1
\$\begingroup\$

I want to know, if an antenna which transmits and receives the reflected signal can be modeled as an impedance. For example, the antenna is coupled to a Collpits oscillator and sends out an electromagnetic wave which is reflected by, for simplicity, a wall of a conductor. So it receives something which superimposes the sent signal. Can this configuration be modeled as an impedance?

The idea is to model this configuration as a "kind of lossy transmission line" where a transmission line (tl) has a certain impedance depending on its length.

If so, the impedance varies with the distance of the reflecting wall and tunes the oscillator to a slightly different frequency, which is what I want to find out about. The goal is to have a function of frequency vs. distance.

I can simulate the S-Parameters of the antenna (not done yet), if this is helpful for you to know.

Edit: Today I simulated the antenna and the software returns Z-Values and S-Parameters that vary over exciting-frequency and distance of the reflector. So it shows, that the antenna can be modeled as a complex impedance which varies with distance of the reflector. So we can use this impedance shift to tune the oscillator by putting it into the feedback loop. But: As said above the impedance varies also with frequency. So imagine we change the distance of the reflector and get a different impedance, then the oscillator changes its frequency which also changes the impedance. So we have a self-backcoupling massively non-linear equation and I don't know how to get around this problem. Am I at least on the right track? Is there a way of solving this interrelatedness?

\$\endgroup\$
8
  • 1
    \$\begingroup\$ Yes this should be possible providing the reflector is within less than qtr of a wavelength (estimate). Rule of thumb is that near-field stuff dies out at lambda therefore it should be reasonable at lambda/2 but because there is a return path to take from the re-radiator it should be within qtr lambda (approx) \$\endgroup\$
    – Andy aka
    Commented Jun 10, 2015 at 20:20
  • \$\begingroup\$ I don't know about tuning the oscillator. But I think the antenna aimed at the wall is very similar to a lossy transmission line with some kind of termination at the end. Radar losses are R^4. R^2 on transmit and R^2 after reflection. So the effect can become vanishingly small in a hurry. I don't really get why Andy said the re-radiator should be within qtr lambda, other than maximizing strength of the return. But it does seem like the useful range is from 0 to lambda/4 anyway, as far as phase shift goes. \$\endgroup\$
    – user57037
    Commented Jun 10, 2015 at 21:12
  • \$\begingroup\$ @mkeith I'm saying that if the OP wants to be able to put something close to the antenna in order to alter its reactive impedance then this has to be done within qtr lambda. Beyond that the signal is a bona fide EM wave and reactive impedance modification is out of the window i.e. the reflector needs to operate in the near field. You seem to read the question differently to me. \$\endgroup\$
    – Andy aka
    Commented Jun 10, 2015 at 21:58
  • \$\begingroup\$ Yes. I get your point now. I wasn't thinking of nearfield effects. My point was that the wall will generate a reflection, and that reflection will come back to the antenna and interfere with the signal (creating a phase shift) no matter how far away the wall is. But the effect will be vanishingly small if the wall is not very close. \$\endgroup\$
    – user57037
    Commented Jun 10, 2015 at 22:08
  • \$\begingroup\$ so am I getting it right, nearfield for maximum effect? I think we will be in the nearfield anyway, but thanks. \$\endgroup\$
    – jjstcool
    Commented Jun 12, 2015 at 0:38

1 Answer 1

Reset to default
1
\$\begingroup\$

If you want frequency proportional to range of wall, consider using an FM continuous wave (FMCW) or linear FM homodyne (LFMH) radar. LFMH and FMCW are almost synonyms. Transmit and receive occur (or can occur) simultaneously. The transmit waveform consists of a linear frequency ramp. The TX and RX are mixed, and the difference frequency is selected using a low-pass filter. When there is a single dominant radar target (such as a wall) the difference frequency is basically a sine wave with frequency proportional to range.

This is one of the simplest types of radar.

http://en.wikipedia.org/wiki/Continuous-wave_radar#Modulated_continuous-wave http://demonstrations.wolfram.com/FrequencyModulatedContinuousWaveFMCWRadar/

\$\endgroup\$
2
  • \$\begingroup\$ thanks, but I think we will try to keep it simple and use no more advanced techniques than building a simple sine oscillator if not necessary. I think your ideas go to far for our small project \$\endgroup\$
    – jjstcool
    Commented Jun 12, 2015 at 0:41
  • \$\begingroup\$ Makes sense, if the limitations are not a problem in this case. \$\endgroup\$
    – user57037
    Commented Jun 12, 2015 at 1:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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