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

Re: [softrock40] audio isolation transfomers

Expand Messages
  • Jack Smith
    Comparisons of the Radio Shack transformer and Triad SP-70 at http://www.cliftonlaboratories.com/softrock_lite_6_2.htm Jack K8ZOA
    Message 1 of 10 , Jul 2 3:12 AM
    • 0 Attachment
      Comparisons of the Radio Shack transformer and Triad SP-70 at
      http://www.cliftonlaboratories.com/softrock_lite_6_2.htm

      Jack K8ZOA


      w5xz@... wrote:
      >
      >
      > i have a nasty ground loop with my emu1212; since it HAS
      > balanced line inputs, i put a couple of cheapie radio
      > shack isolation transformers (p/n 271-1374) in, and
      > re-wired to the balanced inputs, with twisted pairs only.
      > at $3 each, it was too cheap to pass up.
      >
      > now, there is a narrow but distinct notch around 0 Hz.
      > about 20 db or so. beats the heck out of that noise hump, though...
      >
      > i suppose this is low freq roll-off? pleasantly surprised to
      > see plenty of signal many many khz away from 0 hz, though.
      > plus & minus 75 khz, plus.......sampling at 192k with the emu.
      >
      > overall, rx sensitivity seems better, also. with no particular
      > attempt at impedance matching either; 1:1 transformers, supposedly
      > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
      > the emu is 15k ohms in........
      >
      > do the good triad isolation transfomers do this also?
      >
      > 73, w5xz, dan
      >
      >
    • dan edwards
      thanks, jack   i have looked at your site. great work. guess i overlooked this, though.   so, quick easy answer is No great advantage with the triad over
      Message 2 of 10 , Jul 2 4:57 AM
      • 0 Attachment
        thanks, jack
         
        i have looked at your site. great work. guess i overlooked this, though.
         
        so, quick easy answer is "No great advantage with the triad over the r.s. cheapie" ??
         
        even with a passing attempt at 'better' impedance matching? i find a variety
        of z ratio triads are available. is my read of the tech specs of the 0202 daughter
        card input z of 10k ohms correct?
         
        73, w5xz, dan


        --- On Thu, 7/2/09, Jack Smith <Jack.Smith@...> wrote:

        From: Jack Smith <Jack.Smith@...>
        Subject: Re: [softrock40] audio isolation transfomers
        To: softrock40@yahoogroups.com
        Date: Thursday, July 2, 2009, 10:12 AM

        Comparisons of the Radio Shack transformer and Triad SP-70 at
        http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm

        Jack K8ZOA

        w5xz@... wrote:
        >
        >
        > i have a nasty ground loop with my emu1212; since it HAS
        > balanced line inputs, i put a couple of cheapie radio
        > shack isolation transformers (p/n 271-1374) in, and
        > re-wired to the balanced inputs, with twisted pairs only.
        > at $3 each, it was too cheap to pass up.
        >
        > now, there is a narrow but distinct notch around 0 Hz.
        > about 20 db or so. beats the heck out of that noise hump, though...
        >
        > i suppose this is low freq roll-off? pleasantly surprised to
        > see plenty of signal many many khz away from 0 hz, though.
        > plus & minus 75 khz, plus.......sampling at 192k with the emu.
        >
        > overall, rx sensitivity seems better, also. with no particular
        > attempt at impedance matching either; 1:1 transformers, supposedly
        > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
        > the emu is 15k ohms in........
        >
        > do the good triad isolation transfomers do this also?
        >
        > 73, w5xz, dan
        >
        >
      • Jack Smith
        Dan: I believe the 10K is correct. The output impedance of the Softrock is near zero. After all, it is an op amp in a feedback circuit so it tries to maintain
        Message 3 of 10 , Jul 2 5:28 AM
        • 0 Attachment
          Dan:

          I believe the 10K is correct.

          The output impedance of the Softrock is near zero. After all, it is an
          op amp in a feedback circuit so it tries to maintain the same output
          voltage regardless of the load on it. Hence its output impedance is in
          theory zero, limited only by the op-amps ability to source or sink
          current plus the impedance of the blocking capacitor.

