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

Re: [softrock40] KX3 and Si570 tuneing step noise

Expand Messages
  • Sid Boyce
    AA0ZZ detailed the issue back in 2010 in this email and Tony s ATTiny85 code is a downlevel version of Fred s.
    Message 1 of 34 , Dec 8, 2012
    • 0 Attachment
      AA0ZZ detailed the issue back in 2010 in this email and Tony's ATTiny85 code is a downlevel version of Fred's.
      ===============================================================
      Message: 19
      Date: Thu, 28 Oct 2010 19:53:12 -0500
      From: "Craig Johnson, AA0ZZ" <aa0zz@...>
      Subject: Re: [QRP-L] Si598
      To: "QRP-L" <qrp-l@...>
      Message-ID: <F6468F27BDE04088B0F390931A17CB1C@P8>
      Content-Type: text/plain; format=flowed; charset="iso-8859-1";
              reply-type=original
      
      Hi Paul,
      
      
      Anyone taken a look at the new Si598 yet?
      Reported to be an improved Si570 and at a lower price.
      
      Paul, 5B8BA
      
      Yes!  I just took a 1-month "detour" to investigate this even-newer
      part because it
      looked like there were some significant advantages in switching from
      the Si570 to the
      Si598,  The Si598 is several dollars cheaper, it looks like Mouser is
      going to stock
      them (at least some varieties) and it looked like the Si598 had a new
      capability - the
      FREEZE-M - which was not in the Si570.  The FREEZE-M (not to be
      confused with the
      FREEZE-DCO which must always be used for large frequency excursions
      and causes the 10ms
      RF dropouts) is used for small frequency changes with the DCO
      continuing to run.  The
      FREEZE-M prevents the "spraying" of RF of unknown frequencies while
      the six control
      bytes are being loaded into the device.  This capability was new in
      the Si598 spec and
      was not in the Si570 spec. I thought these reasons were sufficient so
      I was ready to
      jump.  I updated the tables in my code (to accommodate the new
      internal clock
      frequency), bought some parts from Mouser and gave it a spin.  Yes, it
      works fine.
      The fact that I am using the CMOS variety, the max frequency of 160
      MHz was the same
      as for the Si570 so it was not an issue.
      
      However, as of a few days ago I have just switched back to the Si570.
      The reason -
      performance.  At one time my "friendly local salesman" told me that he
      had heard that
      the Si570 was the "Cadillac" and the Si598 is the "Chevy" but couldn't
      give me details
      as to why.  In making the change, I had a minor issue getting the
      FREEZE-M capability
      in the Si598 to work properly so I had a good Email discussion with a
      Silicon Labs tech
      support guy.  After he answered my questions and I was able to get it
      working, I asked
      him about the advantages of the Si570 over the Si598.  He told me that
      the Si570 has
      lower phase noise than the Si598 primarily due to crystal selection.
      The Si570 uses a
      3rd overtone 114.285 MHz crystal as a reference versus the more
      economical 39.17 MHz
      fundamental mode crystal used by the Si598. Therefore the update rate
      of the Si570's
      PLL is higher than the Si598's PLL.  Generally speaking, the higher
      the update rate,
      the higher the S/N ratio, and the lower the overall phase noise.  This
      is most
      noticeable in the mid-range offset frequencies where phase detector
      noise contributes.
      
      By the way, he also told me that the Si570 ALSO has the FREEZE-M
      capability; it is shown
      in the latest version of the spec sheet and just wasn't in the older
      version of the spec
      sheet that I was using.  That meant that the only real difference that
      remained was the
      price.
      
      Granted, the Si598 performance may be perfectly sufficient for some
      applications but I
      am not willing to sacrifice better phase noise for a couple of
      dollars.
      
      Best Regards,
      -Craig, AA0ZZ
      
      =================================================================================================================
      73 ... Sid.
      

