|
Websites Found an interesting website? Post the details here and share it with the rest of us. Please stick to websites that are in some way related to our hobby/interest. |
|
Thread Tools |
21st Nov 2014, 10:51 am | #1 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Transmission Line
Here is a link to an old Tektronix film about transmission lines, it compliments in an electrical sense the film on Wave Motion which was posted in another Website thread yesterday, It's a good film for anyone trying to grasp the basics.
http://www.youtube.com/watch?v=I9m2w4DgeVk Lawrence. |
21st Nov 2014, 1:48 pm | #2 |
Rest in Peace
Join Date: Jun 2006
Location: Chard, South Somerset, UK.
Posts: 7,457
|
Re: Transmission Line
Interesting lecture in transmission lines. With the other lecture that you mention, the two together provide a good explanation of the basics. As far as the Tektronix lecture was concerned, it did examine the three cases of matched load and mis-matched load with resistive termination not equal to Zo: fair enough. However, I was expecting an extension of mis-matching where the load was of the form R ± jX. Perhaps there is a part 2 of that lecture covering those situations.
Interesting that the sample of coax line used in the demonstration was 93 Ω. I wonder why the more common 50 or 75 Ω line was not chosen. Finally, the technique for determining the Zo of a coaxial line - which I have used with considerable success - is with the aid of an L/C measuring instrument. The technique is as follows: 1. Open-circuit the far end of the sample of line: measure the capacitance at the 'sending' end. Let this be C Farads. 2. Short-circuit the far end of the sample of line: measure the inductance at the 'sending' end. Let this be L Henries. 3. Then Zo = √(L/C) Ω. Experience has shown that the longer the test sample of coax, the greater the accuracy of the result, but in actual practice, extreme accuracy is not required, since coax cable comes in known values of Zo, so a calculated result in the range 45 → 55 Ω indicates Zo = 50 Ω, as opposed to any other value. Al. / Nov. 21, '14 // |
22nd Nov 2014, 12:51 am | #3 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: Transmission Line
Funnily enough, I have spent most of this week explaining transmission line theory and effects to a class at work. It's a surprisingly hard topic to explain to people who have little electronic engineering background. Of course, there is much more to broadcast engineering than electronics - so much today is about IT and "systems", so we don't necessarily need to recruit people with in-depth electronics knowledge that we once did, but they still need to know the basics.
I've seen that video before, and I have seriously considered playing it in class. It would be a 20 minute break from me, and as everyone has a different learning style, it really doesn't hurt to hear more than one version of the story from different people. The "retro" style is very charming - I love the bow tie and row of pens in the shirt pocket. And I haven't seen a 93Ω cable for a good few years The practical demonstration is very good, and I do a very similar thing myself. And then we see it in several other contexts - e.g. with video signals, with digital audio signals, with square waves. When it comes to measuring characteristic impedance, I agree that the bridge method is good. I also agree that long cables are better - I've just tried it with a 1.5 metre cable, and found that it's really tricky to get the bridge to balance at the low values of inductance (about 0.5µH). To be fair, I've not used the old Marconi TF2700 for a while, and it feels like it's perhaps time for a bit of contact cleaner in some of the potentiometers Measuring C on the bridge was a bit easier, and I could sensibly verify that with a Peak Atlas LCR meter and a DVM. I got 203pF. In fact, using a bridge, the only way I could get a sensible reading was to calculate what L should be (knowing Z, L=Z²C), and set the dial in approximately the right place to start. Then a delicate dance with the Q control finally got it to balance. As I say, it's time for a service... But, once sorted, I got a value of 50.8Ω. Not bad Out of interest, I tried an alternative method. I set up the pulse experiment, but used a 200Ω variable resistor as a terminator. Once the reflection had been minimised, I simply measured the value of resistance that I'd dialled up. That gave 53Ω. Not as close as the bridge, but a *lot* quicker and easier than fiddling about with the LCR bridge. Certainly close enough to distinguish between 50Ω and 75Ω. Of course, you do need a pulse generator of some sort, but it does work with a square wave at a high-ish frequency. I used an old Philips PM5715 for this. With a 1.5m cable, the reflection was too fast, and not properly separated from the pulse. Adding another 1.5m cable gave a separation of about 33ns. Which gives a VF of 0.61, which is in the right ballpark for RG58. The next obvious practical would be to use this knowledge to measure an unknown length of the cable. Anyway, it's always fun to mess about with this stuff once in a while - very relaxing after a busy week... |
22nd Nov 2014, 2:52 am | #4 |
Nonode
Join Date: Oct 2014
Location: West Midlands, UK.
