Forum

When measuring noise power [ e.g. noise floor values ]in satellite communications or microwave sytems you either have to describe the resolution bandwidth of the measuring device [ e.g. a spectrum analyzer res B/W value of 30 khz ] or have the noise power normalized [ via manual calaculation, DSP etc ] to a value of x dBm/Hz.

I noticed in most wireless systems that values of [ say ] -85 dBm are touted. What measurement bandwidth is used ?

Dave

I don't know if this helps, but the 802.11T Draft document
(Recommended Practice for Evaluation of Wireless
Performance)
had this as a requirement for spectrum analysers.


B.3.2.4 Average displayed noise level
The average displayed noise level should not exceed -130 dBm when measured above 1 GHz using a 1 Hz
RBW.

Tks

That is very useful. It's actually the first time I've seen res b/w mentioned. When I get time, I'll put a posting on using a spectrum analyzer. Depending on the settings for resolution bandwidth, video bandwidth, sweep time, video filtering etc., spec an's can give totally misleading results in many cases.

Whenever we refer to a noise or signal measurement using a spec an, we should really give the res bandwidth value as well as the dBm value. The value given by a spectrum analyzer set to a res b/w of 30 khz will be totally different from that of one set to 1 khz [ unless a clean carrier unmodulated signal is being measured ].

Sadly, the industry does not "insist" on this being done, and it can lead to a lot of confusion.

It's a bit like saying to a French guy and an American guy, how tall is that other guy ? One will probably give a value different from the other due to the imperial and metric systems used in those countries. However, if we say "How tall is that guy in feet ?" then we have a common reference to work to.

Dave