Back terminated cable?

[Skywave]

I've recently met the expression "A back-terminated transmission line" and I'm a bit puzzled as just what that means. Yes, I know what a transmission line is and understand the need to terminate it in a resistive load equal to its Zo to eliminate reflections from that load, but what does the added adjective 'back' mean and why is it necessary, as in "back-terminated transmission line" ?

Al.

[Radio Wrangler]

There's a neat trick. I can run a transmission line fed fro a Zo matched source, and NOT terminated at the receiving end... just a high-Z load.

Naturally there is a huge reflection from the receiving end. THe reflected wave returns to the source and is absorbed in the source drive impedance Zo with no further reflection.

Advantage: no 6dB loss of voltage at the receiving end.

Disadvantage: apart from at the end, waveforms are distorted as the forwards wave passes its reflection. So you can't do multi-drop this way.

David

[kalee20]

It refers to driving the transmission line from a source with internal impedance equal to the line's characteristic impedance.

Ideally, the transmission line should be terminated at its remote end by a resistive load equal to the characteristic impedance. Then, there are no reflections, and the source 'thinks' it is just loaded locally by the resistive load. So the source can be a stiff voltage source, a high-impedance current source, or anything in between. But that's an ideal situation.

In practice, what if the load at the remote end isn't quite spot-on? There will be a reflection back to the source end. If the source impedance is not equal to the line's characteristic impedance (such as if the source was a good solid voltage source), then this reflection would bounce back from the source, and head again to the remote end... back and forward, with the amplitudes each reflection being determined by the mismatch at the ends.

However, if the source impedance IS equal to the line's characteristic impedance, then any reflection due to mismatch at the remote end, would be absorbed by the source. So there won't be a continual back and forth of energy - you won't get standing waves set up. Down side is that you lose half the voltage, in the simple method such as feeding a 50 ohm cable from a stiff source, but via a 50 ohm resistor.

[Skywave]

Posts 2 & 3: thank you for those. Everything that has been written in those posts I am familiar with and understand. A transmission line (for most applications) needs to be terminated in a resistive load that matches the Zo of that line. Hence, the need to "terminate a transmission line" (ideally in Zo). But perhaps my question wasn't phrased clearly: it is yet to be answered. I'll try again.

► Why is it necessary to add the word 'back' in order to change the phrase "a terminated transmission line" to "a back-terminated transmission line"?

Aside: So, incidentally, a corollary thus arises: if we can have a 'back' terminated line, then surely we can also have a 'front' terminated line?

Al.

[cmjones01]

A transmission line has two ends, either of which may or may not be terminated. A 'back-terminated' line, also known as a 'source-terminated' line, definitely is correctly terminated at the source, or sending, end of the line. It says nothing about the receiving end. This is important in applications like time-domain reflectometry, in which the properties of the transmission line may be unknown, and may cause all sorts of reflections (which we want to measure). If we know that the source is correctly terminated, any reflections from along the transmission line will ultimately be re-absorbed into the source impedance, so they will only happen once. This is very desirable for measuring those reflections accurately.

Chris

[Skywave]

Thank you, Chris: a clear, concise answer.

So a 'back-terminated' cable is a cable where the impedance of the source which drives that cable equals the cable's Zo. Now I do not wish to sound like an 'arrogant tech. know-it-all' in this matter, but to me, I would have thought that anyone using a transmission line in a typical application would also have the elementary knowledge to realise the need for that source Z to match the cable's Zo, therefore rendering the need for the expression 'back-terminated' to be simply redundant. In conclusion, to me, the use of that phrase is simply 'tech. over-kill'.

Al.

[Radio Wrangler]

You can drive a transmission line from an unterminated source. A voltage or a current source. Heve the receiving end of the line terminated accurately in Zo and the wave that hits the far end is entirely absorbed. There is no reflection back to the source and so it doesn't matter what Z of source is applied to the line, there is no reverse wave to reflect or not. The Zo load at the far end matches the line and the operating impedance down the line says on the centre of the Smith chart. The source looks into the line and sees Zo - no reflection.

A fully terminated line, with Zo attached at both ends is needed where you are playing with signals in both directions.

If your terminations aren't perfect, terminating both ends produces lossy reflections and helps attenuate the Fabry-Perot resonator formed by the spaced reflections.