Posts: 2,181
|
Re: Transmission Line
mhennessy- perhaps another example of the effects of standing waves on transmission theory is a practical demonstration using a signal generator and a measuring set. Connect together with a coax cable with frequency around 4 Mhz and level of Zero dB , through a T joiner. Then add lengths of cable to the non-terminated branch and watch the received level drop.
|
22nd Nov 2014, 9:00 am | #5 |
Nonode
Join Date: Aug 2007
Location: South Bradford, West Yorkshire, UK.
Posts: 2,571
|
Re: Transmission Line
Back when I was doing A level physics we were introduced to the transmission line and one Wednesday afternoon I decided to try simulating one with inductors and capacitors. I raided the radio club cupboard for as many large inductors as I could find. I set them up as the classic LC circuit with 470uF capacitors. We had 8 small Telequipment scopes so I set them all up with one scope across each capacitor. A 6v battery and switch provided the input pulse and it was fascinating to watch the progression of the pulse down the line. I think it took about 2 seconds to propagate from one end to the other! It was demonstrated in class the following day.
Keith |
22nd Nov 2014, 10:31 am | #6 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
On the subject of transmission lines and reflections, here's an interesting series of articles by the late Walter Maxwell:
http://homepages.ipact.nl/~pa1are/tuner/reflections.pdf Lawrence. |
22nd Nov 2014, 11:08 am | #7 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
Walt Maxwell used to put in frequent appearances on rec.radio.amateur groups. Much of what he wrote was OK, but one aspect was not. There was a long standing argument about whether there was a matched condition between a power amplifier valve and the transmission line from the transmitter onwards.
The people with the opposite view said that the impedance presented by the line to the transmitter was transformed in some form of network or transformer in order to present the right impedance to the output valve or transistor in order to extract the planned amount of power. This could sometimes be close to a true match, but it wasn't always, nor did it have to be. Maxwell cited the maximum power transfer theorem and more or less said 'QED' His opponents said that most usually you did not want the maximum theoretical power which could be extracted from a power source. Efficiency suffers and things break. THe leading opponent was Warren Bruene of Collins... a very seasoned designer with plenty of paper to his name as well. You can imagine how it went. Neither side of the argument would change their views. It wasn't nice at times. As for me, I'm a design engineer. I was designing signal generators, where I did need to create carefully controlled output impedance and I was aware of what it took to do it and what the costs in wasted power were. At sig gen levels it wasn't going to affect the electricity bill, but the controlled impedance was necessary for its application. I was also designing transmitters where I wanted to get every watt I could from a given device and get it radiated by the antenna. I sided with Bruene. The maths and the logic weren't hard. I couldn't see why Maxwell and his followers couldn't see the flaw in wanting true matching all the time. I would advise caution in reading Walt Maxwell's material on transmission lines. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Nov 2014, 11:13 am | #8 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
I forgot one bit.
In a class-C or class-B transmitter, the output device is being driven into cut-off over part of the cycle. It's effective output impedance is changing wildly all the time. No network of fixed, linear components and transmission lines can match this, anyway. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Nov 2014, 11:49 am | #9 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
Yes, there was obviously some friction going on:
http://www.w1npp.org/ARES/REFLEC~1/R...2/BRUENE~1.PDF I'm not sure what to make of it all, I just like reading this stuff. Lawrence. |
22nd Nov 2014, 12:27 pm | #10 | |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: Transmission Line
Quote:
Of course, this is PAL, at 4.43MHz... |
|
22nd Nov 2014, 12:30 pm | #11 | |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: Transmission Line
Quote:
|
|
22nd Nov 2014, 12:32 pm | #12 |
Dekatron
Join Date: Jul 2007
Location: Evesham, Worcestershire, UK.