      On 08/12/12 16:56, MIKE DURKIN wrote:
       

      Perhaps the large jumps are broken up into smaller ones .... else the blips are removed in dsp in the tented
      Just guessing...


      To: softrock40@yahoogroups.com
      From: sboyce@...
      Date: Sat, 8 Dec 2012 07:51:18 +0000
      Subject: Re: [softrock40] KX3 and Si570 tuneing step noise

       
      On 08/12/12 07:43, Alan wrote:
       


      ----- Original Message -----
      Subject: Re: [softrock40] KX3 and Si570 tuneing step noise

      >
      > There is a signal that is used to hold off the si570 lock until some
      > period after it settles, this signal was missing from the early si570's.
      >

      Sid,

      The current data sheet says
      "Note that changing the DCO frequency
      outside of the ±3500 ppm window will cause the output
      to momentarily stop and restart at any arbitrary point in
      a clock cycle. Devices sensitive to glitches or runt
      pulses may have to be reset once reconfiguration is
      complete."

      Which seems to say that the switch is not likely to be completely smooth.

      I was under the impression that Fred uses the lock code but I still see/hear jumps when large changes are made.
      Don't you?

      73 Alan G4ZFQ


      Hi Alan,
      I think you are correct about Fred's code.

      I haven't had that problem on the UHFSDR/SDR-Widget.
      I'll have to fire up the SR63ng/Mobo/SDR-Widget and check it. It was either greatly reduced or absent with the latest Widget code.
      73 ... Sid.

      -- 
      Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
      Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
      Senior Staff Specialist, Cricket Coach
      Microsoft Windows Free Zone - Linux used for all Computing Tasks
      


      -- 
      Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
      Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
      Senior Staff Specialist, Cricket Coach
      Microsoft Windows Free Zone - Linux used for all Computing Tasks
      
    • Alan
      ... Subject: [softrock40] Re: KX3 and Si570 tuneing step noise ... Paul, Please let us know when you have tested this. And may I remind anyone with a recent
      Message 34 of 34 , Jan 4, 2013
      • 0 Attachment
        ----- Original Message -----
        Subject: [softrock40] Re: KX3 and Si570 tuneing step noise


        > I have built some three standalone SI570 based oscillators and they all suffer from tuning noise, as you claim.
        >
        > I have built another one who claimS to have eliminated this noise, but I have not tested it yet. He claims to have adjusted the
        > Freeze M DCO:
        >

        Paul,

        Please let us know when you have tested this.

        And may I remind anyone with a recent Softrock with a Si570 manufactured since the beginning of 2011 that an AVR USB controller with
        V 15.15 firmware has the Freeze_M feature. The AVR would need replacing because Tony still uses V 15.12. Fred would be very
        interested if this could be tested.

        73 Alan G4ZFQ

        > http://cbjohn.com/aa0zz/index.html
        >

        >>
        >> Yes that's what got me thinking, looking at the block diagram of the KX3
        >> Hi Fred,
        >> which can be downloaded at the address below
        >> http://www.elecraft.com/manual/KX3%20Manual%20Block%20Diagram.pdf is
        >>
        >> It looks almost like a softrock rx from the front end filters to the
        >> opamp before the DSP,
        >> I have searched for the schematic but no luck so far. If anyone has a
        >> KX3 can you please
        >> verify that on large tuning steps there is no chuffing . hickups ?
        >>
        >> Kelvin
        >>
        >> > I don't think the Freeze-M bit will remove the hick-up of the
        >> receiver,
        >> > for changing freq. (at TX (fsk) it is mandatory!), it can make the
        >> time a
        >> > little smaller, maybe.
        >> > At the block schema the mixing of the KX3 looks like the softrock,
        >> even
        >> > with the 2m osc signal from the si570. I expect there is no detail
        >> internal
        >> > schema available. When using the I&Q outputs on a PC with HDSDR it
        >> > maybe easier to see on the screen.
      Your message has been successfully submitted and would be delivered to recipients shortly.