I like the way that PowerSDR allows you to make relatively precise noise floor measurements for either the receiver or for the current band conditions. The RX meter tends to bounce around a lot on noise, making it hard to determine a good average. However, if your receiver has been calibrated, you can switch the RX meter from "Signal" to "Sig Ave" and get a much longer term average reading of the noise.
I like to average over a longer time that used for the defaults, which can be changed by going to Setup, selecting the Display tab, looking for the Multimeter section in the lower right corner, and changing Average Time to something in the 5000 (5 sec) range. Hit "Apply" when done. This feature gives an average that settles down gradually (converges?) to within 0.1 db over 20 to 30 seconds. In my mind this is an extremely useful feature. MDS (Minimum Discernable Signal) is often defined as injecting enough signal to make the output increase by 3 db. Increasing the output by 3 db means that the noise power equals the injected signal power and the two of them add together to double the total power (a 3 db increase). Thus, at the 3 db point, the injected signal level is at the same level as the noise floor of the receiver. Receiver MDS measurements should be done using a 50 ohm load. With PowerSDR and the above averaging, we can measure the noise power directly, and thus know directly the MDS of the receiver (or the noise floor of the current band conditions) without the need of injecting an actual measurement signal. The receiver bandwidth must be specified for MDS to be meaningful, and 500 Hz one of the bandwidth standards that is often used for this specification. I think it is extremely useful to take note of the noise level of the band and to see how it changes over the course of the day. A note of caution: In my home measurements of MDS, I have seen time of day variations in MDS results. I have come to realize that because the shielding of coax is not perfect (30 dB of attenuation?), a noisy band can impact the measurement of receivers in the 130+ dbm MDS region. I don't have an isolation chamber, so I just have to deal with it. You know you are in trouble when you can weakly hear signals across the band when the dummy load is connected. - Dan, N7VE -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dave Haupt Sent: Friday, June 15, 2007 12:05 PM To: flexradio@flex-radio.biz Subject: Re: [Flexradio] Noise floor, specifying You're getting very close to instrumentation norms. Noise, by its very nature, varies in apparent strength by the bandwidth in which you measure it. When measuring truly random noise, the power (in watts) is in direct proportion to the bandwidth of the filter used in the measurement. Man-made noise often consists of thousands of semi-correlated carriers, and the noise power can grow faster than simply being proportional to BW. To be accurate, an engineer specifies noise either as "dBm/Hz" which means "dBm measured in a one Hertz bandwidth" or "watts/hz" or something like that. It would also be equivalent to say "dBm measured using a 200Hz filter" or something like that. Audio-frequency guys tend to measure amplitudes in volts, at which point a square root gets into the picture, and you will read things like "nanovolts per root Hertz". Sine waves and noise are described using different mathematical terms and as an interesting result, to make a truly accurate spectrum analyzer, instrumentation manufacturers often use an algorithm to attempt to distinguish a sine wave from a noise function. Anybody who's used the digitally-enhanced analog spectrum analyzers by HP (8566/68, 8590 and 8560 series) has probably encountered the "noise marker". The instruments are calibrated for accuracy with sine wave signals (or sums of sine waves - any repetitive waveform), and are in error, fundamentally, for noise. When you turn on the "noise marker", the instrument makes additional calculations to make the readings accurate for random noise. When you get down to the nitty-gritty of noise, it's not at all a simple subject. 73, Dave W8NF >>From: "Doug McCormack" <[EMAIL PROTECTED]> >>Subject: Re: [Flexradio] Noise Floor S meter vs Panadapter >> Thanks everyone for the quick and detailed explanation. I now understand >> the S meter displays one sum across the entire width of the green filter, >> while the panadapter displays hundreds of sums across the width of each >> individual pixel. When I set the filter very narrow (approaching one pixel) >> the S metter reading approaches the panadapter reading. >> My old Kenwood was similar in that selecting the narrow CW filter caused a >> drop in S meter noise floor. This behavior makes perfect sense for any radio. >> I guess if someone asks about the noise floor, they need to specify at what >> filter width. >> 73, Doug, VE3EFC ________________________________________________________________________ ____________ Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. http://mobile.yahoo.com/mobileweb/onesearch?refer=1ONXIC _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Knowledge Base: http://kb.flex-radio.com/ FlexRadio Homepage: http://www.flex-radio.com/ _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Knowledge Base: http://kb.flex-radio.com/ FlexRadio Homepage: http://www.flex-radio.com/