          So you are not matching the Softrock to the sound card's input impedance
          in any normal sense of the term 'matching.' Rather the transformer
          ratios serve as a voltage step up more than anything else.

          You can think of the sound card as if it's a voltmeter - you don't want
          to match the volt meter's high input impedance to the voltage source you
          are measuring. If you don't need the voltage step up, a 1:1 transformer
          ratio is fine. Could be 600:600 or 10K:10K. I doubt very much that you
          will find any benefit in a ratio other than 1:1 with an E MU 0202.

          Of course there are some practical issues related to stray capacitance
          in the transformer windings and how they alter the transformer's high
          frequency response into various loads. You can see some of this in my
          various transformer pages. Hence, I would stick with the 600:600
          versions and deviate from that only if you know what you are doing.

          Jack

          dan edwards wrote:
          >
          >
          > thanks, jack
          >
          > i have looked at your site. great work. guess i overlooked this, though.
          >
          > so, quick easy answer is "No great advantage with the triad over the
          > r.s. cheapie" ??
          >
          > even with a passing attempt at 'better' impedance matching? i find a
          > variety
          > of z ratio triads are available. is my read of the tech specs of the
          > 0202 daughter
          > card input z of 10k ohms correct?
          >
          > 73, w5xz, dan
          >
          >
          > --- On *Thu, 7/2/09, Jack Smith /<Jack.Smith@...>/* wrote:
          >
          >
          > From: Jack Smith <Jack.Smith@...>
          > Subject: Re: [softrock40] audio isolation transfomers
          > To: softrock40@yahoogroups.com
          > Date: Thursday, July 2, 2009, 10:12 AM
          >
          > Comparisons of the Radio Shack transformer and Triad SP-70 at
          > http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm
          > <http://www.cliftonlaboratories.com/softrock_lite_6_2.htm>
          >
          > Jack K8ZOA
          >
          > w5xz@...
          > <http://us.mc1802.mail.yahoo.com/mc/compose?to=w5xz%40att.net> wrote:
          > >
          > >
          > > i have a nasty ground loop with my emu1212; since it HAS
          > > balanced line inputs, i put a couple of cheapie radio
          > > shack isolation transformers (p/n 271-1374) in, and
          > > re-wired to the balanced inputs, with twisted pairs only.
          > > at $3 each, it was too cheap to pass up.
          > >
          > > now, there is a narrow but distinct notch around 0 Hz.
          > > about 20 db or so. beats the heck out of that noise hump, though...
          > >
          > > i suppose this is low freq roll-off? pleasantly surprised to
          > > see plenty of signal many many khz away from 0 hz, though.
          > > plus & minus 75 khz, plus.......sampling at 192k with the emu.
          > >
          > > overall, rx sensitivity seems better, also. with no particular
          > > attempt at impedance matching either; 1:1 transformers, supposedly
          > > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
          > > the emu is 15k ohms in........
          > >
          > > do the good triad isolation transfomers do this also?
          > >
          > > 73, w5xz, dan
          > >
          > >
          >
          >
        • Jeff Blaine AC0C
          Dan, Jack is the master. Lot of worse ways to waste time than camping out on his page for some review... I had an instance where the ground loop was causing a
          Message 4 of 10 , Jul 2 6:09 AM
          • 0 Attachment
            Dan,
             
            Jack is the master.  Lot of worse ways to waste time than camping out on his page for some review...
             
            I had an instance where the ground loop was causing a gigantic rise near the zero and extending up maybe 10 kHz.  Unbelievable mess.  And rather than actually track down the source of the ground loop (the shack wiring was a mess as well and that would have been a week long project), I stuck in the transformers based on Jack's review of the RS cheap transformer.  Great results - very similar to what you have seen.  A much improved noise profile - but quite deep nulls near the zero point.
             
