discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Quadrature demodulator update


From: John Turner
Subject: Re: [Discuss-gnuradio] Quadrature demodulator update
Date: Wed, 12 Feb 2003 09:05:54 -0800

Hi Folks,

In many radios the AGC is distributed in practical chunks. The front end may
have a switchable of say 20dB attenuator. This is used under high signal
level conditions to avoid signal overload at the expense of noise figure. An
AGC with finer control and greater range is built into the IF to set the
demodulator level, or ADC input level. It would make sense that the IF is
not necessarily gain controlled in the AOR or ICOM receivers.

An example it the Intersil Prism II 802.11 chipset. There is a bulk
attenuation switch of about 20dB at RF(2.4GHz) and a variable gain IF
amplifier used for the bulk of the adjustment (70dB range?) with a fine step
control. The whole chain is controlled by the baseband level driving the
A/D. I seem to recall it acquires a signal in 10s of usec over it's entire
dynamic range.

To maximize the utility of a system the gain needs to be adjustable. A few
cautions about AGCs, they are closed loop control systems. As such they can
oscillate, have slow acquisition response, level overshoot etc.. As David
points out AGCs typically function only with the signal of interest,
introducing multiple carriers will cause fluctuations in the measured
signals which may either introduce unwanted modulation or cause the system
to function at inappropriate gain levels. There are a number of references
to AGC circuits for use in Single Sideband Radios which I would assume are
available through the ARRL.

On more positive comments, as the digital dynamic range expands the analog
AGC requirement is decreasing, i.e. if a signal range of 120 dB is required,
90 dB may come from the ADC and decimation leaving only a 30dB analog
requirement.

JT

----- Original Message -----
From: Dave Emery <address@hidden>
To: John Turner <address@hidden>
Cc: Philip Mackenzie <address@hidden>; <address@hidden>
Sent: Tuesday, February 11, 2003 11:29 PM
Subject: Re: [Discuss-gnuradio] Quadrature demodulator update


> On Tue, Feb 11, 2003 at 03:30:36PM -0800, John Turner wrote:
> > Hi Philip,
> >
> >
> > Depending on the input signal level and the gain of the unit the
amplifiers
> > could still saturate or generate some undesired intermodulation
products. In
> > the AOR information they quote a high level signal of 5uV, this is most
> > likely for a narrowband signal not broadcast FM, i.e. 20uV (-80dBm) is a
> > typical limit for stereo reception with 50dB quieting, and you may find
that
> > typical signals are 20 to 30dB above this limit. As an example, we are
> > located quite close to some local FM transmitters and experience levels
> > of -47dBm, these levels are sufficient to casue overloading of
commercial FM
> > radios.
>
> My experiance with ICOM radios is that the 10.7 mhz IF output
> is located ahead of most of the AGC action of the radio so it maybe
> sees 20 to 30 db of AGC action at the most when the radio is tuned to
> a strong signal.   And indeed gain to the 10.7 mhz output is not
> huge - maybe 20-25 db or so - it is not a constant level AGC controlled
> high power output as are the more traditional IF outputs of many
> other receivers.
>
> This does indeed have implications for the dynamic range of
> quadrature downconverters and A/Ds.   Given that that dynamic range
> of signals of interest is from maybe -125 dbm to perhaps as high as
> -25 dbm (sitting under the antennas) - 100 db or more - one does
> have to think about either using some kind of software controlled
> AGC (gain controlled by a DAC), hardware AGC, or careful tweeking
> to make the hardware work over the required dynamic range.
>
> RF modem systems I am familiar with actually adjust gain ahead
> of the A/D by looking at the data coming out.   This, however works
> better if one is digitizing only the signal of interest and not
> a whole band of stuff with carriers coming and going unpredictably
> including perhaps some strong nearby ones that are much more powerful
> than the desired signal.
>
>
>
> --
> Dave Emery N1PRE,  address@hidden  DIE Consulting, Weston, Mass.
> PGP fingerprint = 2047/4D7B08D1 DE 6E E1 CC 1F 1D 96 E2  5D 27 BD B0 24 88
C3 18
>





reply via email to

[Prev in Thread] Current Thread [Next in Thread]