1
\$\begingroup\$

This is regarding some Radiated Emissions testing unintentional radiator:

enter image description here

Critical frequency is at 314.986 MHz

  • Polarization is Horizontal

  • Peak detector is 53 dBμV/m

  • Quasi-Peak detector is 56.68 dBμV/m

Now my question is how is it possible for the quasi-peak detector to be higher than the peak detector?

\$\endgroup\$

1 Answer 1

2
\$\begingroup\$

Could be that the dwell time for peak detector wasn't as long as for QP, and whether due to AM or FM at/around the peak, the true peak was simply missed.

In other words, an aliasing artifact, and the peak would be seen to dance up and down a bit from sweep to sweep.

Normally, QP is done separately, testing point frequencies or narrow ranges, as it's slow to perform. This may vary with test equipment used (I think some can read QP as fast as anything else these days?), but this has been my experience with the tools at the local lab. The longer dwell time may be more likely to catch the true peak at the frequency.


As far as I know, the signal path, receiver, and detector are identical between cases, except for the time constant following the detector, which for QP has a somewhat average-reading effect for fast changes (near the RBW), and peak-reading for most rates of change. The characteristic at very low frequencies (~Hz) is not defined, and would depend on the display mode (peak hold or average), or again would be seen to fluctuate between acquisitions.

The detector reads magnitude within the receiver bandwidth (RBW), so reads AM directly, or FM by slope detection when the modulating signal is within RBW and some of the signal spreads outside the RBW. (FM for signal BW > RBW reads as separate resolved peaks.)

It's also important whether the peak detector behaves as the name suggests (detecting any rate of peak within the RBW, and holding that peak for the duration of acquisition at that frequency), which also depends on instrument design. One would have to check the manual to be sure, or even ask the manufacturer.

Note that these terms originate with frequency-sweep type spectrum analyzers, where, to be absolutely precise, the acquisition is not done by dwelling on each frequency point, but sweeping across the whole span; the LO signal is a long chirp, not a staircase. For many situations, this makes little if any difference against say an FT, but the readings differ most for signals with very narrow peaks/troughs (basically, that give apparent modulation at a time constant comparable to the sweep or acquisition rate). DSAs (Digital Signal Analyzer, i.e. ADC + FFT) may give different results, as they use (or can use) a holistic or simultaneous acquisition (everything at once in one big snapshot), rather than sweeping frequency in a chirp. DSAs may use algorithms to simulate peak or QP detection (a raw magnitude FT gives the average over the acquisition period).

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for your reply, I need some time to process all the information, and I will follow up if I have a question. \$\endgroup\$ Dec 8, 2023 at 15:45

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.