            So I played around with the coupling cap value, and put a resistor on the sound-card side in an attempt to fill in the center a bit.  Some improvement was had - but it's not picture perfect.  The nulls are about 10db, and pretty narrow.  It looks a bit funky on the screen, but in a practical sense, it's rare that the signal of interest is right in the null - I run a 7-15khz offset depending on sound card frequency - I try to use the highest offset possible, just to minimize the effect of the zero frequency which can be minimized, but never eliminated.
             
            I want to say the cap value went from 0.1->0.2 which helped fill in null a but but gave rise to a larger "bump" near the CF.  The loading resistor was about 6K and that seemed to be about the best overall compromise value.
             
            Putting in the tx was a huge improvement.  And the cap/resistor were probably mostly window dressing, in the overall scheme.
             
            Good luck and hope this helps a bit.
             
            73/jeff/ac0c
             

             
            Sent: Thursday, July 02, 2009 8:28 PM
            Subject: Re: [softrock40] audio isolation transfomers

            Dan:

            I believe the 10K is correct.

            The output impedance of the Softrock is near zero. After all, it is an
            op amp in a feedback circuit so it tries to maintain the same output
            voltage regardless of the load on it. Hence its output impedance is in
            theory zero, limited only by the op-amps ability to source or sink
            current plus the impedance of the blocking capacitor.

            So you are not matching the Softrock to the sound card's input impedance
            in any normal sense of the term 'matching.' Rather the transformer
            ratios serve as a voltage step up more than anything else.

            You can think of the sound card as if it's a voltmeter - you don't want
            to match the volt meter's high input impedance to the voltage source you
            are measuring. If you don't need the voltage step up, a 1:1 transformer
            ratio is fine. Could be 600:600 or 10K:10K. I doubt very much that you
            will find any benefit in a ratio other than 1:1 with an E MU 0202.

            Of course there are some practical issues related to stray capacitance
            in the transformer windings and how they alter the transformer' s high
            frequency response into various loads. You can see some of this in my
            various transformer pages. Hence, I would stick with the 600:600
            versions and deviate from that only if you know what you are doing.

            Jack

            dan edwards wrote:

            >
            >
            > thanks, jack
            >
            > i have looked at
            your site. great work. guess i overlooked this, though.
            >
            > so,
            quick easy answer is "No great advantage with the triad over the
            > r.s.
            cheapie" ??
            >
            > even with a passing attempt at 'better' impedance
            matching? i find a
            > variety
            > of z ratio triads are available. is
            my read of the tech specs of the
            > 0202 daughter
            > card input z of
            10k ohms correct?
            >
            > 73, w5xz, dan
            >
            >
            > --- On
            *Thu, 7/2/09, Jack Smith /<Jack.Smith@cox. net>/* wrote:
            >
            >
            > From: Jack Smith <
            href="mailto:Jack.Smith%40cox.net">Jack.Smith@cox. net>
            >
            Subject: Re: [softrock40] audio isolation transfomers
            > To:
            href="mailto:softrock40%40yahoogroups.com">softrock40@yahoogro ups.com
            >
            Date: Thursday, July 2, 2009, 10:12 AM
            >
            > Comparisons of the Radio
            Shack transformer and Triad SP-70 at
            >
            href="http://www.cliftonl">http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm
            > <
            href="http://www.cliftonlaboratories.com/softrock_lite_6_2.htm">http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm>
            >
            >
            Jack K8ZOA
            >
            >
            href="mailto:w5xz%40att.net">w5xz@...
            > <
            href="http://us.mc1802.mail.yahoo.com/mc/compose?to=w5xz%40att.net">http://us.mc1802. mail.yahoo. com/mc/compose? to=w5xz%40att. net> wrote:
            > >
            > >
            > > i have a nasty ground loop with my
            emu1212; since it HAS
            > > balanced line inputs, i put a couple of
            cheapie radio
            > > shack isolation transformers (p/n 271-1374) in,
            and
            > > re-wired to the balanced inputs, with twisted pairs
            only.
            > > at $3 each, it was too cheap to pass up.
            > >
            > > now, there is a narrow but distinct notch around 0 Hz.
            > > about
            20 db or so. beats the heck out of that noise hump, though...
            > >
            > > i suppose this is low freq roll-off? pleasantly surprised
            to
            > > see plenty of signal many many khz away from 0 hz,
            though.
            > > plus & minus 75 khz, plus.......sampling at 192k with
            the emu.
            > >
            > > overall, rx sensitivity seems better, also.
            with no particular
            > > attempt at impedance matching either; 1:1
            transformers, supposedly
            > > i'm guessing the rxtx6.3 is about 50 ohms
            out, and i think
            > > the emu is 15k ohms in........
            > >
            > > do the good triad isolation transfomers do this also?
            > >
            > > 73, w5xz, dan
            > >
            > >
            >
            >

