Noise power over bandwidth - RF Cafe Forums

 Post subject: Noise power over bandwidth

Posted: Wed Mar 24, 2010 1:26 pm 

I have an opportunity to get a spectrum analyzer for a few days and I want to take some measurments, as I've never been able to get my hands on one. I have put together a multiple stage amplifier that should give me about 30 dB of gain at 450 MHz. I want to measure noise power over a bandwidth. How do I calculate total power dissipated out of the amplifier, with the input terminated? I am writing a utility to grab the data and store it in a file, so I'll have 100 data points, and I want to get total noise power out. It's my understanding I can do this directly with a power meter and power sensor, as it's a wideband detector, but I can not get access to one.

Thanks

 Post subject: Re: Noise power over bandwidth

Posted: Mon Mar 29, 2010 8:55 pm 

Greetings newenglandguy:

Agilent has published a nice application note on the subject, and can be found here -

http://cp.literature.agilent.com/litweb/pdf/5966-4008E.pdf

It gives the relationship of the resolution noise bandwidth (RBW) setting on the SA to the equivalent noise bandwidth of the filer used. Knowing that, you can sample and sum the power in the adjacent bands to get total power across your total bandwidth.

You mention having 30 dB of gain, which should be plenty to assure that you are measuring the noise power of the amplifier rather than that of the SA itself. As long as the noise floor level of the terminated SA rises more than 10-15 dB with the DUT connected, the SA contribution will be negligable.

_________________
- Kirt Blattenberger
RF Cafe Progenitor & Webmaster

 Post subject: Re: Noise power over bandwidth

Posted: Wed Apr 07, 2010 3:34 pm 

Depending on the Spectrum Analyzer used, you may have the feature (option) to measure integrated power over a band. This would also give you what you want. It's not typical to get Total Output Noise power from an Amp. I'm not clear why you want that. More typical is to get Noise Power over some bandwidth or Spectral Noise Power (dBm/Hz) and either of these values can be used to figure the Noise Figure of the amp as a function of frequency of interest. Without the Integrated Band Power feature you have to do the power summation yourself and make allowances for the RBW and equivalany noise bandwidth, so the sum of power is not overlapping measurments, but true integrated power. Don't forget to convert to linear (watts) units for the summation and then afterwards convert back to dBm.

_________________
Sunshine Design Engineering Services