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

Si570 frequency ranges

Expand Messages
  • Kees
    I previously tested a Si570 LVDS part which is a BBC000141G which means 3.3V LVDS, 10-280Mhz supported (correction from before), 90ppm/V tuning slope. I see
    Message 1 of 7 , Aug 31, 2008
      I previously tested a Si570 LVDS part which is a "BBC000141G" which
      means 3.3V LVDS, 10-280Mhz supported (correction from before),
      90ppm/V tuning slope. I see basically the same results 3.5Mhz to
      473Mhz looks fine, some "default" frequency from 474Mhz to 485Mhz, OK
      from 486Mhz to 566Mhz, some "default" frequency from 567Mhz to
      607Mhz, OK from 608Mhz to 709Mhz.

      Now what's interesting is that the Si570 CMOS part which is
      a "CAC000141G" has the same breaks at 474Mhz to 485Mhz, 567Mhz to
      607Mhz, and for 710Mhz and up. Guess that's to be expected if it has
      the same internals in front of the output drivers. Now that's the --
      CMOS-- part operating way outside it's range. Tuning up to 702.1Mhz
      it's a nice sine wave and the scope measures 704.25Mhz (little
      calibration issue). Of course, at these Si570 "out of spec" ranges
      other spec parameters are subject to question.

      ********************
      However, what's even more interesting is that the "A" parts have a
      frequency range of 10-945Mhz, 970-1134Mhz, and 1213-
      1417.5Mhz ....that's exactly 2x the numbers we observed with the "C"
      parts and apparently a basic characteristic of the Si570. If we take
      the "B" part with a max 810Mhz and divide by 2 we get
      405Mhz .....didn't someone have his Si570 quit around 400Mhz ?

      I believe all these results would be indicative of, and normal for,
      a "speed sort".

      73 Kees K5BCQ
    • Alan
      ... From: Kees ... I can t help feeling John needs to look at the software again. I will know for sure when I get my replacement. The Si570 I have been
      Message 2 of 7 , Aug 31, 2008
        ----- Original Message -----
        From: "Kees"
        >
        > Now what's interesting is that the Si570 CMOS part which is
        > a "CAC000141G" has the same breaks at 474Mhz to 485Mhz, 567Mhz to
        > 607Mhz, and for 710Mhz and up.

        I can't help feeling John needs to look at the software again. I will know
        for sure when I get my replacement.
        The Si570 I have been testing is a 10MHz startup BBA, allowing for your
        scope's calibration these breaks are perilously close to what my frequency
        counter and ICR7000 are saying with your original IC.
        Using a PIC I can fill in these gaps (and detect signals up to 1.4GHz).

        73 Alan G4ZFQ
      • Kees & Sandy
        Alan, Please check my most recent note about the gaps, originally found by Jose Bonanca, matching the Silicon Labs spec d gaps for the A parts (times two).
        Message 3 of 7 , Aug 31, 2008

          Alan,

          Please check my most recent note about the gaps, originally found by Jose Bonanca, matching the Silicon Labs spec'd gaps for the "A" parts (times two). That data made John and I feel much better about the code and board layout, because, from all indications, those particular gaps are a characteristic of the Si570 part. All the parts, LVDS and CMOS, we have tested here are "C" parts as provided by Tom Hoflich. No telling what you can actually squeeze out of an "A" part. You "A" part owners will have to tell us that. 

          73 Kees K5BCQ

          -- "Alan" <g4zfq@...> wrote:


          ----- Original Message -----
          From: "Kees"
          >
          > Now what's interesting is that the Si570 CMOS part which is
          > a "CAC000141G" has the same breaks at 474Mhz to 485Mhz, 567Mhz to
          > 607Mhz, and for 710Mhz and up.

          I can't help feeling John needs to look at the software again. I will know
          for sure when I get my replacement.
          The Si570 I have been testing is a 10MHz startup BBA, allowing for your
          scope's calibration these breaks are perilously close to what my frequency
          counter and ICR7000 are saying with your original IC.
          Using a PIC I can fill in these gaps (and detect signals up to 1.4GHz).

