0
\$\begingroup\$

I would like to find the impedance of a transmission line, which in my case happens to be conductive spray paint on paper.

However, I want to do this simply with a single DAC (TX) and ADC (RX) pair that can send arbitrary pulses/CW waves etc. (no VNA or multimeter involved here)

My idea is to place the TX and RX close to each other. Send a pulse, measure the reflection, determine the reflection coefficient, then calculate the impedance of the transmission line.

Does this method make sense? If not what are possible problems I might face? I'm wondering if I would even be able to see the reflection, or if it might be too small.

\$\endgroup\$
2
  • \$\begingroup\$ You have to use the frequency of the signal you are going to use in the final application. A pulse only works in theory as you will face considerable dispersion because of non-linear behaviour (simply said: a non-constant epsilon of the paper.) \$\endgroup\$
    – Janka
    Aug 17 '17 at 19:53
  • \$\begingroup\$ 1st define coefficients of physical line for L, R, C. if you understand how per unit length by aspect ratio,gap and resistance with d constant for paper, it will make more sense. So this assumes microstrip or stripline, otherwise waste of time. \$\endgroup\$ Aug 17 '17 at 20:51
4
\$\begingroup\$

Does this method make sense?

Yes. This is called time-domain reflectometry (TDR). You can buy attachments for certain oscilloscopes that do this, or dedicated TDR instruments.

If not what are possible problems I might face?

In order to get a good result (one that can distinguish important features in the transmission structure), your DAC (or other kind of function generator) must generate a very fast rising edge. On the order of 100 ps is probably a good starting point, unless you have a very long transmission line (10's of meters, at least) and are only interested in gross structures of the line (is it terminated with a short or an open, for example).

Similarly your ADC must have fairly high bandwidth (100's of MHz at least, GHz preferred) and corresponding sampling rate for the results to be informative.

I'm wondering if I would even be able to see the reflection, or if it might be too small.

You will be able to tell the difference between short and open termination even with a very simple set-up. Measuring, for example, the propagation delay through the line will require more care, and the precision of the result will depend on the bandwidth of your measurement system.

\$\endgroup\$
1
\$\begingroup\$

It's OK as an idea. Cable radars have been in use tens of years.

You can go with much lower spec equipment if you use sinewaves. Connect an easy to adjust rf sinewave generator to the beginning of your test line and put an oscilloscope in parallel with the generator.

You must use frequency so high that your test line surely is longer than half a wavelength. Assume at first propagation velocity 300 m/us. That's the theoretical maximum. Probably the real velocity is much smaller.

Let's assume your test line is 1,5 meters long. Then you are fine with 100 MHz generator.

Have at first a short circuit at the end of the line. If your line has reasonable low losses, you should notice that the voltage at the beginning of the line depends on the frequency. This is because the just produced wave and the reflection have different phase angles. Find the smallest frequency that gives the voltage minimum in the beginnig of the line. On that frequency your line is exactly a half wavelength long. That gives to you the propagation velocity. The result is not accurate until your signal generator is matched to the line ompedance, but that's not your problrm in the question, you wanted the impedance.

If you leave the end open, the same voltage minimum at the beginning should occur at about 50% smaller frequency.

Now try different load resistors (=preferably mass resistors with wires shorter than 2% of your test line, they are not inductive) until you find the one that kills the reflection. That's the line impedance. When you have found it, you will not see any specific frequency that gives a voltage minimum at the beginning of the line.

Instead of trying separate resistors, you also can try a small trimpot that has a flat resistive part.

When you have found the line impedance, you can measure the propagation velocity with sinewaves. Only add the needed extra parallel or serial resistor to make the beginning of the line matched and find the frequency where your line is half a wavelength long.

NOTE: You must use a high z probe in the oscilloscope to prevent it to disturb the measurement. The cable that connect your signal generator to the test line must be only few centimeters long if you operate at 100 Mhz maximum.

\$\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.