The single-ended termination is a lot more common than it appears. Most transmitters have output Z which is a lousy match to the usual 50 Ohm cables. But it doesn't matter if the antenna is tuned to an accurate 50 Ohms. There is no reflection back to the transmitter. A directional power meter shows real power going out, nothing coming back therefore 1:1 VSWR if you view it that way. Most transmitters have "50 Ohms" written next to their output, but it means ~Please connect me to..." not "I am".

David

[Nuvistor]

Hi David,
I did not realise that about the transmitter matching, you last sentence.

So would this be the case for a receiving aerial, it’s match to the transmission line does not matter as long as the receiver is a “perfect “ match to the transmission line, appreciate that a perfect match is unlikely in most non commercial radios?

[Thyristor]

This is something I too, struggled with, both at college, and at lectures when I was at BT.

Purely by chance, I found this film some time ago, and its the clearest explanation I've found.

By who else, but Tektronix.

https://www.youtube.com/watch?v=I9m2w4DgeVk

[G8HQP Dave]

Originally Posted by Nuvistor

So would this be the case for a receiving aerial, it’s match to the transmission line does not matter as long as the receiver is a “perfect “ match to the transmission line, appreciate that a perfect match is unlikely in most non commercial radios?

No. You want some sort of reasonable match at the antenna end to get good power transfer from the antenna into the line. You then want a reasonable match at the receiver end to get good power transfer from the line into the receiver - although this is complicated by two issues:
1. at HF it often doesn't matter too much as you will be limited by atmospheric noise anyway (in most cases);
2 at VHF and up you may want a good 'noise match' which may be different from a good impedance match.

Incidentally the point made by Radio Wrangler that transmitters are rarely 50 ohms output impedance has been the cause of no end of arguments on amateur radio websites. Many simply cannot believe this. The fact that in radio and audio some manufacturers persist in saying 'output impedance' when they actually mean 'optimum load impedance' simply adds to the confusion.

[Nuvistor]

G8HQPDave,
Yes I appreciate what your saying and that was how I understood the situation on receive.

Reading the the post by David, Radio Wrangler made me think about the receiving aerial as the “transmitter” into the transmission line with a matched receiver at the other end.

[cmjones01]

Originally Posted by Skywave

I would have thought that anyone using a transmission line in a typical application would also have the elementary knowledge to realise the need for that source Z to match the cable's Zo, therefore rendering the need for the expression 'back-terminated' to be simply redundant. In conclusion, to me, the use of that phrase is simply 'tech. over-kill'.

Sometimes there's a choice of whether to terminate the source or not. For example, using a tunnel diode pulser to drive a fast edge in to a transmission line. If you just connect the tunnel diode straight across the line, you get about a 400mV step, but the source impedance is very low (a couple of ohms) so if the line isn't properly terminated at the other end, you'll get odd reflections bouncing back and forth. You can add a series resistor to terminate the source, thus avoiding the reflections, but then the step is only about 200mV. You want the bigger step, you have to live without the termination.

Correct termination is not always necessary to make good use of a transmission line, and sometimes it's better not to have it. The other answers regarding matching of radio receivers and transmitters are excellent cases in point.

Chris

[Radio Wrangler]

Originally Posted by Nuvistor

Hi David,
I did not realise that about the transmitter matching, you last sentence.

So would this be the case for a receiving aerial, it’s match to the transmission line does not matter as long as the receiver is a “perfect “ match to the transmission line, appreciate that a perfect match is unlikely in most non commercial radios?

Receiving, you still want a rough match at the receiver end of the antenna feeder or else there is no flow of signal power into the receiver and the signal to noise ratio will be terrible. Theoretically a good match would improve things by getting more signal into the receiver.

But in one of those quirks fate is justly famous for, the noisiness of input devices depends on the impedance a signal is presented to them from and the optimum noise figure doesn't correlate with the ideal match (the one which gives the best gain)

So the design of an input stage is an inevitable compromise.... trade a bit less gain for a bit less noise, but the lower gain disguises the noise of the next stage less... oh heck.

Lots of receivers (HF, MF, LF) are on frequencies where man made and cosmic noise dominates. They can and do work with an abysmally bad match and the worsened noise performance isn't noticeable. Frightful liberties are taken and the receivers still work usefully.