          73 Alan G4ZFQ

        • jhops99
          ... Kees, Tried the bad frequency ranges on my BBC000141G but using the UBW Kenwood Si570 control software and didn t see any problem. Tried some 10kHz steps
          Message 4 of 7 , Aug 31, 2008
            --- In softrock40@yahoogroups.com, "Kees" <windy10605@...> wrote:
            >
            > I previously tested a Si570 LVDS part which is a "BBC000141G" which
            > means 3.3V LVDS, 10-280Mhz supported (correction from before),
            > 90ppm/V tuning slope. I see basically the same results 3.5Mhz to
            > 473Mhz looks fine, some "default" frequency from 474Mhz to 485Mhz, OK
            > from 486Mhz to 566Mhz, some "default" frequency from 567Mhz to
            > 607Mhz, OK from 608Mhz to 709Mhz.
            >
            >
            > 73 Kees K5BCQ
            >
            Kees,
            Tried the bad frequency ranges on my BBC000141G but using the UBW
            Kenwood Si570 control software and didn't see any problem. Tried some
            10kHz steps in each range.

            73 Joe WB8DNO
          • John H. Fisher
            This might be considered inconclusive because you didn t test with Kees s Si570 and Kees didn t test with your Si570. So you can t say it s the controller. It
            Message 5 of 7 , Aug 31, 2008
              This might be considered inconclusive because you didn't test with
              Kees's Si570 and Kees didn't test with your Si570. So you can't say
              it's the controller. It could still be that some Si570's work better
              than others.

              Regards,
              John

              --- In softrock40@yahoogroups.com, "jhops99" <jhops@...> wrote:
              >
              > --- In softrock40@yahoogroups.com, "Kees" <windy10605@> wrote:
              > >
              > > I previously tested a Si570 LVDS part which is a "BBC000141G" which
              > > means 3.3V LVDS, 10-280Mhz supported (correction from before),
              > > 90ppm/V tuning slope. I see basically the same results 3.5Mhz to
              > > 473Mhz looks fine, some "default" frequency from 474Mhz to 485Mhz, OK
              > > from 486Mhz to 566Mhz, some "default" frequency from 567Mhz to
              > > 607Mhz, OK from 608Mhz to 709Mhz.
              > >
              > >
              > > 73 Kees K5BCQ
              > >
              > Kees,
              > Tried the bad frequency ranges on my BBC000141G but using the UBW
              > Kenwood Si570 control software and didn't see any problem. Tried some
              > 10kHz steps in each range.
              >
              > 73 Joe WB8DNO
              >
            • Kees & Sandy
              What I see using a CAC000141G CMOS part (the LVDS part will be the same) stepped at 10Khz increments is a nice sine wave at 473Mhz (scope indicates
              Message 6 of 7 , Aug 31, 2008

                What I see using a "CAC000141G" CMOS part (the LVDS part will be the same) stepped at 10Khz increments is a nice sine wave at 473Mhz (scope indicates 469.4Mhz). Tuning beyond that shows a very unstable 56Mhz (according to the scope) "square" wave (sound like a familiar frequency ?) of much greater amplitude which then changes to an unstable ~113-130Mhz sine wave. This is when the part has definitely lost lock. At 486Mhz the output returns to a nice sine wave.

                Again, this is just for interest and well beyond where the part is spec'd. And well beyond the frequency range where nearly all people would use a "C" speed part. The Silicon Labs spec sheet for "A" speed parts shows 10-945Mhz, 970-1134Mhz, and 1213-1417.5Mhz. Please note the frequency "gaps" where it is spec'd to NOT be operational.   

                73 Kees K5BCQ

                -- "John H. Fisher" <k5jhf@...> wrote:

                This might be considered inconclusive because you didn't test with
                Kees's Si570 and Kees didn't test with your Si570. So you can't say
                it's the controller. It could still be that some Si570's work better
                than others.

