Minimize Problems of Corona, Multipaction, and Stability,” NASA Contractor Report, NASA CR-61387, NASA-George C. Marshall Space Flight Center, AL, December 1971.