          • dan edwards
            jack & jeff   thanks for the feedback.   jack, my rxtx6.3 has 100 ohms in series with the outputs of the op amps, along with .1 ufd.  so, that means the
            Message 5 of 10 , Jul 2 7:12 AM
            • 0 Attachment
              jack & jeff
               
              thanks for the feedback.
               
              jack, my rxtx6.3 has 100 ohms in series with the outputs of the op amps, along
              with .1 ufd.  so, that means the output impedance is 100 ohms or so, right?
              a higher impedance ratio, in to out, would just give me a small voltage gain, as
              long as i dont create some funky resonances, right? but not really a significant
              advantage to try to match 100 to 10000 ohms?
               
              jeff: yes, i thought about switching the blocking cap to a much larger value, say
              1 or 10 ufd. this is a good place for a tantalum cap, instead of a b-flat electrolytic?
              the 3 dollar radio shack unit is right in line with the budget for this project!!! ha.
              dirt cheap and actually works pretty well.
               
              73, w5xz, dan

              --- On Thu, 7/2/09, Jeff Blaine AC0C <keepwalking188@...> wrote:

              From: Jeff Blaine AC0C <keepwalking188@...>
              Subject: Re: [softrock40] audio isolation transfomers
              To: softrock40@yahoogroups.com
              Date: Thursday, July 2, 2009, 1:09 PM

              Dan,
               
              Jack is the master.  Lot of worse ways to waste time than camping out on his page for some review...
               
              I had an instance where the ground loop was causing a gigantic rise near the zero and extending up maybe 10 kHz.  Unbelievable mess.  And rather than actually track down the source of the ground loop (the shack wiring was a mess as well and that would have been a week long project), I stuck in the transformers based on Jack's review of the RS cheap transformer.  Great results - very similar to what you have seen.  A much improved noise profile - but quite deep nulls near the zero point.
               
              So I played around with the coupling cap value, and put a resistor on the sound-card side in an attempt to fill in the center a bit.  Some improvement was had - but it's not picture perfect.  The nulls are about 10db, and pretty narrow.  It looks a bit funky on the screen, but in a practical sense, it's rare that the signal of interest is right in the null - I run a 7-15khz offset depending on sound card frequency - I try to use the highest offset possible, just to minimize the effect of the zero frequency which can be minimized, but never eliminated.
               
              I want to say the cap value went from 0.1->0.2 which helped fill in null a but but gave rise to a larger "bump" near the CF.  The loading resistor was about 6K and that seemed to be about the best overall compromise value.
               
              Putting in the tx was a huge improvement.  And the cap/resistor were probably mostly window dressing, in the overall scheme.
               
              Good luck and hope this helps a bit.
               
              73/jeff/ac0c
               

               
              Sent: Thursday, July 02, 2009 8:28 PM
              Subject: Re: [softrock40] audio isolation transfomers

              Dan:

              I believe the 10K is correct.

              The output impedance of the Softrock is near zero. After all, it is an
              op amp in a feedback circuit so it tries to maintain the same output
              voltage regardless of the load on it. Hence its output impedance is in
              theory zero, limited only by the op-amps ability to source or sink
              current plus the impedance of the blocking capacitor.

              So you are not matching the Softrock to the sound card's input impedance
              in any normal sense of the term 'matching.' Rather the transformer
              ratios serve as a voltage step up more than anything else.

