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

RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

Expand Messages
  • MIKE DURKIN
    It might be listed .. but not available to absorb .... its actually linked to the MicEqu.zip Mike KC7NOA To: softrock40@yahoogroups.com From:
    Message 1 of 12 , Feb 1, 2013
      It might be listed .. but not available to absorb .... its actually linked to the MicEqu.zip

      Mike KC7NOA




      To: softrock40@yahoogroups.com
      From: alshovk@...
      Date: Fri, 1 Feb 2013 23:12:11 -0500
      Subject: Re: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

       
      Hi Joel,

      Image suppression in SoftRock before the software correction is around 40 dB,
      the rest depends on the correction algorithm. The programs that use manual
      correction at a single point have balance errors varying over the passband in
      the range of 50-70 dB, while automatic correction, such as the one in Rocky, was
      measured to achieve 90+ dB of image suppression. Now that the source code of the
      Rocky balance correction module is available on my web site (dxatlas.com/dev),
      you can expect more programs to have good I/Q balance.

      73 Alex VE3NEA

      -----Original Message-----
      From: joelweder
      Sent: Friday, February 01, 2013 9:00 PM
      To: softrock40@yahoogroups.com
      Subject: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

      I'm currently testing the receiver of my Ensemble RxTx with various software.
      (currently enjoying the simplicity and reliability of HDSDR) I heard 9K2MU at
      14.220 this morning which was great fun!

      I've tweaked the I/Q amplitude and phase adjusters to get the best image
      rejection I can. At 10 MHz while monitoring WWV I have gotten to a little better
      than 50 dB of rejection after considerable fussing. At 5 MHz and 15 MHz the
      numbers aren't quite so good, perhaps 35-40 dB. (though I assume the rejection
      would need to be fine tuned as you move from band to band?)

      My question is this: What is a reasonable rejection specification? Not the
      ideal, but a value that is seen by most successful builders? Am I in the right
      ballpark? Somewhere I seem to recall a post from Tony talking about something
      like 75 dB but I can't get near that.

      Joel Weder
      VE6EI


    • Simon G4ELI/HB9DRV
      Yes, The link is wrong. It would be interesting to see what you are doing Alex, myself I m using a brute force and little intelligence approach, simply
      Message 2 of 12 , Feb 2, 2013

        Yes,

         

        The link is wrong. It would be interesting to see what you are doing Alex, myself I’m using a ‘brute force and little intelligence’ approach, simply continually iterating to nirvana.

         

        Simon G4ELI/HB9DRV

        http://dit-dit-dit.com

         

        From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf Of MIKE DURKIN

         

        It might be listed .. but not available to absorb .... its actually linked to the MicEqu.zip

      • Alex VE3NEA
        Fixed, here is a direct link: (http://dxatlas.com/Dev/Files/Dsp/IqBal.zip) 73 Alex VE3NEA ... From: MIKE DURKIN Sent: Saturday, February 02, 2013 2:56 AM To:
        Message 3 of 12 , Feb 2, 2013
          Fixed, here is a direct link: (http://dxatlas.com/Dev/Files/Dsp/IqBal.zip)

          73 Alex VE3NEA



          -----Original Message-----
          From: MIKE DURKIN
          Sent: Saturday, February 02, 2013 2:56 AM
          To: softrock
          Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?




          It might be listed .. but not available to absorb .... its actually linked to
          the MicEqu.zip

          Mike KC7NOA





          To: softrock40@yahoogroups.com
          From: alshovk@...
          Date: Fri, 1 Feb 2013 23:12:11 -0500
          Subject: Re: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?


          Hi Joel,

          Image suppression in SoftRock before the software correction is around 40 dB,
          the rest depends on the correction algorithm. The programs that use manual
          correction at a single point have balance errors varying over the passband in
          the range of 50-70 dB, while automatic correction, such as the one in Rocky, was
          measured to achieve 90+ dB of image suppression. Now that the source code of the
          Rocky balance correction module is available on my web site (dxatlas.com/dev),
          you can expect more programs to have good I/Q balance.

