20676Perseus SDR architecture
- Jul 1, 2013What I think I've learned (loosely-translated) from the gurus on this list:
1. The S-meter displays the signal after filtering so you can get a true
handle on noise power in the channel.
2. The sensitivity of the S-meter moves in 10 db steps, tracking the pads,
so the signal level always appears the same.
3. The pads are not passive-loss devices as once was traditional in
fixed-gain front ends. Instead, the pads change the "gain" of the
preamplifier. That in turn changes the measured noise level because with
"x" db pad the channel has "x" db less gain.
4. OR: I am still "all wet" as we say over here <g>
On the Road
From: Roelof Bakker
Sent: Sunday, June 30, 2013 4:38 AM
Subject: Re: [perseus_SDR] Perseus SDR architecture
> The observed signal levels seem to be independent of the amount of pad IIn professional receivers the S-meter displays the signal level at the
> have switched in. This would lead one to think that this measurement is
> an absolute signal strength.
antenna port of the receiver.
This level does not change when you switch in the internal attenuator. The
S-meter reading is compensated for the amount of attenuation.
This is not the case in amateur radio equipment; however as this is what
we are used to it is considered the norm, which it is not. In other words
the software compensates for the attenuator setting. E.g. with 20 dB
attenuation it measures a 20 dB lower signal, but the software adds 20 dB,
so the S-meter displays the same level.
> And yet it's affected by the bandwidth adjustment.The S-meter Level (dBm) displays the power measured at the antenna port.
This level is dependent of the bandwidth.
This can be easily demonstrated by disconnecting the antenna and measuring
the noise level.
When you double the bandwidth, the noise power is doubled as well, which
equals an increase of 3 dBm on the S-meter. A difference in bandwidth of
four, gives a change of 6 dBm on the S-meter. And a ten times wider
bandwidth will give 10 dBm more power.
I strongly advise to play a bit with the bandwidth settings, which will be
> The related head-scratcher is that the amount of pad switched in seems toThis is only logical, as the attenuator really attenuates the noise by 30
> affect the noise floor...inversely. 30 db of pad produces noise levels
> that are 30 db BETTER than the level seen with 0 pad.
dB, though this is not displayed on the S-meter!
As an aside, almost all analogue receivers have a S-meter that derives the
reading from the AGC.
The reading starts not at the noise floor, but at a signal strength of
This is still the case with new offerings. E.g. Funkamateur Magazin June
2013, page 600, which gives a graph for the new TS-990 S-meter.
So the PERSEUS is one of the first receivers where the relationship
between bandwidth and S-meter / Level reading can be properly observed.
(I have build a home made receiver 20 years ago that did the same by using
a 12 section logarithmic amplifier)
I hope this helps!
Roelof Bakker, pa0rdt
- << Previous post in topic Next post in topic >>