              You can think of the sound card as if it's a voltmeter - you don't want
              to match the volt meter's high input impedance to the voltage source you
              are measuring. If you don't need the voltage step up, a 1:1 transformer
              ratio is fine. Could be 600:600 or 10K:10K. I doubt very much that you
              will find any benefit in a ratio other than 1:1 with an E MU 0202.

              Of course there are some practical issues related to stray capacitance
              in the transformer windings and how they alter the transformer' s high
              frequency response into various loads. You can see some of this in my
              various transformer pages. Hence, I would stick with the 600:600
              versions and deviate from that only if you know what you are doing.

              Jack

              dan edwards wrote:
              >
              >
              > thanks, jack
              >
              > i have looked at your site. great work. guess i overlooked this, though.
              >
              > so, quick easy answer is "No great advantage with the triad over the
              > r.s. cheapie" ??
              >
              > even with a passing attempt at 'better' impedance matching? i find a
              > variety
              > of z ratio triads are available. is my read of the tech specs of the
              > 0202 daughter
              > card input z of 10k ohms correct?
              >
              > 73, w5xz, dan
              >
              >
              > --- On *Thu, 7/2/09, Jack Smith /<Jack.Smith@cox. net>/* wrote:
              >
              >
              > From: Jack Smith <Jack.Smith@cox. net>
              > Subject: Re: [softrock40] audio isolation transfomers
              > To: softrock40@yahoogro ups.com
              > Date: Thursday, July 2, 2009, 10:12 AM
              >
              > Comparisons of the Radio Shack transformer and Triad SP-70 at
              > http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm
              > <http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm>
              >
              > Jack K8ZOA
              >
              > w5xz@...
              > <http://us.mc1802. mail.yahoo. com/mc/compose? to=w5xz%40att. net> wrote:
              > >
              > >
              > > i have a nasty ground loop with my emu1212; since it HAS
              > > balanced line inputs, i put a couple of cheapie radio
              > > shack isolation transformers (p/n 271-1374) in, and
              > > re-wired to the balanced inputs, with twisted pairs only.
              > > at $3 each, it was too cheap to pass up.
              > >
              > > now, there is a narrow but distinct notch around 0 Hz.
              > > about 20 db or so. beats the heck out of that noise hump, though...
              > >
              > > i suppose this is low freq roll-off? pleasantly surprised to
              > > see plenty of signal many many khz away from 0 hz, though.
              > > plus & minus 75 khz, plus.......sampling at 192k with the emu.
              > >
              > > overall, rx sensitivity seems better, also. with no particular
              > > attempt at impedance matching either; 1:1 transformers, supposedly
              > > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
              > > the emu is 15k ohms in........
              > >
              > > do the good triad isolation transfomers do this also?
              > >
              > > 73, w5xz, dan
              > >
              > >
              >
              >
            • Matt Palmer
              I d avoid tantalums, green polystyrenes are better for this. Matt W8ESE Former KD8DAO http://blog.MattIsKichigai.com
              Message 6 of 10 , Jul 2 7:15 AM
              • 0 Attachment
                I'd avoid tantalums, green polystyrenes are better for this.


                Matt
                W8ESE
                Former KD8DAO
                http://blog.MattIsKichigai.com



