What is the difference between a waveguide/horn and regular metal pipes?

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
4 replies · 2K views
BlackPowder
Messages
14
Reaction score
2
What is the difference between those horns/waveguides and regular metal pieces with a same geometry? Why the microwave companies sell those parts at hundreds and thousands dollars? Why we cannot buy some metal sheets or pipes on McMaster-Carr with very low price and make some microwave components by ourselves?
 
Engineering news on Phys.org
There is quite a lot of black magic talked about with waveguides, because the waves will go through any old copper pipe. However, a rectangular WG has the advantage that there is only one mode, so it avoids multiple signals which can be a problem for high bandwidth transmission of data or TV. It is also difficult to use ordinary pipe bends because a "proper" waveguide will have carefully chosen bending radius and will often have tuning screws to reduce unwanted reflections. I cannot see any problem in principle about using circular tubing, although more modes may spring up and polarisation may rotate due to imperfections. One issue is that it is hard to make WG to coax transitions, so that is why peope buy them and then stick to standard sizes. Incidentally, for 50 years it has been the practice for microwave links to use corrugated elliptical waveguide, which can to some extent be bent like a cable.
 
BlackPowder said:
Why we cannot buy some metal sheets or pipes on McMaster-Carr with very low price and make some microwave components by ourselves?
We can and some do.

Losses in a waveguide are less than in coaxial cable because there is no dielectric. Losses in the waveguide are therefore due to the internal surface resistance. That suggests cheap aluminium or expensive copper tube, or a silver or gold plated base metal. But reactive metals gradually develop a surface film that becomes a lossy dielectric, which may also absorb water. Surface protective films are also lossy dielectrics at GHz frequencies. Filling the guide with dry nitrogen is sometimes used, but that is expensive and requires a seal or dielectric window at each end.

Aluminium is chemically too reactive, copper oxidises with losses gradually increasing over time, gold plating is more lossy than copper in the short term, but losses with gold do not increase with time, silver plating is usually a good compromise.
 
  • Informative
Likes   Reactions: berkeman and Klystron
The first demonstrations of microwave transmission through waveguides were performed around 1930 by G. C. Southworth. He used round copper rain downspouts one mile long visible in the photo below.

George_Southworth_%26_waveguides.jpg

Southworth at Bell Labs in 1943 standing in front of experimental waveguide line he used in his original research into waveguides during the 1930s. (Wikipeadia)

As tech99 pointed out, however, round waveguides mix modes and polarizations, as do imperfections in rectangular guides. It's also true that pipe joints give poor electrical performance unless they are carefully soldered or welded. Commercial waveguides, on the other hand, have flanges with built-in quarter wave chokes that provide nearly lossless performance in a bolt-together joint.
1571975209365.jpeg
 
  • Like
  • Love
Likes   Reactions: Klystron, dlgoff and berkeman
BlackPowder said:
What is the difference between those horns/waveguides and regular metal pieces with a same geometry?

At some fundamental level I suppose there isn't a huge difference but I would question if those pieces of metal are actually the same geometry. Precision matters. Horns and waveguides are manufactured to very tight tolerances and that is not cheap.

BlackPowder said:
Why the microwave companies sell those parts at hundreds and thousands dollars?

Let's consider some of the costs to the manufacturer. Many geometries are possible. After all pieces of metal come in many shapes and sizes. Someone has to pay an engineer to determine what geometry will solve the particular problem at hand and that costs money. The precision machining mentioned above costs money. And, very important and often left out, someone has to go an do the measurements that the design and particular part you will be sold meets the specification. Those measurements are often very expensive.

If the part is unique or exotic in some way there will not be many of them to amortize the costs over hence the high price. This is very typical as one gets away from the main stream, almost by definition.
 
  • Informative
Likes   Reactions: Klystron