1
\$\begingroup\$

I would like to simulate RF transmission lines in how they propagate, and how impedance mismatches at the end of the transmission line cause reflections and power transfer does not work.

It would be nice if I could send an audio frequency through the "transmission line" and watch the output, hooking up different resistors until it is matched to the characteristic impedance of the line, and max power transfer occurs.

A fun hands-on simulation project built from electronic components would help me grasp a better understanding of radio wave propagation through transmission lines, and would be a fun project for my electronic club in high school, attempting to understand RF a little more.

Any ideas on a circuit? It isn't really practical to make a big line multipule wavelengths of the Audio Frequency and I am thinking maybe using large inductors and capacitor arrays, possibly a circuit similar to the picture below.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ At audio? Forget this madness - for a start the characteristic impedance at audio is more complex than just 50 ohm or 75 ohm and this means you realistically learn not much if you are hoping to emulate/simulate RF. At audio you do realize that you will be simulating several (if not tens) of km in order to get the types of effect you want to generate. Use RF, use 50m coax and get a signal generator capable of driving the line. \$\endgroup\$ – Andy aka Sep 8 '14 at 15:07
  • \$\begingroup\$ Does your electronics club have a oscilloscope? \$\endgroup\$ – Olin Lathrop Sep 8 '14 at 15:08
  • \$\begingroup\$ I have a 25Mhz one but yeah i think he has a analog 20mhz one somewhere. The probelem is I don't know if we have a way of generating RF other than my 144/440 mhz ham radio which is way too high freq \$\endgroup\$ – skyler Sep 8 '14 at 20:06
4
\$\begingroup\$

You can simulate a transmission line with lots of series inductors and parallel capacitors, but that is a lot of trouble and I'm not sure will give you the insight you really want.

It's actually not that hard to see obvious transmission line effects at high frequencies if you have a oscilloscope. If your electronics club doesn't have one, that would be a great piece of equipment to ask for. In the mean time, find some company in the area that does electronics and is willing to help out high school students. I suspect most of them would be happy to help if asked.

We did this in a lab in college, and I remember being surprised how clear and obvious the results were. Get a spool of some cable. Coax would be great, but 100 feet of ethernet twisted pair cable will work well too. Most likely whoever handles the network in your school has a spool of "CAT5" or similar cable they can let you borrow.

Use a signal generator or a simple circuit with a digital output that makes a square wave. Connect ground and this square wave to one end of a twisted pair, and get access to the other end of the same twisted pair.

First just look at the signal as injected at the transmitting end with and without about a 50 Ω resistor in series. Especially with the resistor, you should be able to see stair steps as the reflection from the other end gets back to the transmitting end. Now you can put a 50 Ω resistor accross the far end and see the effect it has on the transmitting end. Also look a the signal at the far end with and without each of the resistors in place. I think you'll see clear artifacts from the reflections, and how things quiet down but are half the voltage with the resistors in place.

Other things to do is to adjust the resistors for minimal ringing, which means finding the characteristic impedance of the line. It would also be a interesting exercise to measure the total propagation time thru the cable, measure the length of cable, and compute the actual propagation speed on that cable. You may be surprised what you find.

\$\endgroup\$
3
\$\begingroup\$

You may want to make a mechanical model instead. That would make it much easier to get something that works at frequencies so low that you can see the effect with your bare eyes.

Like this famous video from Bell Labs, where Dr. J.N. Shive explains waves so a layman can understand it :-)

\$\endgroup\$
0
\$\begingroup\$

That's an interesting idea. (I'd get rid of the R's) I had this idea of putting together different lengths of (say) 50 ohm and 75 ohm coax, the end goal to make 1/4 wavelength impedance matching networks. (And other structures.) The problem is that for reasonable lengths of coax you need fairly high frequencies. It would be nice to get something to work at lower frequencies (say 10-20 MHz.) such as are available with inexpensive signal generators. One issue I wonder about with the lumped element approach is how many elements do you need? You've got a discrete vs continuous structure, at some point the discreteness will show up. (This is perhaps just a long comment and I'll delete it if/when someone complains.)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.