                On Thu, Jul 2, 2009 at 9:12 AM, dan edwards<w5xz@...> wrote:
                >
                >
                > jack & jeff
                >
                > thanks for the feedback.
                >
                > jack, my rxtx6.3 has 100 ohms in series with the outputs of the op amps,
                > along
                > with .1 ufd.  so, that means the output impedance is 100 ohms or so, right?
                > a higher impedance ratio, in to out, would just give me a small voltage
                > gain, as
                > long as i dont create some funky resonances, right? but not really a
                > significant
                > advantage to try to match 100 to 10000 ohms?
                >
                > jeff: yes, i thought about switching the blocking cap to a much larger
                > value, say
                > 1 or 10 ufd. this is a good place for a tantalum cap, instead of a b-flat
                > electrolytic?
                > the 3 dollar radio shack unit is right in line with the budget for this
                > project!!! ha.
                > dirt cheap and actually works pretty well.
                >
                > 73, w5xz, dan
                > --- On Thu, 7/2/09, Jeff Blaine AC0C <keepwalking188@...> wrote:
                >
                > From: Jeff Blaine AC0C <keepwalking188@...>
                > Subject: Re: [softrock40] audio isolation transfomers
                > To: softrock40@yahoogroups.com
                > Date: Thursday, July 2, 2009, 1:09 PM
                >
                > Dan,
                >
                > Jack is the master.  Lot of worse ways to waste time than camping out on his
                > page for some review...
                >
                > I had an instance where the ground loop was causing a gigantic rise near the
                > zero and extending up maybe 10 kHz.  Unbelievable mess.  And rather than
                > actually track down the source of the ground loop (the shack wiring was a
                > mess as well and that would have been a week long project), I stuck in the
                > transformers based on Jack's review of the RS cheap transformer.  Great
                > results - very similar to what you have seen.  A much improved noise profile
                > - but quite deep nulls near the zero point.
                >
                > So I played around with the coupling cap value, and put a resistor on the
                > sound-card side in an attempt to fill in the center a bit.  Some improvement
                > was had - but it's not picture perfect.  The nulls are about 10db, and
                > pretty narrow.  It looks a bit funky on the screen, but in a practical
                > sense, it's rare that the signal of interest is right in the null - I run a
                > 7-15khz offset depending on sound card frequency - I try to use the highest
                > offset possible, just to minimize the effect of the zero frequency which can
                > be minimized, but never eliminated.
                >
                > I want to say the cap value went from 0.1->0.2 which helped fill in null a
                > but but gave rise to a larger "bump" near the CF.  The loading resistor was
                > about 6K and that seemed to be about the best overall compromise value.
                >
                > Putting in the tx was a huge improvement.  And the cap/resistor were
                > probably mostly window dressing, in the overall scheme.
                >
                > Good luck and hope this helps a bit.
                >
                > 73/jeff/ac0c
                >
                >
                > From: Jack Smith
                > Sent: Thursday, July 02, 2009 8:28 PM
                > To: softrock40@yahoogro ups.com
                > Subject: Re: [softrock40] audio isolation transfomers
                > Dan:
                >
                > I believe the 10K is correct.
                >
                > The output impedance of the Softrock is near zero. After all, it is an
                > op amp in a feedback circuit so it tries to maintain the same output
                > voltage regardless of the load on it. Hence its output impedance is in
                > theory zero, limited only by the op-amps ability to source or sink
                > current plus the impedance of the blocking capacitor.
                >
                > So you are not matching the Softrock to the sound card's input impedance
                > in any normal sense of the term 'matching.' Rather the transformer
                > ratios serve as a voltage step up more than anything else.
                >
                > You can think of the sound card as if it's a voltmeter - you don't want
                > to match the volt meter's high input impedance to the voltage source you
                > are measuring. If you don't need the voltage step up, a 1:1 transformer
                > ratio is fine. Could be 600:600 or 10K:10K. I doubt very much that you
                > will find any benefit in a ratio other than 1:1 with an E MU 0202.
                >
                > Of course there are some practical issues related to stray capacitance
                > in the transformer windings and how they alter the transformer' s high
                > frequency response into various loads. You can see some of this in my
                > various transformer pages. Hence, I would stick with the 600:600
                > versions and deviate from that only if you know what you are doing.
                >
                > Jack
                >
                > dan edwards wrote:
                >>
                >>
                >> thanks, jack
                >>
                >> i have looked at your site. great work. guess i overlooked this, though.
                >>
                >> so, quick easy answer is "No great advantage with the triad over the
                >> r.s. cheapie" ??
                >>
                >> even with a passing attempt at 'better' impedance matching? i find a
                >> variety
                >> of z ratio triads are available. is my read of the tech specs of the
                >> 0202 daughter
                >> card input z of 10k ohms correct?
                >>
                >> 73, w5xz, dan
                >>
                >>
                >> --- On *Thu, 7/2/09, Jack Smith /<Jack.Smith@cox. net>/* wrote:
                >>
                >>
                >> From: Jack Smith <Jack.Smith@cox. net>
                >> Subject: Re: [softrock40] audio isolation transfomers
                >> To: softrock40@yahoogro ups.com
                >> Date: Thursday, July 2, 2009, 10:12 AM
                >>
                >> Comparisons of the Radio Shack transformer and Triad SP-70 at
                >> http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm
                >> <http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm>
                >>
                >> Jack K8ZOA
                >>
                >> w5xz@...
                >> <http://us.mc1802 mail.yahoo. com/mc/compose? to=w5xz%40att. net> wrote:
                >> >
                >> >
                >> > i have a nasty ground loop with my emu1212; since it HAS
                >> > balanced line inputs, i put a couple of cheapie radio
                >> > shack isolation transformers (p/n 271-1374) in, and
                >> > re-wired to the balanced inputs, with twisted pairs only.
                >> > at $3 each, it was too cheap to pass up.
                >> >
                >> > now, there is a narrow but distinct notch around 0 Hz.
                >> > about 20 db or so. beats the heck out of that noise hump, though...
                >> >
                >> > i suppose this is low freq roll-off? pleasantly surprised to
                >> > see plenty of signal many many khz away from 0 hz, though.
                >> > plus & minus 75 khz, plus.......sampling at 192k with the emu.
                >> >
                >> > overall, rx sensitivity seems better, also. with no particular
                >> > attempt at impedance matching either; 1:1 transformers, supposedly
                >> > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
                >> > the emu is 15k ohms in........
                >> >
                >> > do the good triad isolation transfomers do this also?
                >> >
                >> > 73, w5xz, dan
                >> >
                >> >
                >>
                >>
                >
                >
              • Alan
                With regard to transformer ratio, if it steps up the voltage to the soundcard then would the Softrock gain need to be reduced? Softrocks are normally sensitive
                Message 7 of 10 , Jul 2 8:14 AM
                • 0 Attachment
                  With regard to transformer ratio, if it steps up the voltage to the
                  soundcard then would the Softrock gain need to be reduced?
                  Softrocks are normally sensitive enough so extra voltage could reduce the
                  dynamic range.

