Loading ...
Sorry, an error occurred while loading the content.

Re: SDR-IQ, Perseus, QS1R bench measurementsSDR]

Expand Messages
  • Leif Asbrink
    Hi Clemens, ... I had a look at the Perseus IM2 today. Actually, the signals at the I2 frequencies are not caused by second order intermodulation but by some
    Message 1 of 33 , Apr 30, 2008
    • 0 Attachment
      Hi Clemens,

      > >Whoa, hold your horses...
      > >Wideband SDR is NOT just red roses!!
      > >Comparing best-of-the-best HF receiver with the new wide-band
      > >technology will often show the SDR superior, with DR above 100 dB,
      > but >it put extremely high demand on the preselector. If you have one
      > or >several strong signals this will set the level which the A/D will
      > <approach full scale. I.e. there is no AGC to save the day! With the
      > >old-tech HF receiver, having a roofing filter with 5-20 kHz this
      > will >allow much less strong signals to enter latter stages and thus
      > will >handle the strong out-of-band signals better .
      >
      > I second that.
      > On my 2x46m dipole I see and hear on the perseus quite strong IM2
      > products on the 20m band, generated by extremely strong bc signals on
      > 41m,*even with preselector on*,while on my IC-7800 I detect not the
      > faintest signs of IM2 even *without* the tracking preselector,only
      > the bpf do it.
      > I can even switch on the preamp without any problems at all.
      > In central europe on 40m you do need a *wideband* DR3 of 110dB like
      > the ic-7800 (or E1800 or hilberling) to stay IM3 free without signal
      > input attenuation.
      > In the US/Canada IP2 and wideband DR3 are far less important than
      > here in central europe.
      > I think we'll have to wait for 18bit enob adc's to outperform
      > the best analogue designs also in these areas.
      I had a look at the Perseus IM2 today. Actually, the signals at
      the I2 frequencies are not caused by second order intermodulation but
      by some other phenomenon. The interference drops rapidly with the input
      signal and at -5 dBm the interference is gone.

      The builtin attenuator of Perseus is useless, it does not allow
      more input signal without IM2, but adding a 10 or 20 dB resistive
      attenuator would remove the IM2 that you hear on 14 MHz.
      Presumably the sensitivity would suffer a bit too much.....

      The IM2 problem does not require a better A/D converter. 16 bit is
      plenty, but the preselectors and associated switches could be improved.
      On the other hand, a simple external preselector would eliminate the problem:-)

      73

      Leif / SM5BSZ
    • Leif Asbrink
      Hello Lasse and all, ... Yes. Sure:-) ... Hmmm, it does.... ... With the preselector on and when listening on 14 MHz, full scale for 7 MHz signals is at +15
      Message 33 of 33 , May 2, 2008
      • 0 Attachment
        Hello Lasse and all,

        > Just a few notes, remember that the Perseus do show the input level at
        > the antenna prior to the attenuator. I.e. if you switch the
        > attenuator on, the receiver will still show the same input power.
        Yes. Sure:-)

        > And dither... I was surprised that dither do affect IM2! Dither is used to
        > average away spurs within the A/D and stationary signals should not be
        > affected by this!
        Hmmm, it does....

        > Finally most measurements on A/D's are made with two
        > tones @ f.s. -7 dB which means one have to account for all input power.
        > If I remember correctly the f.s. equals to -6dBm on the Perseus.
        With the preselector on and when listening on 14 MHz, full scale
        for 7 MHz signals is at +15 dBm. Two equal tones at +8 dBm
        do not saturate the A/D.

        > It is *very* interesting to see how the IM products do behave when input
        > levels are reduced, as often the A/D is not showing the nice "normal"
        > IMD-ratios then.
        OK. I have uploaded an IM2 test for the Perseus here:
        http://www.sm5bsz.com/perseus/perseus.htm

        As it turns out, IM2 behaves similarly to IM3. My guess is that
        the "intermodulation" is caused by internal coupling inside the A/D.
        The non-randomized digital output of the A/D reach its analog input
        about -120 dBm. (preamp and preselector ON.)
        This coupling is stronger in the SDR-14:
        http://www.sm5bsz.com/dynrange/qex/digital-imd.pdf
        http://www.sm5bsz.com/digdynam/practical.htm

        The IM3 coupling in Perseus is about 30 dB lower than the
        coupling the SDR-14 (when measured below full scale -90
        and -120 dB respectively.)

        In case the IM2 signal is a problem, a simle wideband amplifier could be
        the solution in case the 7 MHz signals would stay below about 0 dBm
        while the noise floor on 14 MHz is lifted high enough above the second
        order signals.

        In case full dynamic range is needed on 14 MHz due to strong local signals
        on that band, one would have to attenuate the 7 MHz signals down below
        -70 dBm which would require a rather sophisticated filter.

        73

        Leif / SM5BSZ
      Your message has been successfully submitted and would be delivered to recipients shortly.