          73 Alex VE3NEA

          -----Original Message-----
          From: joelweder
          Sent: Friday, February 01, 2013 9:00 PM
          To: mailto:softrock40%40yahoogroups.com
          Subject: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

          I'm currently testing the receiver of my Ensemble RxTx with various software.
          (currently enjoying the simplicity and reliability of HDSDR) I heard 9K2MU at
          14.220 this morning which was great fun!

          I've tweaked the I/Q amplitude and phase adjusters to get the best image
          rejection I can. At 10 MHz while monitoring WWV I have gotten to a little better
          than 50 dB of rejection after considerable fussing. At 5 MHz and 15 MHz the
          numbers aren't quite so good, perhaps 35-40 dB. (though I assume the rejection
          would need to be fine tuned as you move from band to band?)

          My question is this: What is a reasonable rejection specification? Not the
          ideal, but a value that is seen by most successful builders? Am I in the right
          ballpark? Somewhere I seem to recall a post from Tony talking about something
          like 75 dB but I can't get near that.

          Joel Weder
          VE6EI
        • MIKE DURKIN
          starting from assumed no error in phase/balance .... are you going by a random %(+/-) of max assumed possible error correction and checking if there was an
          Message 4 of 12 , Feb 2, 2013
            starting from assumed no error in phase/balance ....
            are you going by a random %(+/-) of max assumed possible error correction and checking if there was an improvement ...
            or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular wire in a bundle -- that some bone head before you didn't mark or label.

            just curious .... like George.

            Mike KC7NOA

            To: softrock40@yahoogroups.com
            From: simon@...
            Date: Sat, 2 Feb 2013 11:31:08 +0100
            Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

             

            Yes,

             

            The link is wrong. It would be interesting to see what you are doing Alex, myself I’m using a ‘brute force and little intelligence’ approach, simply continually iterating to nirvana.

             

            Simon G4ELI/HB9DRV

            http://dit-dit-dit.com

             

            From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf Of MIKE DURKIN

             

            It might be listed .. but not available to absorb .... its actually linked to the MicEqu.zip


          • Alex VE3NEA
            Hi Mike, Rocky does not search for the balance, it computes the correction factors from the observed signal and image amplitudes directly, using the closed
            Message 5 of 12 , Feb 2, 2013
              Hi Mike,

              Rocky does not search for the balance, it computes the correction factors from
              the observed signal and image amplitudes directly, using the closed form
              expressions. It then interpolates the correction factor as a function of
              frequency, and constructs a correction filter. The theory is described on the
              Rocky web page: (http://dxatlas.com/Rocky/Advanced.asp)

              73 Alex VE3NEA




              -----Original Message-----
              From: MIKE DURKIN
              Sent: Saturday, February 02, 2013 11:34 AM
              To: softrock
              Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?




              starting from assumed no error in phase/balance ....
              are you going by a random %(+/-) of max assumed possible error correction and
              checking if there was an improvement ...
              or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular
              wire in a bundle -- that some bone head before you didn't mark or label.

              just curious .... like George.

              Mike KC7NOA


              To: softrock40@yahoogroups.com
              From: simon@...
              Date: Sat, 2 Feb 2013 11:31:08 +0100
              Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?





              Yes,



              The link is wrong. It would be interesting to see what you are doing Alex,
              myself I’m using a ‘brute force and little intelligence’ approach, simply
              continually iterating to nirvana.





              Simon G4ELI/HB9DRV

              http://dit-dit-dit.com





              From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf
              Of MIKE DURKIN





              It might be listed .. but not available to absorb .... its actually linked to
              the MicEqu.zip
            • Simon G4ELI/HB9DRV
              Mike, I continually iterate the correction values using spare CPU cycles. Alex s approach is 1,000% better, first I must read the legalese in the MPL licence
              Message 6 of 12 , Feb 2, 2013

                Mike,

                 

                I continually iterate the correction values using spare CPU cycles. Alex’s approach is 1,000% better, first I must read the legalese in the MPL licence before I consider using it.

