Introduction to waveguide
Waveguide is a huge topic for anyone studying microwave engineering, we have barely touched on it here.
Waveguides are metallic transmission lines that are used at microwave frequencies, typically to interconnect transmitters and receivers (transceivers) with antennas. OK, some purists will tell you that waveguide is not a transmission line, because it doesn`t have two conductors, but we don`t draw such a distinction here. We will be discussing rectangular waveguides for the time being here at Microwaves101, but you should know that other waveguide structures such as circular and double-ridged are available.
Waveguide has a number of advantages over coax, microstrip and stripline. It is completely shielded (excellent isolation between adjacent signals can be obtained), it can transmit extremely high peak powers and it has very low loss (often almost negligible) at microwave frequencies.
One disadvantage of waveguide is its high cost. Manufacturing volumes are low, and waveguide materials such as copper and silver are relatively expensive. Other disadvantages include unwieldy size and mass, particularly at lower frequencies. If your cell phone employed waveguide components, it would need wheels because it would be too heavy to lift! A final disadvantage of waveguide is that you can`t pass DC currents along with your RF signal. Fear not, the advantages of power handling often outweigh all of waveguide`s perceived short-comings!
To reach megawatt power levels waveguide can be pressurized with special gasses that increase the peak power level before the waveguide short circuits with electrical arcing between the top and bottom walls. Silver plating used on the inside walls of the waveguide decreases the resistance loss making the common aluminum or copper waveguides even more efficient. The end of a waveguide can be flared out to form a horn antenna, the most common antenna used to illuminate parabolic dishes.
Waveguide transitions
Waveguide can be interfaced with coaxial cable by using simple antenna probes reaching into the waveguide to excite the waveguide mode. There are many methods of building microstrip-to-waveguide transitions, a common one is an E-plane probe with a backshort. The backshort is positioned 1/4 wave away from the probe, and reflects EM energy that made it past the probe back to the probe where it combines in phase with the incident wave.
Many shapes of waveguide sections, switches, twists etc. with coupling flanges on the ends can be screwed together to form the complex shapes to fit inside aircraft, spacecraft, ships and other applications. Even flexible waveguides made from spring-like (Slinky) material are used; however, these are not as efficient in transmitting microwave energy.
E-plane and H-plane
Within a waveguide cross-section the electric field is normal to the broad wall and the magnetic field line is normal to the short wall. The maximum positive and negative voltage crests of the wave travel down the center of the waveguide and the voltage decreases to zero along the waveguide side walls. When high power waveguide systems fail, the electrical arcs are usually between the top and bottom walls of the waveguide in the center where the voltage is greatest.
Waveguide E-plane and H-plane Rule of Thumb
Somebody in the lab asks you to get them an E-plane bend or an H-plane bend. You can`t remember which way the fields go in the waveguide, but you don`t want to look stupid by asking. Don`t panic, there is an easy easy way to remember which is which. The E-plane bend is bent the easy way , and the H-plane bend is bent the hard way , which you can see in the photo below. If it isn`t obvious to you what is meant by easy and hard way when you are bending a rectangular rod, it is not too late to consider a career shift to the software industry.
H-plane bend and E-plane bend (WR-28)
Waveguide components
All manner of waveguide components exist, including circulators, isolators, attenuators, loads, mixers, amplifiers, you name it.
Below are some pictures of some waveguide splitters you may find in your lab. Note that basic network theory says that you can`t make a three-port splitter that is lossless and matched at all three ports, so if you want to split a signal, your best bet is the magic tee, just feed the sum port, terminate the delta port and the outputs are the co-linear ports.