S Meter & dBm Hz display - df8oe/UHSDR GitHub Wiki

An S-Meter should be able to do two things:

1.) Accurately measure the absolute signal power level 2.) Allow the relative comparison of two signal power levels.

For these purposes, signal power level S9 is defined as an absolute level of -73dBm (a power level -73dB below 1 milliwatt) and one S-meter category is defined as 6dB, making an S8 signal 6dB stronger than a S7 signal. Although this S-Meter-System has been clearly defined as early as 1981 in a technical recommendation of the IARU, S-Meters of almost all commercial transceivers are mostly not usable, because they do not fulfill these two clearly defined criteria. This is indicated in the test results by WB9KMW and W8WWV.

A second flaw is the following: consider a noise signal of a constant level. If you measure this signal in a bandwidth of 500Hz, you will measure for example a signal power level of -100dBm. Now what if you measure the same signal in a 1000Hz bandwidth? It will be twice the signal level, i.e. -97dBm. That means, a signal power level measurement is pretty useless if you do not state the bandwidth in which you have measured your signal. These two points now apply to the S-Meter of the UHSDR.

So, how can we give other hams an indication on signal power level without having to use an unreliable S-meter and without having to state the filter bandwidth, in which this level occurs? The UHSDR software measures the signal power level in the filter passband. All signals outside the filter bandwidth do not contribute to this measurement. Additionaly the UHSDR can measure PSD (power spectral density, if S-meter is set to dBm/Hz in the menu) which measures the total signal power inside the filter passband and divides that by the filter bandwidth (measured in dBm/Hz). That ratio is the power spectral density, which is independent of the chosen filter bandwidth and it is given in dBm/Hz. This display is given in a blue box below the S-Meter. Please be aware of certain characteristics of this display, which you may find strange in the first place, if you are used to a conventional and unreliable S-Meter in your old transceiver:

• If you are listening to a noise signal (a noise signal generator or your antenna on a free frequency), the display will/should be totally independent of your chosen filter bandwidth
• If you tune into an AM station (with its very powerful carrier signal) and you decrease the filter bandwidth, the signal display in dBm/Hz will increase!
• The power spectral density is given in dBm/Hz, not in S-units
• The dBm/Hz display is comparable between different antennas/setups/receivers

If you want to change measurement of signal power level (dBm) vs. power spectral density (dBm/Hz), please change the setting of the basis of the S-Meter in the menu.

What can you do with this display?

• Measure your antenna noise and compare it with other antennas
• Measure a signal of your QSO partner and give an objective and comparable feedback on signal strength

Measurements for this display have shown a surprisingly high accuracy verified with a direct sampling SDR (Perseus) and independently with a precision signal generator. Accuracy is +-2dBm down to -102dBm for a 2.4kHz bandwidth and down to -119dBm for a 300Hz bandwidth !

Thanks, Christian DL9NL, for your accurate measurements! HERE

You have the possibility to choose between the dBm/Hz display, a dBm display (which sums up all the signal energy inside the filter passband) or you can switch off the display in the Display Menu.

S-Meter:

• based on dBm (measures signal power)
• based on dBm/Hz (measures power spectral density)

dBm display:

• OFF
• dBm (measured signal power in a selected bandwidth)
• dBm/Hz (calculated power spectral density)

User choice is stored in EEPROM.

***

Table of S units reference (HF)

S units	P (dBm)	U on 50ohm (uV)
+10dB	-63	158,3
9	-73	50,1
8	-79	25,1
7	-85	12,6
6	-91	6,30
5	-97	3,16
4	-103	1,58
3	-109	0,79
2	-115	0,40
1	-121	0,20