                 

                Simon G4ELI/HB9DRV

                http://dit-dit-dit.com

                 

                From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf Of MIKE DURKIN

                 

                starting from assumed no error in phase/balance ....
                are you going by a random %(+/-) of max assumed possible error correction and checking if there was an improvement ...
                or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular wire in a bundle -- that some bone head before you didn't mark or label.

              • Alex VE3NEA
                Hi Simon, MPL is simple: if you want to use it as it is, just use it; if you improve it, share your improvements. 73 Alex VE3NEA ... From: Simon G4ELI/HB9DRV
                Message 7 of 12 , Feb 2, 2013
                  Hi Simon,

                  MPL is simple: if you want to use it as it is, just use it; if you improve it,
                  share your improvements.

                  73 Alex VE3NEA



                  -----Original Message-----
                  From: Simon G4ELI/HB9DRV
                  Sent: Saturday, February 02, 2013 1:52 PM
                  To: softrock40@yahoogroups.com
                  Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?






                  Mike,



                  I continually iterate the correction values using spare CPU cycles. Alex’s
                  approach is 1,000% better, first I must read the legalese in the MPL licence
                  before I consider using it.





                  Simon G4ELI/HB9DRV

                  http://dit-dit-dit.com





                  From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf
                  Of MIKE DURKIN





                  starting from assumed no error in phase/balance ....
                  are you going by a random %(+/-) of max assumed possible error correction and
                  checking if there was an improvement ...
                  or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular
                  wire in a bundle -- that some bone head before you didn't mark or label.
                • Simon G4ELI/HB9DRV
                  Alex, Thanks. I’ll spend some cycles and see how it goes… Simon G4ELI/HB9DRV http://dit-dit-dit.com From: softrock40@yahoogroups.com
                  Message 8 of 12 , Feb 2, 2013

                    Alex,

                     

                    Thanks. I’ll spend some cycles and see how it goes…

                     

                    Simon G4ELI/HB9DRV

                    http://dit-dit-dit.com

                     

                    From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf Of Alex VE3NEA
                    Sent: 02 February 2013 21:13
                    To: softrock40@yahoogroups.com
                    Subject: Re: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

                     

                     

                    Hi Simon,

                    MPL is simple: if you want to use it as it is, just use it; if you improve it,
                    share your improvements.

                    73 Alex VE3NEA

                    -----Original Message-----
                    From: Simon G4ELI/HB9DRV
                    Sent: Saturday, February 02, 2013 1:52 PM
                    To: softrock40@yahoogroups.com
                    Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

                    Mike,

                    I continually iterate the correction values using spare CPU cycles. Alex’s
                    approach is 1,000% better, first I must read the legalese in the MPL licence
                    before I consider using it.

                    Simon G4ELI/HB9DRV

                    http://dit-dit-dit.com

                    From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf
                    Of MIKE DURKIN

                    starting from assumed no error in phase/balance ....
                    are you going by a random %(+/-) of max assumed possible error correction and
                    checking if there was an improvement ...
                    or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular
                    wire in a bundle -- that some bone head before you didn't mark or label.

                  • MIKE DURKIN
                    delphi ... i had a hard time following the code ... i was using Qt5 just to view it. To: softrock40@yahoogroups.com From: alshovk@dxatlas.com Date: Sat, 2 Feb
                    Message 9 of 12 , Feb 3, 2013
                      delphi ...

                      i had a hard time following the code ... i was  using Qt5 just to view it.


                      To: softrock40@yahoogroups.com
                      From: alshovk@...
                      Date: Sat, 2 Feb 2013 15:13:14 -0500
                      Subject: Re: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

                       
                      Hi Simon,

                      MPL is simple: if you want to use it as it is, just use it; if you improve it,
                      share your improvements.

                      73 Alex VE3NEA

                      -----Original Message-----
                      From: Simon G4ELI/HB9DRV
                      Sent: Saturday, February 02, 2013 1:52 PM
                      To: softrock40@yahoogroups.com
                      Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?

                      Mike,

                      I continually iterate the correction values using spare CPU cycles. Alex’s
                      approach is 1,000% better, first I must read the legalese in the MPL licence
                      before I consider using it.

