Please forgive the long question. I have split it into 2 parts.

I am very confused about how harmonic suppression is measured in RF system (in my case, ham radios). I have picked up a couple of the Baofeng radios: The UV-5R+ and the GT-3, and I am keen to understand how bad the harmonics are whist using them.

I have read the following 2 blog articles that do measurements on these devices: http://hamgear.wordpress.com/2013/01/30/baofeng-uv-5r-spectrum-analysis-revisited/ http://hamgear.wordpress.com/2014/01/24/review-baofeng-gt-3/

Also, referring to the EU standard EN-301-166-1 Section 7.4.3:

"The power of any spurious emission, occurring more than 2,5 × CSP from the centre of the channel on which the transmitter is intended to operate, shall not exceed [0,25 µW (-36,0 dBm)]"

Part 1

When doing these experiments, you configure a certain attenuation on the Spectrum Analyser. However, by applying different levels of attenuation, doesn't that mean that the measured harmonics will vary?? Does this mean that to obtain the absolute dBm value coming out of the radio, you need to add the attenuation figure onto any measure figures? For example, if you are applying 20dB attenuation, and you measure -50dBm on your analyser, does this mean that your radio is actually putting out -30dBm?? Or have I got this all wrong?? Also, what does reference dB mean? Is that simply the top line on the graph?

Part 2

I don't have a spectrum analyser, but I do have a very cheap RTL-SDR dongle (DVB-T stick hacked to act as a wideband receiver). Using the program SDR-Sharp, this gives about 50dB dynamic range (i.e. the amount I can see on the FFT display). The FFT display in SDR-Shaper has a y-axis of dB (not dBm, unfortunately). With the absence of any calibration equipment, I had to make do with relative measurements, rather than absolute values...

Using this dongle, an experiment I did, was set my radio to a certain VHF frequency (144.500Mhz), and by way of adjusting the SDR gain and moving the radio closer/further away from the USB dongle, made sure that the main frequency was just touching 0dB on the SDR-Sharp FFT display. I then went on to measure the harmonics, up to the 7th one.

The values I got from the UV-5R+ radio, from the 1st harmonic (main frequency) up to the 7th harmonic were as follows (in dB): 0, -55, -47, -45, -48, -52, -54

Have I measured anything useful here? And can I relate these figures to the EU standard at all?

Any help is appreciated.

Many Thanks


1 Answer 1


A spectrum analyzer will automatically add the attenuator setting to the measured absolute power.

The units you need to know are

  • dB -- power relative to some reference value
  • dBm -- power relative to 1mW
  • dBc -- power relative to carrier power

The problem with using the RTL-SDR is that these devices are not designed for absolute power measurements, but rather for extracting digital data. As such, they do not have any compensation for their own frequency response, which is overlaid with the measurement data you have.

The main issue, however is that you are measuring in dBc, by normalizing your carrier power to 0dB and then looking at the harmonics. dBc give you a good estimate of the quality of your amplifier, but for regulatory purposes, only absolute power in dBm is relevant, which requires you to use a device that has been calibrated.

The attenuation setting on the spectrum analyzer needs to be chosen so the entire RF range is overload free, as overloads in the RF frontend would produce harmonics in the first mixer stage, before the selection filter.

Some analyzers include a high pass filter option, where higher harmonics are measured by replacing the attenuator with a high pass filter, allowing for better resolution while still remaining overload free; also, analyzers often provide a dedicated harmonics measurement that automatically searches the base frequency, adjusts the attenuator and reports back absolute and relative values with error margins.

  • \$\begingroup\$ Thanks for this. So looking at the hamgear blog posts, if he measures -47dBm using his Rigol analyser, then that can be compared to the regulatory -36dBm? \$\endgroup\$
    – Jonny
    Commented Aug 25, 2014 at 14:36
  • \$\begingroup\$ Yes, as long as the transmission was comparable to a real world scenario. The problem I have with this blog article is that they mix up the units. There is no such thing as "56dBm down", it is -56dBc. \$\endgroup\$ Commented Aug 25, 2014 at 14:44
  • \$\begingroup\$ Is is definitely dBc? The marker on the Rigol says dBm... Is it possible that the main carrier is higher than 0dB and simply "off the chart"? \$\endgroup\$
    – Jonny
    Commented Aug 25, 2014 at 14:51
  • \$\begingroup\$ The analyzer won't display dBc unless told which is the reference carrier. It is possible for the carrier to be off the chart, but unlikely as attenuation is usually automatically set from the reference level (i.e. so that a signal reaching the top of the chart will not cause an overload). \$\endgroup\$ Commented Aug 25, 2014 at 14:53
  • \$\begingroup\$ Hmm ok. But if the attenuation setting makes the main carrier 0db on the chart, how is the measured harmonic -56dBm? This goes back to my original question of does the attenuation setting affect the measured value... \$\endgroup\$
    – Jonny
    Commented Aug 25, 2014 at 14:57

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.