I have an application where I am characterising the output of a radar transceiver. Because of the high peak power, I am using a coupler and some attenuator padding to stay within my spectrum analyser maximum input power level. The problem is that as my pulses are not nice rectangular pulses I get very different results when define pulse width as teff (as specified in page 5 of Application Note 150-2) compared to when using pulse width given by FWHM.

The two definitions give me -5.87dBm and -4.27dBm peak power respectively. This may not seem like much, but when I add the 79dB (due to the coupler and attenuator padding), the two definitions give me peak power of 20.56KW and 29.71KW respectively, which is now very significant difference!

I attach a file showing the pulse shapes.

Any help appreciated

The two definitions give me -5.87dBm and -4.27dBm peak power respectively. This may not seem like much, but when I add the 79dB (due to the coupler and attenuator padding), the two definitions give me peak power of 20.56KW and 29.71KW respectively, which is now very significant difference!

I attach a file showing the pulse shapes.

Any help appreciated

These are definitely not nice rectangular pulses! However, the answer to your question lies in the definition of Teff. In Figure 5, page 6, of Application Note 150-2, the definition of Teff is given as:

Teff = Width of Rectangular Pulse of same height and area as pulse applied to analyzer

If you can calculate the area under the pulse envelope (in Volt-seconds, or Watt-seconds), you would then divide the area by the peak value and the result would be Teff.

For example, if the waveform were a triangle wave, and the width at the base were 100ns and height at the peak was 1V, the area would be 5x10E-8 V-s (one-half the base times the height). Divide this by 1V to get Teff = 50ns.

In your situation, of course, the calculation of the area under the pulse will take a bit more effort.

Regards -