In general, expect receiver inputs to be fairly poor matches on all except hyper-critical applications (moonbounce etc.)

David

[kalee20]

Originally Posted by Skywave

Aside: So, incidentally, a corollary thus arises: if we can have a 'back' terminated line, then surely we can also have a 'front' terminated line?

Well, the back-termination is so-called because it is a termination for the returned reflection coming backwards from the mismatched load at the remote end. So the back-termination is actually at the front

Hopefully the sketches may make things clearer... Case 1 is the ideal situation, a transmission line terminated by a load exactly matching its characteristic impedance. It's how David Radio Wrangler said (post #7).

Case 2 is where the match isn't exact, and you get a reflection heading back to source. If the source impedance is not equal to the cable characteristic impedance, then there will be a further reflection.

Case 3 matches the cable at the source end, so that any back-reflection is absorbed by the source.

If the back-termination match isn't exact here either, then there will be a small reflection - but if things are matched pretty nearly, then after a double-reflection (at load and at source), amplitudes may well be low enough to ignore.

[GMB]

Most transmitters have "50 Ohms" written next to their output, but it means ~Please connect me to..." not "I am".

Yes, but I measured one or two examples and was surprised to find that their source impedance was pretty good - not what I had expected.

[dseymo1]

Originally Posted by Thyristor

This is something I too, struggled with, both at college, and at lectures when I was at BT.

Purely by chance, I found this film some time ago, and its the clearest explanation I've found.

By who else, but Tektronix.

https://www.youtube.com/watch?v=I9m2w4DgeVk

I don't suppose you were the only one at BT!
We had an Openreach engineer here the other day investigating a fault, who said that broadband would never work properly over a mix of copper and aluminium cable. When I ventured the suggestion that the impedence mismatch at the junction of the two would probably result in reflections, he looked at me blankly.

[Skywave]

Lots of interesting, informative and relevant posts: thank you.

But I will just add this. During the last 55 years of my life, I have read many books on Radio Theory; studied Radio Communications and Electronics at college (HNC+); have been employed in many companies dealing with R.F. technology (including high-power VHF transmitters) and have pursued VHF Amateur Radio. Until very recently, I had never met the phrase 'a back-terminated transmission line': hence, this thread.
My research suggests that the origin of this phrase is the U.S.A.

Al.

[Skywave]

Originally Posted by cmjones01

Correct termination is not always necessary to make good use of a transmission line and sometimes it's better not to have it.

Yes, indeed. Using a piece of Tx line to form a tunable stub being one example.

Al.

[Thyristor]

Originally Posted by dseymo1

I don't suppose you were the only one at BT!
We had an Openreach engineer here the other day investigating a fault, who said that broadband would never work properly over a mix of copper and aluminium cable. When I ventured the suggestion that the impedence mismatch at the junction of the two would probably result in reflections, he looked at me blankly.

Comparing Openreach and BT, is a bit like comparing apples and oranges !

Back in the day, the 'chosen ones' were sponsored to study up to HND; sent on more advanced courses, etc. and were earmarked to become TOs.
I came second, in an intake of about 40 (the person that came first was a woman) The first 5 or 6 were put forward for sponsorship.

Nowadays, I think Openreach engineers are sent on a networking course.

I had a similar incident that you describe, happen to me at work recently. I said to a (new) engineer "....it won't work, because t'other end isn't terminated...." which was met with a blank look.

its a much maligned, and mis-understood topic.

[Radio Wrangler]

"Transmission lines" is a phrase which strikes fear into a lot of hearts, which is a shame because they are damned useful things and transmission line concepts make a lot of other things clear. Approached from the right direction they are quite understandable and not even difficult. It's just that the standard way of teaching them beginning with the maths not ending with it scares people and teaches them that they should be scared. Same thing goes for Laplace, Fourier and filters.

I\ve usually used the phrases "source terminated" and 'load terminated" for single-terminated lines because their meaning is clearer.

People get sent on courses and if they get taught twaddle then that is the 'knowledge' they go away with. Someone who really understood the subject would fail the end of course exam for not spouting the approved twaddle. rather a lot of people get taught a mumbo jumbo approach to transmission lines. They know to repeat the magic words, but they don't know what they mean and anythng else guts the blank look.

Once people have been taught fear and twaddle, it's an awkward job to get them out of it.

David