by Mark Blackwood – Fairview Microwave
From antenna apparatus in transmit/receive chains to cellular tower installations, waveguides have offered high-powered and extremely low-loss solutions for signal transmission. Still, not every setup is the same, and the connectorization between the waveguides can prove to be challenging as the waveguide build is often bulky and less agile than its coaxial cable counterpart; the solid rigid metal alloy casing can heavily limit routing capabilities. Oftentimes, the waveguide installation can be a tail-end consideration; this often leads to a misalignment where the two connecting flanges are offset and cannot be fitted or bolted together. This is an unfortunate scenario as waveguides require precision manufacturing and skilled labor to assemble these components within tolerance; an attempt to force two waveguides together can result in damage and a significant degradation in performance. Still, size and weight restrictions force creative routing solutions such as upgrades and retrofits; waveguides that can bend, twist, and flex are specifically engineered for these types of situations.
The typical rectangular rigid waveguide is proportioned according to specific wavelengths to generate a precise electric or magnetic field pattern in order to effectively pass signals of certain frequencies,\; this is why there is a “cut off” frequency for a waveguide or a “starting” frequency for ideal transmission. The signal transmission in a waveguide can only take place in either transverse electric wave (TE) or transverse magnetic wave (TM) modes, where the electrical field or magnetic field is wholly perpendicular, or transverse, to the direction of signal propagation. A waveguide can also be described according to its width and height, where the width corresponds to the H-plane, or the plane that the magnetic field is aligned with, and the height corresponds to the E-plane, or the plane that the electric field runs parallel to. An installation that requires a change in direction, or a bend in the E- or H-planes, must be done with a level of exactness to avoid signal distortion at the output—an improper bend can cause significant loss and reflections.
There are a variety of considerations when it comes to bending waveguides, depending on the application. General practice during an installation is to inspect a waveguide for dents, corrugations, or other rapid changes in cross section in order prevent any impedance mismatch. While it’s not always possible, it is ideal to minimally utilize bends, twists, joints, and couplings to lessen any variability in performance of the overall system.
E- and H-Bends
A bend in a waveguide is a 90 degree turn that distorts either the magnetic or electric field; an E-bend (also known as an “easy” bend) will distort the electric field ,while the H-bend (also known as a “hard” bend) distorts the magnetic field (Figure 1). Unless carefully fabricated, a bend in the waveguide can appear as a discontinuity and cause reflections of the signal; a gradual bend can prevent this. The general rule of thumb is that the waveguide bend radius must be greater than two wavelengths to ensure peak performance.
The military standardized handbook for the fabrication of rigid waveguide assemblies, MIL-HDBK-660B, lists a number of methods to bend waveguides properly. Initially, the waveguide is annealed to a certain temperature for an allotted period of time and cooled at room temperature. Then, the hollow portion of the waveguide is filled with either a mandrel or some solid material to prevent any bump or discontinuities on the internal surface during the bending process. There are a variety of machines to form a waveguide, from precision bending apparatus to hydraulic push presses and hand bending machines. All of these methods essentially guide the waveguide in position and gently bend it to a specific radius. Finally, the waveguide is cleaned and degreased in a caustic bath and rinsed in an acidic bath before the flanges are braze welded on. Figure 2 shows the flowchart for this.
Mitered 45º Bends and Sharp 90º Bends
While having a waveguide that changes directions increases accessibility in various installation applications, the large bend radius may still take up too much real estate; there are situations that require a 90º turn in less space with comparable performance. For more rapid changes in directions, there are angled bends where a 90º turn is accomplished by leveraging segments of 45º angles, or a double miter (Figure 3a and 3b); the two consecutive 45º turns can provide a lower VSWR while still accomplishing a sharp turn. Ideally the length of corners, or the distance between two 45º bends, is an odd multiple of ¼ wavelengths apart; this way the reflections that are generated by a corner are cancelled out by the second corner.
