Hello

I have an Agilent E4400B Signal Generator. I noticed in the output recently that there are harmonics of the set frequency showing on a spectrum analyzer at lower power levels. I thought that harmonics should be filtered out from the output of this machine.

Is this true? If the instrument is malfunctioning here, what might be the cause and repair?

Richard

 

"Harmonics (≤+4 dBm (≤+7.5 dBm, Option UNB) output level) < –30 dBc(typical below 1 GHz)"

If they are more than that, it's out of spec.of pure tone, no modulation.

 

I thought that harmonics should be filtered out from the output of this machine.

They are filtered. You provide no values for us to determine if they are out of spec. There is no such thing as a filter that can remove all harmonics. What were you expecting?

If the instrument is malfunctioning here, what might be the cause and repair?

I really doubt that the instrument is malfunctioning, however, without any context, anything I say here is just guessing.

A signal with harmonics below 30 dBc is classified as "clean" for most applications. Viewing on an oscilloscope, -30dBc will look pure. Viewed on a spectrum analyzer, virtually no signal will look clean, harmonic wise, unless it is a DDS (direct digital synthesis) signal, which can have very low (if any at all) harmonic content due to the way the signal is constructed.

 

Thanks for the help! I am still learning about test equipment, so was not familiar with how harmonics are handled. A friend who is an electronics engineer was visiting and noticed a harmonic. He commented that this did not look right and should be filtered. So, I started checking this out.

To respond to the first reply: So, the specification says that the harmonics should be less than or equal to +4 dBm as compared to the primary signal. Is that how to read the spec? I am not familiar with measuring dBc yet.

Here are some data: For an unmodulated output from the E4400B at 500 MHz set at -10 dBm, the primary signal measures about -12 dBm and the harmonic at 1 GHz measures about -50 dBm on a spectrum analyzer. So that is well within the specification, correct?

 

That's 38 dBc. I don't know what the specs should be - RTM. 38dBc ain't all that bad. Your EE friend should probably change careers to baking cookies or something.

 

Yes, I started wondering about his expertise in this area...

What is "RTM"?

Thanks for giving the dBc. I learned a new metric!

 

RTM = Read the Manual

 

Thanks, yes I did look at the manual. Oddly, the specs are in a separate document that I had to search for.

Richard

 

Thanks, yes I did look at the manual. Oddly, the specs are in a separate document that I had to search for.

Richard

Well? Is it in spec?

 

Yes. "Harmonics (≤+4 dBm (≤+7.5 dBm, Option UNB) output level) < –30 dBc(typical below 1 GHz)"