Posts: 4,241
|
Re: Transmission Line
I did watch a bit more on Youtube last night, and while I haven't seen them all yet, this guy has a series of videos demonstrating some of this stuff:
https://www.youtube.com/watch?v=-eAn...teQPA&index=14 He certainly has a lot of kit! Tek fans are in for a treat |
22nd Nov 2014, 3:33 pm | #13 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
Yes, that was from the time things were getting rather nasty. I was watching from afar (Scotland) and chipping in the occasional comment. If you want to find out about Bruene, he was a pleasant, mild mannered man. He has a track record of a great many successful and high-performance transmitters as well as transmission line instrumentation.
http://www.collinsradio.org/warren-bruene-1916-2013/ Maxwell, too, had done a lot but he'd made an unnecessary assumption that would not be true under most conditions and he stuck to it. He did not argue his case graciously and he did not allow the possibility that he could have got something wrong. The Sabin mentioned is Bill Sabin at Collins, who was one of the compilers/editors/authors of Sabin and Schoenike "Single Sideband Circuits and Systems" Having watched the whole show and having a serious involvement in RF design, I was left with the feeling that I could trust the writings of Warren Bruene and Bill Sabin, but I'd be careful with those of Walt Maxwell. Maxwell was certainly right about a lot of things, but he certainly had a bee in his bonnet over this matter. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Nov 2014, 4:19 pm | #14 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
Do you know what was the original disagreement that got Bruene and Maxwell going ?
Lawrence. |
22nd Nov 2014, 5:36 pm | #15 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
As I remember it, there had been a discussion going on about the impedance transformation used to couple power from the final active device of a transmitter into the antenna feeder. It was being explained to some beginners in the field.
Maxwell said that the transformation had best present a matched load to the anode/collector/drain of the power device. Others disagreed, and said that the optimum presented load impedance wasn't necessarily a conjugate match. People then started taking sides and positions hardened. It wasn't pretty. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Nov 2014, 5:40 pm | #16 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
Thanks for that.
Lawrence. |
22nd Nov 2014, 8:17 pm | #17 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
Here's a question, the answer to which I'm not sure of.
I watched one of the utube video's showing how to find out the characteristic impedance of a transmission line when the lines impedance initially isn't known. It used TDR with a variable load resistance at the far end of the line, I understand that and how it works, but what about the output impedance of the TDR generator compared to the initially unknown impedance of the line it's feeding into, in other words will an impedance mismatch occur and if so what effect will it have on the measurement and what's to do about it ? Lawrence. Last edited by ms660; 22nd Nov 2014 at 8:18 pm. Reason: typo |
22nd Nov 2014, 8:55 pm | #18 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
Easy!
You can have a monumental mismatch from the TDR to the line and you get a big spike showing the reflection from that mismatch. So many nanoseconds/microseconds later you see a quite separate spike from the mismatch at the end of the line. Fit a variable resistor to the end of the line and you will see the end of line spike change as you trim it. You won't see any change of the start of line spike. There is still the same big echo there. It takes time for the signal to reach the end of the line and for the reflection to make it back. You can work out the impedance from the screen on some TDRs, but it's easier to use a resistor. Let's play this another way.... You've built something, and you have various connectors, cables and then a length of track on a PCB. Something is wrong with the impedance. You can work at it with network analysers and smith charts etc, but if you connect the TDR instead, you can slide a bit of metal on a plastic handle (those plastic trimmer tools with atiny steel blade are great for this) along the track etc and you'll see its tiny reflection. Move it to put its reflection in coincidence with that of your problem, and your little prod is in the right place! Simples! (Once you've seen a TDR used) David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Nov 2014, 9:07 pm | #19 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: Transmission Line
To clarify, does that mean that a mismatch between the pulse generator and the line under test doesn't make any difference to the test ?
Lawrence. |
22nd Nov 2014, 9:25 pm | #20 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
|
Re: Transmission Line
It means that the test shows both mismatches completely separately and the generator end mismatch doesn't interfere with your ability to see the echo (if any) from the far end.
It also allows you to work on the generator end match if you need to without being bothered by whatever is at the far end. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |