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

RX difference

Expand Messages
  • g8bup
    Hi, For a guy who is not so upto date, can someone point me to the pros and cons of a direc conversion receiver which uses FPGA, like Perseus, and one which
    Message 1 of 12 , Sep 1, 2009
    • 0 Attachment
      Hi,

      For a guy who is not so upto date, can someone point me to the pros and cons of a direc conversion receiver which uses FPGA, like Perseus, and one which uses DSP, like SDR14. Both seem to come to similar conclusion.

      Tnx
    • hfdecoding
      Hi, http://pudxk.blogspot.com/2008/01/testing-perseus-sdr.html has a good write up about the Perseus and does some comparisons with SDR-IQ/14 s if that helps.
      Message 2 of 12 , Sep 1, 2009
      • 0 Attachment
        Hi,

        http://pudxk.blogspot.com/2008/01/testing-perseus-sdr.html has a good write up about the Perseus and does some comparisons with SDR-IQ/14's if that helps.

        The Perseus and QS1R are true SDRs there are a lot that are not true SDRs so a direct comparison may not be possible. Sherwood also lists test data for classic radios and SDRs - but with new technology the tests are not always the final reason for owning one.

        But myself and others on this list have sold high end radios once the Perseus arrived as the other radios were never used ...

        If you can check out some of the web sites that the owners of Perseus blog or write about on most of your questions will get answered -or- you could just buy one and find out how happy you are then :-)

        73

        Steve

        --- In perseus_SDR@yahoogroups.com, "g8bup" <g8bup@...> wrote:
        >
        > Hi,
        >
        > For a guy who is not so upto date, can someone point me to the pros and cons of a direc conversion receiver which uses FPGA, like Perseus, and one which uses DSP, like SDR14. Both seem to come to similar conclusion.
        >
        > Tnx
        >
      • Marco IK1ODO -2
        ... G8BUP DE IK1ODO Have a look at the attached article, published one year ago in RadCom. 73 - Marco IK1ODO
        Message 3 of 12 , Sep 1, 2009
        At 18.44 01/09/2009, you wrote:
        >
        >
        >Hi,
        >
        >For a guy who is not so upto date, can someone point me to the pros
        >and cons of a direc conversion receiver which uses FPGA, like
        >Perseus, and one which uses DSP, like SDR14. Both seem to come to
        >similar conclusion.
        >
        >Tnx

        G8BUP DE IK1ODO

        Have a look at the attached article, published one year ago in RadCom.

        73 - Marco IK1ODO
      • Marco IK1ODO -2
        sorry for BW, previous msg was meant as PR to G8BUP ... but went to the group :-) 73 - Marco IK1ODO
        Message 4 of 12 , Sep 1, 2009
        • 0 Attachment
          sorry for BW, previous msg was meant as PR to G8BUP ... but went to
          the group :-)

          73 - Marco IK1ODO
        • g8bup
          Thanks Marco. I was looking for a guide as to which process is best, that of Perseus FPGA way or SDR-IQ way. John
          Message 5 of 12 , Sep 1, 2009
          • 0 Attachment
            Thanks Marco. I was looking for a guide as to which process is best, that of Perseus FPGA way or SDR-IQ way.

            John

            --- In perseus_SDR@yahoogroups.com, Marco IK1ODO -2 <ik1odo@...> wrote:
            >
            > At 18.44 01/09/2009, you wrote:
            > >
            > >
            > >Hi,
            > >
            > >For a guy who is not so upto date, can someone point me to the pros
            > >and cons of a direc conversion receiver which uses FPGA, like
            > >Perseus, and one which uses DSP, like SDR14. Both seem to come to
            > >similar conclusion.
            > >
            > >Tnx
            >
            > G8BUP DE IK1ODO
            >
            > Have a look at the attached article, published one year ago in RadCom.
            >
            > 73 - Marco IK1ODO
            >
          • Marco IK1ODO -2
            ... The Analog Devices AD6620 DDC has inherent limitations in filter shape and image rejection, that are overcome implementing the same functions in the
            Message 6 of 12 , Sep 1, 2009
            • 0 Attachment
              At 17.54 01/09/2009, you wrote:
              >
              >
              >Thanks Marco. I was looking for a guide as to which process is best,
              >that of Perseus FPGA way or SDR-IQ way.
              >
              >John

              The Analog Devices AD6620 DDC has inherent limitations in filter
              shape and image rejection, that are overcome implementing the same
              functions in the Perseus FPGA. The overall logic is very similar, but
              Perseus is more advanced.
              Using an FPGA the designer has many degrees of freedom selecting the
              best configuration and filter coefficients. That makes the
              difference, SDR-IQ has a worst case image rejection of about -43 dBc,
              while Perseus has a worst case at -103 dBc.
              Other circuital differences are in the ADC and in front-end band pass
              filters. I have two Perseus and one SDR-IQ; I use the SDR-IQ mainly
              as last conversion in VHF receivers, under Winrad. In that way it is
              not too bad :-) but when connected to a long wire antenna it needs
              immediately 10 or 20 dB attenuation, otherwise it saturates.

              73 - Marco IK1ODO
            • g8bup
              Perfect Marco, thanks. So it seems both used the DDC function, just SDR-IQ implements this with bought in component, whereas FPGA way allows designer to tweak
              Message 7 of 12 , Sep 1, 2009
              • 0 Attachment
                Perfect Marco, thanks. So it seems both used the DDC function, just SDR-IQ implements this with bought in component, whereas FPGA way allows designer to tweak and play to best advantage.

                One thing I dont understand is image rejection. I understand what this is with super het rx, but with direct conversion, how is an imae generated, and at what freq.

                Tnx - John

                --- In perseus_SDR@yahoogroups.com, Marco IK1ODO -2 <ik1odo@...> wrote:
                >
                > At 17.54 01/09/2009, you wrote:
                > >
                > >
                > >Thanks Marco. I was looking for a guide as to which process is best,
                > >that of Perseus FPGA way or SDR-IQ way.
                > >
                > >John
                >
                > The Analog Devices AD6620 DDC has inherent limitations in filter
                > shape and image rejection, that are overcome implementing the same
                > functions in the Perseus FPGA. The overall logic is very similar, but
                > Perseus is more advanced.
                > Using an FPGA the designer has many degrees of freedom selecting the
              • Marco IK1ODO
                ... It is generated in the decimation/downconversion process, and you will find it at F(input signal) +/- F(output sampling rate). A numeric example: you tune
                Message 8 of 12 , Sep 1, 2009
                • 0 Attachment
                  At 21.15 01/09/2009, you wrote:
                  >
                  >
                  >Perfect Marco, thanks. So it seems both used the DDC function, just
                  >SDR-IQ implements this with bought in component, whereas FPGA way
                  >allows designer to tweak and play to best advantage.
                  >
                  >One thing I dont understand is image rejection. I understand what
                  >this is with super het rx, but with direct conversion, how is an
                  >imae generated, and at what freq.
                  >
                  >Tnx - John

                  It is generated in the decimation/downconversion process, and you
                  will find it at F(input signal) +/- F(output sampling rate).
                  A numeric example: you tune 14100 kHz with 100 kS/s (I'm thinking
                  about SDR-IQ). You see 14050 to 14150 kHz on your waterfall. Sweep a
                  signal generator thru the band, and when you reach 14151 you see a
                  signal again at 14051, sweeping to the right. It is attenuated, but
                  very visible.
                  To improve the image rejection, Nico limited the bandwith to approx.
                  80% of the sampling rate (125 kS/s gives 100 kHz of useable band on
                  the screen). But try using Winrad on the same hardware and sampling
                  rate, and you will see a noise rolloff at band edges - here is where
                  the images may show up. DSP may seem like magic, but some old signal
                  processing problems are always present. The good news are that they
                  may be controlled very effectively.

                  73 - Marco IK1ODO
                • Leif Asbrink
                  Hi, ... The Perseus and the SDR-14 have exactly the same architecture. They sample at VHF and use digital frequency conversion followed by digital filters and
                  Message 9 of 12 , Sep 1, 2009
                  • 0 Attachment
                    Hi,

                    > For a guy who is not so upto date, can someone point me to the
                    > pros and cons of a direc conversion receiver which uses FPGA,
                    > like Perseus, and one which uses DSP, like SDR14.
                    > Both seem to come to similar conclusion.
                    The Perseus and the SDR-14 have exactly the same architecture.
                    They sample at VHF and use digital frequency conversion followed
                    by digital filters and decimation (in several steps.) I am not used
                    to see direct conversion as the label on this type of purely
                    digital radio hardware. To me, direct conversion means that
                    some analog circuitry is used to produce two audio signals
                    I and Q. I and Q can be sent into the soundcard of a computer,
                    but they can also be low-pass filtered and phase shifted to make
                    an analog SSB receiver without any IF filter.

                    There is a significant difference between the Perseus and SDR-14
                    in that Perseus uses a much better A/D converter. The digital
                    filters used in the Perseus are also better than the filters
                    used in the SDR-14. The SDR-14 uses a chip from Analog Devices
                    to do filtering and decimation and that chip is optimized
                    for cost to meet the telecom standards. Perseus goes far beyond
                    that. By selecting other chips one could get the desired
                    performance with the same architecture as that used in the SDR-14.
                    There is one from Graychip that could be used but today an FPGA
                    is presumably more attractive in an amateur product.

                    73

                    Leif / SM5BSZ
                  • g8bup
                    Still perfect Marco, I got it. Now I just need to understand the difference between Perseus and Quicksilver, which essentially seems to be component and
                    Message 10 of 12 , Sep 1, 2009
                    • 0 Attachment
                      Still perfect Marco, I got it. Now I just need to understand the difference between Perseus and Quicksilver, which essentially seems to be component and implementation differences rather than basic design.

                      Thanks again, I'll be quiet now ;-)

                      John

                      >
                      >
                      > It is generated in the decimation/downconversion process, and you
                      > will find it at F(input signal) +/- F(output sampling rate).
                      > A numeric example: you tune 14100 kHz with 100 kS/s (I'm thinking
                      > about SDR-IQ). You see 14050 to 14150 kHz on your waterfall. Sweep a
                      > signal generator thru the band, and when you reach 14151 you see a
                      > signal again at 14051, sweeping to the right. It is attenuated, but
                      > very visible.
                      > To improve the image rejection, Nico limited the bandwith to approx.
                      > 80% of the sampling rate (125 kS/s gives 100 kHz of useable band on
                      > the screen). But try using Winrad on the same hardware and sampling
                      > rate, and you will see a noise rolloff at band edges - here is where
                      > the images may show up. DSP may seem like magic, but some old signal
                      > processing problems are always present. The good news are that they
                      > may be controlled very effectively.
                      >
                      > 73 - Marco IK1ODO
                      >
                    • Leif Asbrink
                      Hi John and Marco, ... That is true only if you use some specific software. Spectravue or perhaps Winrad. If you run SDR-14 with Linrad you have a freedom to
                      Message 11 of 12 , Sep 1, 2009
                      • 0 Attachment
                        Hi John and Marco,

                        > Using an FPGA the designer has many degrees of freedom selecting the
                        > best configuration and filter coefficients. That makes the
                        > difference, SDR-IQ has a worst case image rejection of about -43 dBc,
                        > while Perseus has a worst case at -103 dBc.
                        That is true only if you use some specific software. Spectravue or perhaps
                        Winrad.

                        If you run SDR-14 with Linrad you have a freedom to select filters yourself
                        and it is possible to get a much better spur protection if you can
                        arrange some front end selectivity (which is a good idea anyway to improve
                        the dynamic range.) Perseus has builtin filters, SDR-14 would benefit
                        from external filters. You would get them for free if you use e.g. a
                        144 to 28 MHz converter.

                        > Other circuital differences are in the ADC and in front-end band pass
                        > filters. I have two Perseus and one SDR-IQ; I use the SDR-IQ mainly
                        > as last conversion in VHF receivers, under Winrad. In that way it is
                        > not too bad :-) but when connected to a long wire antenna it needs
                        > immediately 10 or 20 dB attenuation, otherwise it saturates.
                        But, with a 20 dB attenuator in place it will still be perfectly
                        adequate at many frequency bands. Everywhere where the noise floor
                        is more than 13 dB above the internal noise of the SDR-IQ. Saturation
                        during lightening crashes and other transcients make no harm at all,
                        it is not the saturation indicator one should monitor but the generation
                        of false signals. They typically do not start to occur until attenuation
                        is quite a bit below the point where the indicator flashes occasionally.

                        There is a BIG difference between Perseus and SDR-14/SDR-IQ when there
                        are extremely strong signals present. Strong enough to force you to
                        place the antenna noise very low, only a couple of dB above the internal
                        noise floor. The false signals in the SDR-xx are then about 25 dB
                        stronger than the false signals in the Perseus!! Attenuating does not
                        improve, on the contrary. Adding 25 dB gain would remopve the false
                        signals if 25 dB gain were allowed without saturation due to the strong
                        signals. The phenomena are very complex because a strong signal can
                        actually remove false signals that are generated by other undesired
                        signals of modest amplitude (dithering)

                        Under normal circumstances with normal VHF converters one places the
                        noise floor 25 dB above the internal noise of the SDR in order to
                        have a rise in the noise floor of 13 to 15 dB when connecting the
                        preamplifier. With the 25 dB extra noise that is needed to get
                        an adequate front end noise figure, the advantages of Perseus
                        are hidden in the noise almost completely.

                        73

                        Leif / SM5BSZ
                      • Marco IK1ODO
                        ... John, correct. Basically very similar. QS1R lacks a good frontend, so has less sensitivity. The dynamic range is very similar; it goes higher in F than
                        Message 12 of 12 , Sep 2, 2009
                        • 0 Attachment
                          At 21.42 01/09/2009, you wrote:
                          >
                          >
                          >Still perfect Marco, I got it. Now I just need to understand the
                          >difference between Perseus and Quicksilver, which essentially seems
                          >to be component and implementation differences rather than basic design.
                          >
                          >Thanks again, I'll be quiet now ;-)

                          John,

                          correct. Basically very similar. QS1R lacks a good frontend, so has
                          less sensitivity. The dynamic range is very similar; it goes higher
                          in F than Perseus, and has a direct ADC input.
                          A more open design, but still not completed. You may also have a look
                          at the HPSDR Mercury for a complete panorama :-)

                          73 - Marco IK1ODO
                        Your message has been successfully submitted and would be delivered to recipients shortly.