Transmission Line

[ms660]

Ok, I think I understand.

Thanks.

Lawrence.

[G0HZU_JMR]

Originally Posted by pwdrive

On the subject of transmission lines and reflections, here's an interesting series of articles by the late Walter Maxwell:

I've seem links to this before (usually on QRZ.com) and I quickly got the impression that WM was an antenna engineer with a great deal of knowledge about antennas and T Lines but he didn't seem to look much beyond this when looking at the whole system. The whole system includes a transmitter with a PA stage that has special design requirements and limitations of its own. This is especially the case for solid state PA design. These PA design requirements and limitations often clash with the simplistic model or 'vision' of a typical PA stage by an antenna engineer who maybe doesn't have adequate PA design experience.

[Radio Wrangler]

That's a good synopsis, Jeremy.

If you come across things from the amateur radio side of things, you come across Walt Maxwell because of his CQ Magazine and ARRL connections.

Fortunately, designing PAs isn't particularly difficult, it just needs a different viewpoint for things to become clear.

David

[ms660]

And possibly visa versa. Ref.post #22

Lawrence.

[m0cemdave]

I seem to remember a major argument about conjugate matching going on in Wireless World in the 1970's, was this part of the same scientific spat?

[Radio Wrangler]

No, the Maxwell/Bruene business took place in the US with only a few on this side of the water even aware.

The wireless World conjugate matching discussion was a storm in a teacup. Conjugate matching is effectively the tuning out of a source with a reactive term by a load with an equal but opposite reactive term. Effectively they form a parallel tuned circuit, which vanishes...well, at one frequency. The fun came from people forgetting it was the creation of a resonator. Techniques for broadening the bandwidth of resonant transformations are interesting. There's a lot of ideas in some of Helge Granborg's notes, though without explanation of what's going on.

David

[ms660]

General question:

With conjugate matching does the length of the transmission line come into play or is the length immaterial ?

Lawrence.

[Radio Wrangler]

Transmission line lengths do matter where conjugate matching is being used.

Let's say you had a nominally 50 Ohm system, but one of your devices is a bit off-tune and looks a bit capacitive as well. What you want to connect it to is adjustable, so you can fix this by tuning the second device so it looks like 50 Ohms resistive, plus a bit of inductive impedance. You adjust the inductive component until it exactly cancels the capacitance of the first device. On that one spot frequency, the inductance and capacitance terms cancel, and both vanish. If one of your devices is a length of line, it is being run with a non-zero reactive term, so what there has to be at the far end depends on the line length. The length of line depends how far round a fixed radius on a Smith chart the impedance value rotates.

David

[ms660]

Thanks, I'm starting to get the picture, also found this link which I found helpfull:

http://www.navymars.org/national/tra.../14182_ch3.pdf

Lawrence.

[Skywave]

Stub matching can be a time-consuming and a tedious process. Years ago when I was a member of a team doing R & D on an unusual vertical aerial for UHF, that aerial presented a very awkward impedance which had very low R and very high C. Matching was achieved by a stub - but finding the right length of stub and its exact position from the aerial on the feeding coax for minimum VSWR resulted in a lot of bits of coax on the floor of the anechoic chamber! Unfortunately , a 3-stub tuner was not available, but recourse to a Smith Chart did narrow the search range.

Al.

[Radio Wrangler]

You can speed up stub matching and general mucking about with transmission lines with a bit of computation. there are several free programmes out there.

The one I mostly use is LLSMITH written by Lance Lascari. It's downloadable from his site www.rfdude.com.

David

[G0HZU_JMR]

With conjugate matching does the length of the transmission line come into play or is the length immaterial ?

Sometimes you can use a length of transmission line as a matching tool. For example, a useful (but limited) method of matching involves using a series section of transmission line that is not Zo.

eg for a 50R system you could use a series section of 75R coax to match a non 50R load by spacing the 75R section away from this load with a short section of 50R coax. Hope that makes sense.

It's good to play around with this method on a smith chart to see how it works. However it is limited in terms of the range of impedances it can match and this method isn't always an option. But for impedances that lie fairly close to 50R you can often use this method.

eg you could design a matching network to match to 75R or design another for 33R or design another for a complex impedance like 88, j23 using a section of series 75R coax placed at the correct offset from the load.

eg in the example image below I've matched 88, j23 to 50R by placing a section of 75R coax that is 0.147 wavelength long in series with a 0.079 wavelength long piece of 50R coax.
Note: If I calculate the lengths rather than use the smith chart then I get a slightly different length for the 75R section but it is close enough in the smith chart tool.

[ms660]

I found Greg Durgin's video lectures helpful, TDT 01 to TDT 10 is about DC on a transmission line, DC pulse etc.

THT 01 to THT 05 is about sine waves on a transmission line.

Links to the start of both series are below, the rest can be found on utube:

TDT 01

https://www.youtube.com/watch?v=7Oz1sazpekM

THT 01

https://www.youtube.com/watch?v=nmONJpCVq_0


Lawrence.

[G0HZU_JMR]

Another thing that helps with TLine theory is to look at the current and voltage along the line.

This is easiest seen if the line is made from microstrip and you leave the far end open. This will cause a high VSWR and a high ISWR.

Then with a VNA or spectrum analyser + tracking gen you can probe along the line using simple homemade E and H field probes and see the places where there are standing peaks and nulls for each probe type vs frequency.

This is also a good way to test the health + performance of your E and H near field probes because you can test the H field probe for a 90degree rotation null and also a phase inversion for a 180degree rotation of the probe. The E field probe I use is just a bit of semi rigid cable with about 6mm of the centre conductor sticking out of the end of the cable. So it is pretty reliable in operation upt to a few GHz.

I use a big piece of single sided FR4 PCB for the probe tests and I make the topside microstrip transmission line from a strip of sticky backed copper tape. The H field probe is a bit more complex and the above test platform can show the upper limit of the H field probe wrt its size.

Although I really just use this setup for probe checking it is quite interesting to see the various current and voltage nulls and peaks along the transmission line and see that they agree with theory