[Top] [All Lists]

Re: [CQ-Contest] SDR Mythbusters - ADC Overload myths debunked...

To: "'Mark van Wijk'" <pa5mw@home.nl>, "'cq-contest'" <cq-contest@contesting.com>
Subject: Re: [CQ-Contest] SDR Mythbusters - ADC Overload myths debunked...
From: "Marijan Miletic" <s56a@bit.si>
Date: Mon, 9 Nov 2015 16:29:11 +0100
List-post: <cq-contest@contesting.com">mailto:cq-contest@contesting.com>
Mark, I thought for a long time that a carrier as a steady signal but in the
case of ADC it overflows only at the short peaks and it can often be


Weak signal can be detected by human operator with analog RX in the case of
speech and CW.  DSP systems are much better in advanced modulations.  

Hams are amazed by HF PSK while EME in 21st century is done by WSJT with
reasonable hardware.  It is easier for me to make RTTY LP QSO than CW :-(

DSP noise reduction greatly enhances weak signal detection due to human
hearing abilities.  No way to do it with my beloved electronic circuits.


73 de Mario, S56A



From: Mark van Wijk [mailto:pa5mw@home.nl] 
Sent: 9. november 2015 15:57
To: cq-contest; Marijan Miletic
Cc: rwmcgwier
Subject: Re: [CQ-Contest] SDR Mythbusters - ADC Overload myths debunked...



The NPR test is flawed because of its reference MDS.


MDS is measured using a STEADY carrier. 

As soon as you modulate a weak signal carrier, it  disappears in the band
noise of most DSP/SDR based receivers where analog systems still offer a
readable signal.

My measurements showed that analog receivers recover modulated signals at 5
- 8 dB lower levels.

*****try measureing MDS with modulated testignals for a change *****

(a simple dit-dit-dit-dit as used in a S/N setup will do)


My conclusion: 

the MDS reference as a static benchmark does not resemble any real-life

The calculated NPR figure is thus flawed; it does not relate to any
intelligent radio communication reception.

73 Mark, PA5MW



K9YC wrote:

>The Noise Power Ratio Test is especially problematic --


CQ-Contest mailing list

<Prev in Thread] Current Thread [Next in Thread>