Re: Multichannel receiver (SDR based)

On Mar 29, 2:18 pm, Eric Jacobsen <eric.jacob...@xxxxxxxx> wrote:
On Sun, 29 Mar 2009 07:18:33 -0500, "alberto.fuggetta"

<alberto.fugge...@xxxxxxxxx> wrote:
Vladimir Vassilevsky <antispam_bo...@xxxxxxxxxxx> wrote:

For the variety of reasons, the SDR will never reach the performance
level of the traditional superhet. That may or may not be important
depending on the particular application.

I probably agree, but I think you should say what is
fixed and what isn't. If you have a cheap superhet receiver,
and no holds SDR (high sample rate and dynamic range on
the A/D conversion, etc.) then I would probably believe
the SDR was better.

-- glen

Thanks everybody for the feedback.

Just some info more to clarify my questions.
My idea is to translate the band of interest to a very low IF (i.e. 0-25
MHz)by means of a standard analog mixer, then sample the overall band with
several channels.
The target is to receive (at least) two different channels at the same
time with only one analog front end.
I guess, it should not be a problem if the received signal strength of
different channels is almost the same.
Let's simplify the model so that to have only two channels: a strong one
and a weak one.
If I had an AGC system, how should I set the gain?
I was not able to find any similar example in the literature.

Thanks

Alberto

Cellular base stations have done this sort of thing for a long time.
As someone else pointed out, search the web for "near-far problem" and
maybe include "dynamic range" or something like that.

It is a difficult problem if you can't adequately constrain the input
conditions. If the power levels of the signals fluctuate
independently over a broad range it gets bad. If they can change
power levels quickly (or, say, come and go quickly) then it's even
worse.

As Steve mentioned, usually it's advantageous to have an AGC, since it
provide the opportunity to maximize SNR in the analog stages, which is
where much of the noise and distortion comes from, anyway.

I wouldn't say that the analog stages are where the noise and
distortion come in. The ADC is a major limiting factor in a digital
radio. Although they have been steadily improving over the years, the
ADC still has less dynamic range than the analog circuitry driving
it. That is why an AGC is more important for a digital back end than
an analog one. If you don't do a good job of digitizing the signals,
they are hosed from then on.


The
problem with a multi-channel system is that the AGC has to be set
according to the composite power of all signals present, plus some
headroom for fluctuation. If, for example, most channels are empty
and the present channels are low-level, the AGC may crank up the
analog gain to increase the level of the weak signals that are
present. If a strong signal then appears if adequate headroom was
not left in the dynamic range the amps can go into saturation and then
it's essentially game over until it can correct itself.

So a constant-gain system with a lot of dynamic range does have an
advantage for the case where the signal levels can vary broadly and
quickly in that it will avoid such dynamic problems as saturation.
It's very expensive to build such a system, but sometimes it's what
has to be done.

Yes, a fast AGC is important. I was working on a system once that was
trying to do a job on the cheap using commercial handheld radios.
They seemed to do a good job in most respects, but when a signal was
keyed, the AGC took some ms to settle down and in the meantime it
would oscillate a bit pinching off the start of the signal. No one I
was working with would tackle the modification of the AGC to make it
respond without this overshoot, so the cheap units were not usable in
the end. This was one of those lessons in corporate (dis)
functionality. The PM was due for shipping out to a new assignment in
a couple of months. When we did the testing and found the receivers
did not work in narrow band mode; rather than to take the time and
money to fix the problem, the PM swept it under the rug until after he
was gone. The project moved ahead as if there was nothing wrong and
in the final testing stage they found that it would not work in some
environments, so it was literally shelved, never to see the light of
day. When the problem was first found, we could have either fixed it,
or stopped the project without spending all the money. But then the
PM likely would not have gotten his new assignment... I won't tell
you who paid for all of this, because you wouldn't like it. I'll just
say that no one was held accountable because there was no "profit"
incentive.

Rick
.

Relevant Pages

  • Re: MCE 5 no longer updates guide
    ... I mean receiving only analog channels such as channels 2-67 using Media Center. ... A converter box will not help you in any way since it is designed for OTA digital signals and not for analog or for cable digital signals in fact using it will prevent you from receiving anything from your cable provider. ...
    (microsoft.public.windows.mediacenter)
  • Re: Multichannel receiver (SDR based)
    ... The target is to receive two different channels at the same ... If I had an AGC system, how should I set the gain? ... maybe include "dynamic range" or something like that. ... If the power levels of the signals fluctuate ...
    (comp.dsp)
  • Re: IMTS vs MTS was 1965 mobile phone on "Get Smart" [telecom]
    ... digital signal is harder to receive than the older analog signals. ... I'm guessing that this has more to do with the frequencies typically used for ... optional common channels between police, fire, and rescue which often ...
    (comp.dcom.telecom)
  • Re: Converter boxes for digital TV (for when analog TV goes away).
    ... On Analog, it's definitely hard ... its higher output power, provided good reception on Analog but, on ... You do understand that all of the digital signals are in the UHF bands ... using channels 2-13 will go away and be reallocated to another ...
    (rec.outdoors.rv-travel)
  • Re: Digital Cable STB - analog picture quality?
    ... >Digital signals should never be snowy or grainy, ... I guess I was referring primarily here to the analog channels. ...
    (microsoft.public.windows.mediacenter)
Loading