                  From: "Jeff Blaine AC0C"

                  >I had an instance where the ground loop was causing a gigantic rise near
                  >the zero and extending up maybe 10 kHz. Unbelievable mess. And >rather
                  >than actually track down the source of the ground loop (the shack wiring
                  >was a mess as well

                  Some might enjoy the challenge! However it may not be easy if it is all
                  boxed up before you start. It is most likely the proplems lie between what
                  is connected between ground, the computer and to the Softrock. That's the
                  path you have broken.


                  73 Alan G4ZFQ
                • Jack Smith
                  Dan: If you add a 100 to 10000 ohm matching transformer, the result is a 10:1 step up in voltage. The E MU 0202 10K input resistance will then reflect back to
                  Message 8 of 10 , Jul 2 8:18 AM
                  • 0 Attachment
                    Dan:

                    If you add a 100 to 10000 ohm matching transformer, the result is a 10:1
                    step up in voltage.

                    The E MU 0202 10K input resistance will then reflect back to the
                    transformer primary as 100 ohms, so you will see a 6 dB loss due to
                    voltage dividing based on the 100 ohm series R in your Softrock. This
                    nets out at 5:1 voltage step up, measuring from the op-amp output to the
                    E MU 0202 input.

                    My experience is that the E MU 0202 does not need extra voltage gain
                    with the Softrock Lite, and the net effect of 'matching' the impedances
                    will be to cause you to have to turn the E MU 0202's gain down to avoid
                    overloading on peaks.

                    You get to the same place by using a 1:1 transformer. In this case 10K
                    on the secondary reflects back as 10K on the primary, which causes
                    negligible drop across the 100 ohm series resistance and for most
                    practical purposes the full voltage output of the op-amp is delivered to
                    the E MU 0202's input.

                    My belief is that you don't need a 100 ohm to 10K transformer for the
                    reasons stated above.