A direct 90º turn can be frequency sensitive and difficult to impedance match at the corner, thereby causing reflections and an overall lossy system. There is a simple bending technique to overcome this where there is a sharp 90º bend in the innerwall while the outerwall is curved on a predetermined radius. The radius, or the curvature of the outerwall, determines the impedance matching capabilities; it can be adjusted to nearly eliminate all reflections for an ideal impedance match. Other variations of this method include a sharp 90º inner turn with a polygon-like angled exterior with 5 edges.
Mitered Bends Construction
One method of developing miter corners is accomplished by shearing 45º V-shaped segments of predetermined dimensions out of a straight rectangular waveguide. The pieces are the folded together to create the bends in the waveguide. This can allow for more precision cuts and connections to limit reflections and losses (Figure 4).
A waveguide twist is essentially a section of the waveguide in which there is a progressive rotation along the longitudinal axis (Figure 5), which switches the polarization of a wave. In other words, the electromagnetic field in the waveguide is rotated in order to phase-match with a load. To avoid losses due to reflections, the length of the twist is an odd multiple of the quarter guide wavelength; in most cases this length is kept to a minimum of double the guide wavelength. For applications that require more fine-tuned adjustments in polarization there are also 45º waveguide twists.
Flexible waveguides (Figure 6), also known as semi-rigid and hybrid waveguides, offer a more agile solution in field-based waveguide installations. From moving equipment exposed to vibrations to quick fixes for misalignments, there is no one-size-fits-all waveguide—even with the variety of bends and twists, there are scenarios that call for flexibility. The flexible rectangular waveguide is accomplished by leveraging a flexible retaining sheath of either a continuous spiral of interlocking metal, or a convoluted metal rectangular tube; both are molded with a rubber jacket. Flexible waveguides are generally designed with a distance that is an odd multiple of quarter wavelengths between each groove, or dip in the cross section; this provides cancellations of any reflections and energy propagated back.
Flexible Waveguide Construction
Similar to a metal hose, a flexible waveguide is generated with an interlocking metal structure; the rectangular body is normally made of brass and coated in silver for conduction. The flexible waveguide core and jacket can vary depending upon the installation; while some applications require more flexure including bends and twists over electrical performance, others require superior PIM performance over flexibility. Demonstrated in Figure 7c.
Flexible and Twistable
Waveguides that require both flexibility about the E- and H-planes, and the ability to twist, have a core that relies upon a friction joint between interlocking or wound segments of metal (Figure 7a). Additionally, the mechanical tension between the interlocking metals could also be used to maintain an RF metal plane. Twistable and flexible waveguides are analogous to the hand-formable coaxial cable; it can be formed into a particular shape and reformed later without significant degradation in performance. Still, these cores rely on their neoprene jackets for a pressurized seal which may not hold up under certain installation or environmental conditions, and may lead to failures if handled roughly.
To eliminate the twist, the interlock between the joints can be filled with a malleable solder (Figure 7b). This method also retains a pressure seal and can be more desirable for field applications that are exposed to the environment. Another form of a flexible but non-twistable waveguide is the seamless waveguide, which is constructed of a corrugated rectangular shape. This physical structure allows only for a limited number of E- and H-bends. The seamless waveguide generally is less lossy than interlocking waveguides and can also hold up to outdoor installations, such as cell towers with high vibrations, as this method excels in PIM performance. The seamless waveguides are similar to semi-rigid coaxial cables; they can generally be formed only once to hold a particular shape, but they are not reformable.
Flexible waveguide standards exist to establish a level of performance that prevents failures in a variety of installations. Flexible waveguide military mechanical and electrical standards are listed in MIL-DTL-287, and commercial standards in IEC 60636. The electrical performance tests include measurements for attenuation (insertion loss) and power handling, while the environmental and mechanical tests ensure that the waveguide can uphold performance in harsh environments. This would include salt water exposure, vibration, temperature cycling, and flexures.