[hpsdr] Dither on Mercury-EU

Wed Aug 17 09:49:33 PDT 2011

... again, Helmut, 
I don't thick that this is correct, but may be I misunderstood?
You cannot bypass the amplifier. 
If you really could, then you would indeed see the full effect of
worsening the
S/N by activating dithering at the ADC. 
However, the amplifier is always included in the signal path. You only
can add 20dB at the input in front of it.  Doing so, the overall noise
figure
is worse by 20dB, but the influence of dithering stays small, in the
sense of reduced due to the amplifier ahead. The influence is expected
to cause the same reduction given in dB but now on a higher level.

Here are the results of noise power density measurements with my
TAPR-Mercury on 20m:
Input matched with 50 Ohm.

0dB attenuation: noise power density is -132dBm/1kHz
dither enabled: -129dBm/1kHz

20dB attenuation: -113.5dBm/1kHz
dither enabled: -110.2 dBm/1kHz.

Summarizing, Dithering increases noise level in my case by about 3dB,
independent of the additional attenuation at the front, in agreement 
with what I was saying before.

PS:
0dB means "preamp on",
20dB attenuation means "preamp off" as some software versions are still
saying.
But we know, the preamp is always on, only an attenuator is included at
the frontend.

73, Hans, DL2MDQ.

-----Original Message-----
Date: Wed, 17 Aug 2011 12:27:05 +0200
Subject: AW: [hpsdr] Dither on Mercury-EU
From: "Helmut" <dc6ny at gmx.de>
To: <h-hartfuss at t-online.de>

Hi Hans, 

	of course I'm familiar with calculating noise figures in cascaded
arrays,  but I didn't investigate the noise figures of the single
stages. You can ‘bypass' the preamp by  adding  the 20 dB pad in
front of the preamp and  you'll have similar  readouts  concerning SNR
reduction.  It seems that these readouts are ok, because they
correspond with my MDS, blocking and IM tests I made with external
test equipment.  

	73, Helmut_DC6NY  

	Von: h-hartfuss at t-online.de [mailto:h-hartfuss at t-online.de] 
 Gesendet: Mittwoch, 17. August 2011 10:51
 An: Helmut
 CC: hpsdr at lists.openhpsdr.org
 BETREFF: Re: [hpsdr] Dither on Mercury-EU  

	Hi Helmut, 

	I don't think that this is completely correct. The process is
conducted at the ADC by dithering over 2 or 3 quantization steps. Of
course this results in an S/R reduction.  

	However, the ADC is not the first stage in the receiver system. The
first stage is a low noise preamp with 20 db gain.  

	The overall noise temperature of the receiver is determined by the
noise temperature of the first stage plus a minor contribution from
the second which is the noise temperature of the ADC, however, divided
by the gain of the first. That's cascading rule! 

	You can easily calculate that it is impossible to increase the
overall noise temperature by an order of magnitude just by increasing
the ADC noise figure by say a factor of 4. 

	73, Hans, DL2MDQ.  

 -----Original Message-----
 Date: Wed, 17 Aug 2011 09:35:49 +0200
 Subject: Re: [hpsdr] Dither on Mercury-EU
 From: "Helmut" <dc6ny at gmx.de>
 To: "'wully'" <wully at bluewin.ch>, <hpsdr at lists.openhpsdr.org>

 ***** High Performance Software Defined Radio Discussion List *****

 Hi Alfred,

 enabled dithering reduces the signal-noise-ratio at least by 6-10 dB.
If
 less, something is wrong with the efficiency of dithering. My two
 mercuries-eu show similar values.

 73 Helmut, DC6NY

 PS: Perhaps we meet us 40m to continue

 -----Ursprüngliche Nachricht-----
 Von: hpsdr-bounces at lists.openhpsdr.org
 [mailto:hpsdr-bounces at lists.openhpsdr.org] Im Auftrag von wully
 Gesendet: Mittwoch, 17. August 2011 09:07
 An: hpsdr at lists.openhpsdr.org
 Betreff: [hpsdr] Dither on Mercury-EU

 ***** High Performance Software Defined Radio Discussion List *****

 Hi all

 First, I would like to add my congratulations for Phil Harman for
this 
 award!

 I have noticed a strange behavior of Mercury-EU, which might indicate
an 
 individual board failure:

 Antenna input with 50 Ohms terminated, no antenna. When switching on 
 Dither I notice an increase in background noise level of between 7
and 
 11 dB depending on the band! Typical noise value without dither is 
 around -130dBm (slightly better than tapr-mercury).

 Doing the same experiment on my two Tapr-Mercuries, the increase of 
 noise is hardly noticable with a value of 1 dB.

 Since I don't have two Mercury-EU, I can not compare.
 Normally, I am running the HPSDR with Dither off, so this is only a 
 "theoretical" problem, but it is interesting anyway.

 Any suggestions are much appreciated.

 73, Alfred, HB9EPU
 _______________________________________________
 HPSDR Discussion List
 To post msg: hpsdr at openhpsdr.org
 Subscription help:
 http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
 HPSDR web page: http://openhpsdr.org
 Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

 _______________________________________________
 HPSDR Discussion List
 To post msg: hpsdr at openhpsdr.org
 Subscription help:
http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
 HPSDR web page: http://openhpsdr.org
 Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

 1313599773.0