                Regards,
                John

                --- In softrock40@yahoogro ups.com, "jhops99" <jhops@...> wrote:
                >
                > --- In softrock40@yahoogro ups.com, "Kees" <windy10605@ > wrote:
                > >
                > > I previously tested a Si570 LVDS part which is a "BBC000141G" which
                > > means 3.3V LVDS, 10-280Mhz supported (correction from before),
                > > 90ppm/V tuning slope. I see basically the same results 3.5Mhz to
                > > 473Mhz looks fine, some "default" frequency from 474Mhz to 485Mhz, OK
                > > from 486Mhz to 566Mhz, some "default" frequency from 567Mhz to
                > > 607Mhz, OK from 608Mhz to 709Mhz.
                > >
                > >
                > > 73 Kees K5BCQ
                > >
                > Kees,
                > Tried the bad frequency ranges on my BBC000141G but using the UBW
                > Kenwood Si570 control software and didn't see any problem. Tried some
                > 10kHz steps in each range.
                >
                > 73 Joe WB8DNO
                >

              • tom_iphi
                Hi all, I have tested an LVDS connected to a spectrum analyzer on my USB controller and didn t see any problems in the whole range up to 1.4GHz. Of course, the
                Message 7 of 7 , Aug 31, 2008
                  Hi all,

                  I have tested an LVDS connected to a spectrum analyzer on my USB
                  controller and didn't see any problems in the whole range up to
                  1.4GHz. Of course, the spec gaps are there. They are due to the VCO
                  range and possible divider combinations. I don't understand, why
                  SILabs don't spec down to 3.5MHz, as a simple calculation reveals,
                  that this is possible without running the VCO outside spec.

                  I would assume, that the DSPLL core is identical for all SI570s,
                  likely CMOS for all. I remember from a visit to SiLabs some time ago,
                  that they stated, they would do all in CMOS because of lower cost and
                  they were proud that they could do things in CMOS that other people
                  couldn't. I also assume, that there are just different output buffers
                  for the different types. This would also make sense from a
                  development and production cost point of view.

                  BTW: If you hold a fast CMOS buffer close to threshold it is likely
                  to oscillate.

                  73's Tom DG8SAQ

                  --- In softrock40@yahoogroups.com, "Kees & Sandy" <windy10605@...>
                  wrote:
                  >
                  > What I see using a "CAC000141G" CMOS part (the LVDS part will be
                  the same) stepped at 10Khz increments is a nice sine wave at 473Mhz
                  (scope indicates 469.4Mhz). Tuning beyond that shows a very unstable
                  56Mhz (according to the scope) "square" wave (sound like a familiar
                  frequency ?) of much greater amplitude which then changes to an
                  unstable ~113-130Mhz sine wave. This is when the part has definitely
                  lost lock. At 486Mhz the output returns to a nice sine wave.
                  > Again, this is just for interest and well beyond where the part is
                  spec'd. And well beyond the frequency range where nearly all people
                  would use a "C" speed part. The Silicon Labs spec sheet for "A" speed
                  parts shows 10-945Mhz, 970-1134Mhz, and 1213-1417.5Mhz. Please note
                  the frequency "gaps" where it is spec'd to NOT be operational.
                  > 73 Kees K5BCQ
                  >
                  > -- "John H. Fisher" <k5jhf@...> wrote:
                  >
                  > This might be considered inconclusive because you didn't test with
                  > Kees's Si570 and Kees didn't test with your Si570. So you can't say
                  > it's the controller. It could still be that some Si570's work better
                  > than others.
                  >
                  > Regards,
                  > John
                  >
                  > --- In softrock40@yahoogroups.com, "jhops99" <jhops@> wrote:
                  > >
                  > > --- In softrock40@yahoogroups.com, "Kees" <windy10605@> wrote:
                  > > >
                  > > > I previously tested a Si570 LVDS part which is a "BBC000141G"
                  which
                  > > > means 3.3V LVDS, 10-280Mhz supported (correction from before),
                  > > > 90ppm/V tuning slope. I see basically the same results 3.5Mhz
                  to
                  > > > 473Mhz looks fine, some "default" frequency from 474Mhz to
                  485Mhz, OK
                  > > > from 486Mhz to 566Mhz, some "default" frequency from 567Mhz to
                  > > > 607Mhz, OK from 608Mhz to 709Mhz.
                  > > >
                  > > >
                  > > > 73 Kees K5BCQ
                  > > >
                  > > Kees,
                  > > Tried the bad frequency ranges on my BBC000141G but using the UBW
                  > > Kenwood Si570 control software and didn't see any problem. Tried
                  some
                  > > 10kHz steps in each range.
                  > >
                  > > 73 Joe WB8DNO
                  > >
                  >
                Your message has been successfully submitted and would be delivered to recipients shortly.