1. Home
  2. Featured Articles
  3. An Introduction to Passive Interconnect for Millimeter-wave Test

An Introduction to Passive Interconnect for Millimeter-wave Test


by Fairview Microwave

A few years ago, working with electronic signals at millimeter wave frequencies was almost exclusively the domain of advanced military and scientific applications. Consumer products did not exceed the 5.7GHz ISM band, and commercial satellite communications were bounded by the Ka-band at 27GHz.  The cost of high frequency semiconductor technologies limited their application use.  But a “bandwidth-hungry” consumer market has intersected with the advent of inexpensive millimeter-wave electronics on silicon and has produced a tremendous wave of new high-frequency applications, including those targeting the 60GHz IMS band.  Automotive RADAR is moving to 77 GHz; 5G NR FR2 is targeting frequencies from 20 to 60 GHz; and 802.11AD, 802.11AY (WiGig), and WirelessHD applications are in the 60 GHz band.  And who knows what other high frequency applications the future holds?

As a result, a generation of high frequency engineers must grapple with the benefits and pitfalls they are encountering when working in these higher bands.  They require the active and passive components necessary to implement benchtop experiments and to evaluate the performance of devices and systems during the R&D process for both commercial- and consumer-targeted applications. This article covers a portion of related topics.

In Case You Didn’t Know

The term millimeter-wave generally refers to signals with a wavelength (λ) in the range of 1mm to 10mm, which corresponds to frequencies in the range of 30 GHz to 300 GHz; sub-millimeter-wave is often used to refer to frequencies above 100GHz. The two significant transmission issues at these frequencies are atmospheric absorption and diffraction.   

A look at the attenuation between space and the surface shows 100% attenuation for much of the spectrum, but the attenuation drops off dramatically for wavelengths between 1cm and 11 meters.  This is referred to as the Radio Window.  

Much of the above absorption spectrum is related to phenomena that occur in the upper atmosphere. The plot in Figure 2 shows atmospheric attenuation near sea level.  Notably, there is a very high level of attenuation centered right around 60 GHz, which is caused by molecular Oxygen (O2).  As a result, this band is only useful for short-range communications. However, the advantage is that nearby systems are less likely to interfere with each other in this band and transmission signals are also more difficult to intercept, thereby minimizing privacy and surveillance concerns.

 Very low frequencies have large angles of diffraction, allowing them to transmit signals around obstructions.  This is leveraged for communications signals and RADAR to reach beyond the line-of-sight (LOS).  Anywhere ranging from radio and television bands through the traditional cellular bands, this is viewed as an advantageous property.  However, because signals are likely to arrive at the receiver by following multiple paths, multi-path and fading mitigation become important aspects of waveform, channel-mediation, and antenna (MIMO) design for modern communications electronics.  

At millimeter-wave frequencies, the diffraction angle is very small, and signals are primarily received via LOS.  At the same time, the short wave lengths allow for the design of active phased-array antenna that can be implemented on a single chip, allowing optimization of gain along off-boresight links. 

The single biggest problem affecting everyone who works in a high frequency laboratory is connectivity.  Even at more traditional frequencies, the cables, connectors, board-to-coax launches, etc. require skillful attention in design, test, and manufacturing. At millimeter-wave frequencies this becomes more challenging. 


In rectangular waveguides, the losses are fairly low. As you ascend to the terahertz range, they become so significant that NIST has designed a modified TRL calibration for Vector Network Analyzers to account for these losses.  Because of reduced skin depth, by comparison, the losses in coaxial and other transmission lines are much higher.  For instance, a 12-inch 1.85mm coaxial cable can lose well over 4dB. Benchtop work and measurement with coax is simply more difficult.  Even when using waveguide components whenever possible, most benchtop and probe station work will still require a few coaxial cables, and engineers often find it challenging to make the necessary connections while keeping cable lengths short to minimize loss.


Transmission lines, connectors, and other items must scale to match your signal’s wavelength. The manufacturing tolerances for higher frequency components become unforgiving. They are more expensive to produce and more likely to have problems both in construction and use.  An important consideration with transmission lines is their signal propagation mode.  The way that the magnetic (H-Field) and electric (E-Fields) fields couple with the transmission line, in conjunction with the dielectric, determines the phase velocity of the signal.  If the signal frequency is too high for the design, the signal can couple with a transmission line in more than one way, using multiple modes of propagation. These signals will then traverse the transmission line at different velocities, and the result can be severe linear distortion and attenuation.  “Multimoding” is also a problem for connectors, adapters, and board-to-coax interfaces.  For example, typically before multimode becomes a problem, a 3.5mm connector can operate up to 26.5 GHz (sometimes to as high as 30 GHz), a 2.92mm connector up to 40 GHz, a 2.4mm connector up to 50 GHz, a 1.85mm connector up to 67 GHz, and a 1.0mm connector up to 110 GHz.


The potential for millimeter-wave is seemingly endless as more and more applications are being explored in the cellular, imaging, and radar spaces. As described, there are many challenges associated with developing and testing devices at high frequencies. For most consumer and commercial applications, this was once primarily encountered because of the need to evaluate harmonics. With carrier signals themselves now at these frequencies, the type and cost of test instrumentation and tools can be substantial, and there is virtually no margin for error allowed for these tools—whether referring to instrument operation, cleanliness, or the torqueing of connections. For example, the accurate performance characterization of active electronically-scanned antennas requires a wide range of interconnects to accomplish the task. Acquiring a general understanding of millimeter-wave interconnects and their respective strengths, weaknesses, and potential issues for testing purposes can be very beneficial, as the technical considerations can be more nuanced than for microwave frequencies. 

1. Microwave Engineering by V.S. Bagad

2. https://www.microwavejournal.com/articles/26932-rf-waveguide-tips-and-facts-you-cant-learn-in-school

3. https://www.pasternack.com/t-calculator-skin-depth.aspx

4. https://www.mpdigest.com/2017/02/23/your-guide-to-waveguide-plumbing-how-do-waveguides-bend-twist-and-flex/