The external noise figure Fa is defined (from ITU P.372-13) as:
Above, the noise floor in 9kHz bandwidth with a CISPR quasi peak detector is about -78dBm. This is 12dB above the instrument noise floor, sufficient to not bother making a correction and we can take the external noise to be -78dBm (see below for correction calculation if needed). Lets allow 1dB loss in the antenna system, and call it -77dBm at the air interface.
The noise power in a resistor in 1Hz bandwidth is kT0 or -174.0dBm, so we can calculate the power in 9kHz to be -174.0+10log(9000)=-134.5dBm.
The power measured is -77–134.5=57.5dB above the noise power in a resistor at 290K, so the ambient noise figure Fa is 57.5dB. Ta is 290*10^(57.5/10)=1.63e8K.
Above is a chart of expected median ambient noise figure Fa for ITU P.372-13. At 7.1MHz, the expected value for a Residential precinct is 48.9dB, and our measured 57.5dB is 8.6dB above that… as I said it is quite noisy at the moment.
In fact ITU P.372-13 is based on a survey with a vertical monopole antenna, and the test antenna for the measurement in this article is a horizontal antenna and we might expect noise to be a little lower than P.372-13 predicts.
What does this mean in terms of S meter readings?
On a receiver with 2kHz effective noise bandwidth, the received power would be -78+10*log(2000/9000)=-84.5dBm. Taking S9 as -73dBm and 6dB per S unit, it is 11.5dB below S9, so approximately S7 on a calibrated S meter.
Correction for instrument noise floor
The instrument indicates the total of external and internal noise powers, and the adjustment to obtain external noise power component is 10*log(1-10^(RNF/10)) where RNF is the noise floor relative to the measurement. For example if the measurement is -78dBm and instrument noise floor is -90dBm then RNF is -12dB and the adjustment is 10*log(1-10^(-12/10)) =-0.283dB, so external noise is -78-0.283=‑78.283dBm.
References / links
- ITU-R. Sep 2016. Recommendation ITU-R P.372-13 (9/2016) Radio noise.