0
\$\begingroup\$

I am performing some RF tests for 802.11 and measuring path loss of the RF test setup, it has a mix of PSK & OFDM signals. I see the norm is to use a Continuous Wave signal to perform RF path loss, from signal generator to power meter. But when testing an actual signal which can have channel bandwidth of 20 or 40 or 80 MHz, which is a collection of subcarriers or multiple bundled frequencies. Shouldn't the path loss signal be an actual modulated signal instead of just Continuous Wave signal to perform RF tests on high channel bandwidth modulated signals?

\$\endgroup\$
4
  • \$\begingroup\$ Sorry, you mention cable - is this a cable path-loss test? And, are you asking what is the best signal type to use? \$\endgroup\$
    – Andy aka
    Jun 23 '17 at 17:46
  • \$\begingroup\$ Yes that correct. This is for conducted RF testing. I usually see Continuous Wave signal used for performing Cable Loss Calibration to perform any RF tests. If the RF tests involve a 40 MHz signal, would still using the Cable Path Loss values obtained using a Continuous Wave signal stay valid to perform, say TX Power Tests. \$\endgroup\$
    – eecs
    Jun 23 '17 at 17:48
  • 1
    \$\begingroup\$ Usually the expectation is that the bandwidth is a small fraction of the center frequency and that changes in behavior over it are minor. If one of those is not true you may need a more involved test. It's probably going to be easier to do CW measurements at a number of frequencies (or with a slow sweep) than to measure actual modulated signals. \$\endgroup\$ Jun 23 '17 at 18:01
  • \$\begingroup\$ Isn't center frequency a small fraction of the bandwidth? Is it OK to comment that using a modulated signal to characterize the cable loss at a test station would be more accurate, but using a CW signal would be just OK if not accurate, to test a modulated signal? \$\endgroup\$
    – eecs
    Jun 25 '17 at 19:33
1
\$\begingroup\$

If you view the cable as a transmission line, you can model all of its characteristics using standard transmission line formulas. This suggests that single frequency tests at the spectrum limits should provide most of the needed information.

Take care however, that your tests also examine group delays given the wide spectrum of the application. With careful phase change measurements across the application spectrum, you will be able to predict any possible group delay effects.

\$\endgroup\$
1
  • \$\begingroup\$ This would mean, modelling the losses by the transmission line/cable could be done at the chosen frequency within the bandwidth, which can be used as reference to test your actual modulated signal which contains a complex signal of multiple CW signals within that channel? \$\endgroup\$
    – eecs
    Jun 25 '17 at 19:35

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.