[hpsdr] Signal level discrepancy RA Utility / Multimeter

[DK4RW]

Dear All,

while documenting local RFI levels for a complaint to the authorities, I
noticed differences between the signal levels indicated by the PowerSDR
S-Meter (Multimeter) and the Radio Astronomy (RA) utility.

My set up:
Hermes, firmware v2.9, Alex, Pennylane
PowerSDR_mRX_PS_v3.3.6 (11/16/15);  Win7 Pro / 32 bit 
PowerSDR Meter settings: 
	Analog Hold         1000ms
	Digital peak hold 500ms
	Averaging time   1000ms
	Analog refresh    100ms
	Digital refresh     500ms

RA Utility settings used for the measurements:
	sampling period 50ms
	averaging             20 samples    for 1000 ms averaging

Two examples of measured signal levels:
1. 10m basically white noise; AM, bandwidth 10 kHz; 
	Signal: -80 dBm    Sig Avg:  -95 dBm    RA: -82 dBm;    Differences
Signal-RA: +2 dB    SigAvg-RA: -13 dB
	
2. 41m BC Station (7.31 MHz); AM, bandwidth 10 kHz; signal levels only
approx. because of fading
	Signal: -50 dBm    Sig Avg: -58 dBm    RA: -73 dBm;    Differences
Signal-RA: +23 dB    SigAvg-RA: +15dB

I am puzzled by the large difference of the indicated signal powers, even
more so by the fact that the differences seem to depend on the signal
content in the passband. Has anybody else observed similar differences? Or
am I doing something stupid?

According to the FLEX-5000 Owner's manual Chapter 5 p.77 "Signal" and "Sig
Avg" are true RMS power readings of the peak and averaged power,
respectively. The RA utility (RAForm.cs) gets the signal levels from the
wdsp.CalculateRXMeter function, obviously in dBm. RA then averages the dBm
values and adds a constant (RA_cal = -4.23f) to the average. 
Being totally unfamiliar with C# and the development environment I was not
able to find the wdsp.CalculateRXMeter function. So I am not sure which
signal level value wdsp.CalculateRXMeter returns, and how it relates to the
Multimeter reading. I would appreciate if someone could help to understand
how the signal levels are calculated. 
>From my POV, ideally, RA and Multimeter would yield identical RMS power
levels, provided that the same averaging conditions are applied, of course.
Thank you.

73, Wolfgang DK4RW