Spec: Spurious <-110dB in a 100kHz BW, up to 6GHz.

[myRF]

Hi

I have a spec which states "Maximum spurious signal output of < -110dB in a 100kHz bandwidth, up to 6GHz".

My question is, if I simply measure my DUT output with a spectrum analyser set to resolution bandwidth of 30Hz (which allows me to see below -130dBm), and sweep over 1GHz spans (about 13s sweeptime), will this catch any spurious above -110dBm? Or would I need to sweep over narrower spans?

It seems a silly question to me but a colleague has cast some doubt in my mind by saying that we don't have the kit to measure that low!

Thanks in advance.
myRF

 

[D.A.(Tony)Stewart]

30 Hz won't work for capturing the peak amplitude of wideband spurs, but it will help resolve the exact frequency if it is a problem. If the spur was only 100uS long , the peak level would be suppressed in 30Hz BW. If it was continuous, no error in measurement.

 

[myRF]

30 Hz won't work for capturing the peak amplitude of wideband spurs, but it will help resolve the exact frequency if it is a problem. If the spur was only 100uS long , the peak level would be suppressed in 30Hz BW. If it was continuous, no error in measurement.

Thanks. Would you mind explaining why 30Hz wouldn't work for capturing the peak amplitude of spurs? Is it due to the slow sweep time over such a wide span?

 

[G4BCH]

If the spurs a stationary ie it is a cw source then the narrow bandwidth will not be a problem, but if they are modulated in any way then you will not be able to measure the amplitude accurately.
I assume that you mean -110dBc for the spur level. This is a tall order particularly if the output is wideband, not impossible, just difficult. you will need to pay very careful atention to every thing that can go wrong. MxN mixer products, screenig between circuits, leakage across and between filters and even leakage out of and into interconnect cables.
Your colleague is probably right in that you are unlikely to have the kit to do it. Be prepared to spend a lot of time working out what are instrument spurs and what are actual spurs. Once you get to know the instrument it is not too bad but still takes some time.
Peter

 

[biff44]

you need to set the spectrum analyzer to a 100 KHz resolution bandwidth. period. that is the spec. if they wanted it measured in a 30 hz bandwidth they would have written that as the spec.

If you are having trouble with the noise floor of your spectrum analyzer, filter/null out the carrier, and add a 20 dB amplifier in front of the spectrum analyzer.

 

[D.A.(Tony)Stewart]

30Hz resolution means the peak is attenuation due to the bandwidth of the channel responds like a low passfilter response.

For accurate peak level detection, 100KHz is required for capturing intermittent or signals.

In general, one cannot deviate from specified settings, unless one knows the consequences of the differences and does it for a specific reason. One example is quasi-peak measurements which use a defined audio bandwidth for detecting RF interference