Hi all,

I have a question regqrding noise sources.

I would like to do some measurements of RF devices from MHz up to a few GHz. The best would be of course to have a network analyzer. For the near feature I was considering another option: use a spectrum analyzer that is already available, combined with a broadband noise source. Of course, I will miss the phase information and also I cannot measure reflection. I would greatly appreciate any advice on this. Also, do you know how easy it is to get a noise source at a decent price? Someone warned me that a good noise source can also be quite expensive (though I hope it is not that bad, since I do not need a very flat spectral density and large power for example).

Any tip is greatly appreciated,

Adi

Hello,

Don't mix between the use of Network Analyzer and Spectrum Analyzer. Network Analyzer won't help you to measure noise. Please provide information on what kind of parameters you want to measure?

In the RF world one should have a deep pocket for purchasing test equipment. There are no cheap solutions for getting accurate and reliable test results. Any compromise in the price will come on the expense of quality

_________________
Best regards,

- IR

Hello IR,

Thank you for your reply. I see that I did not manage o explain very well what I wanted to do. The idea is that I want to set up a measurement system and for characterization I want to to measure transmission of various two port components (amplifiers, transmission lines, etc). Of course, I would normally use a network analyzer, which gives both the amplitude and phase of the transmission. The problem is that we do not have a network analyzer. But, since we have a spectrum analyzer, I was thinking of a different solution: use a noise source to inject some broadband noise at one port and the spectrum analyzer to measure the transmitted noise at another port. This measurement would give me the amplitude of the transmission for that two port system (unfotunately I will still miss the phase).
So in the end I would need a noise source to do the measurement. The requirement would be that it generates noise in a broadband (say from a few MHz to a few GHz) and that it has an output impedance which is 50 ohm over this band.



I completely agree with you.

Adi

Hi,

if I understand you right you want to see for an example a filter-curve on your spectrum analyzer. This is possible. You will not, as you mention your self, see the angle information.
What can be tricky though is to have enough power from the noise source since these often have very low output power. If you amplify this noise you must be sure to have a good output match!

Hope this helped.
Regards
Jens