                      Simon G4ELI/HB9DRV

                      http://dit-dit-dit.com

                      From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf
                      Of MIKE DURKIN

                      starting from assumed no error in phase/balance ....
                      are you going by a random %(+/-) of max assumed possible error correction and
                      checking if there was an improvement ...
                      or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular
                      wire in a bundle -- that some bone head before you didn't mark or label.


                    • MIKE DURKIN
                      ahhhh ... ok ill look at that next ... thanks for the heads up.. ... ahhhh ... ok ill look at that next ... thanks for the heads up.. To:
                      Message 10 of 12 , Feb 3, 2013
                        ahhhh ... ok ill look at that next ... thanks for the heads up..

                        > To: softrock40@yahoogroups.com
                        > From: alshovk@dxatlas.com
                        > Date: Sat, 2 Feb 2013 13:48:51 -0500
                        > Subject: Re: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?
                        >
                        > Hi Mike,
                        >
                        > Rocky does not search for the balance, it computes the correction factors from
                        > the observed signal and image amplitudes directly, using the closed form
                        > expressions. It then interpolates the correction factor as a function of
                        > frequency, and constructs a correction filter. The theory is described on the
                        > Rocky web page: (http://dxatlas.com/Rocky/Advanced.asp)
                        >
                        > 73 Alex VE3NEA
                        >
                        >
                        >
                        >
                        > -----Original Message-----
                        > From: MIKE DURKIN
                        > Sent: Saturday, February 02, 2013 11:34 AM
                        > To: softrock
                        > Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?
                        >
                        >
                        >
                        >
                        > starting from assumed no error in phase/balance ....
                        > are you going by a random %(+/-) of max assumed possible error correction and
                        > checking if there was an improvement ...
                        > or ... i was thinking 1/3 2/3 ... like you would when hunting for a paticular
                        > wire in a bundle -- that some bone head before you didn't mark or label.
                        >
                        > just curious .... like George.
                        >
                        > Mike KC7NOA
                        >
                        >
                        > To: softrock40@yahoogroups.com
                        > From: simon@sdr-radio.com
                        > Date: Sat, 2 Feb 2013 11:31:08 +0100
                        > Subject: RE: [softrock40] Reasonable Image Rejection in an Ensemble RxTx ?
                        >
                        >
                        >
                        >
                        >
                        > Yes,
                        >
                        >
                        >
                        > The link is wrong. It would be interesting to see what you are doing Alex,
                        > myself I’m using a ‘brute force and little intelligence’ approach, simply
                        > continually iterating to nirvana.
                        >
                        >
                        >
                        >
                        >
                        > Simon G4ELI/HB9DRV
                        >
                        > http://dit-dit-dit.com
                        >
                        >
                        >
                        >
                        >
                        > From: softrock40@yahoogroups.com [mailto:softrock40@yahoogroups.com] On Behalf
                        > Of MIKE DURKIN
                        >
                        >
                        >
                        >
                        >
                        > It might be listed .. but not available to absorb .... its actually linked to
                        > the MicEqu.zip
                        >
                        >
                        >
                        >
                        >
                        >
                        >
                        >
                        >
                        >
                        >
                        > ------------------------------------
                        >
                        > Yahoo! Groups Links
                        >
                        > <*> To visit your group on the web, go to:
                        > http://groups.yahoo.com/group/softrock40/
                        >
                        > <*> Your email settings:
                        > Individual Email | Traditional
                        >
                        > <*> To change settings online go to:
                        > http://groups.yahoo.com/group/softrock40/join
                        > (Yahoo! ID required)
                        >
                        > <*> To change settings via email:
                        > softrock40-digest@yahoogroups.com
                        > softrock40-fullfeatured@yahoogroups.com
                        >
                        > <*> To unsubscribe from this group, send an email to:
                        > softrock40-unsubscribe@yahoogroups.com
                        >
                        > <*> Your use of Yahoo! Groups is subject to:
                        > http://docs.yahoo.com/info/terms/
                        >
                      Your message has been successfully submitted and would be delivered to recipients shortly.