                    Jack K8ZOA


                    dan edwards wrote:
                    >
                    >
                    > jack & jeff
                    >
                    > thanks for the feedback.
                    >
                    > jack, my rxtx6.3 has 100 ohms in series with the outputs of the op
                    > amps, along
                    > with .1 ufd. so, that means the output impedance is 100 ohms or so,
                    > right?
                    > a higher impedance ratio, in to out, would just give me a small
                    > voltage gain, as
                    > long as i dont create some funky resonances, right? but not really a
                    > significant
                    > advantage to try to match 100 to 10000 ohms?
                    >
                    > jeff: yes, i thought about switching the blocking cap to a much larger
                    > value, say
                    > 1 or 10 ufd. this is a good place for a tantalum cap, instead of a
                    > b-flat electrolytic?
                    > the 3 dollar radio shack unit is right in line with the budget for
                    > this project!!! ha.
                    > dirt cheap and actually works pretty well.
                    >
                    > 73, w5xz, dan
                    >
                    > --- On *Thu, 7/2/09, Jeff Blaine AC0C /<keepwalking188@...>/* wrote:
                    >
                    >
                    > From: Jeff Blaine AC0C <keepwalking188@...>
                    > Subject: Re: [softrock40] audio isolation transfomers
                    > To: softrock40@yahoogroups.com
                    > Date: Thursday, July 2, 2009, 1:09 PM
                    >
                    > Dan,
                    >
                    > Jack is the master. Lot of worse ways to waste time than camping
                    > out on his page for some review...
                    >
                    > I had an instance where the ground loop was causing a gigantic
                    > rise near the zero and extending up maybe 10 kHz. Unbelievable
                    > mess. And rather than actually track down the source of the
                    > ground loop (the shack wiring was a mess as well and that would
                    > have been a week long project), I stuck in the transformers based
                    > on Jack's review of the RS cheap transformer. Great results
                    > - very similar to what you have seen. A much improved noise
                    > profile - but quite deep nulls near the zero point.
                    >
                    > So I played around with the coupling cap value, and put a resistor
                    > on the sound-card side in an attempt to fill in the center a bit.
                    > Some improvement was had - but it's not picture perfect. The
                    > nulls are about 10db, and pretty narrow. It looks a bit funky on
                    > the screen, but in a practical sense, it's rare that the signal of
                    > interest is right in the null - I run a 7-15khz offset depending
                    > on sound card frequency - I try to use the highest offset
                    > possible, just to minimize the effect of the zero frequency which
                    > can be minimized, but never eliminated.
                    >
                    > I want to say the cap value went from 0.1->0.2 which helped fill
                    > in null a but but gave rise to a larger "bump" near the CF. The
                    > loading resistor was about 6K and that seemed to be about the best
                    > overall compromise value.
                    >
                    > Putting in the tx was a huge improvement. And the cap/resistor
                    > were probably mostly window dressing, in the overall scheme.
                    >
                    > Good luck and hope this helps a bit.
                    >
                    > 73/jeff/ac0c
                    >
                    >
                    >
                    > *From:* Jack Smith
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=Jack.Smith@...>
                    > *Sent:* Thursday, July 02, 2009 8:28 PM
                    > *To:* softrock40@yahoogro ups.com
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=softrock40@yahoogroups.com>
                    >
                    > *Subject:* Re: [softrock40] audio isolation transfomers
                    >
                    > Dan:
                    >
                    > I believe the 10K is correct.
                    >
                    > The output impedance of the Softrock is near zero. After all, it
                    > is an
                    > op amp in a feedback circuit so it tries to maintain the same output
                    > voltage regardless of the load on it. Hence its output impedance
                    > is in
                    > theory zero, limited only by the op-amps ability to source or sink
                    > current plus the impedance of the blocking capacitor.
                    >
                    > So you are not matching the Softrock to the sound card's input
                    > impedance
                    > in any normal sense of the term 'matching.' Rather the transformer
                    > ratios serve as a voltage step up more than anything else.
                    >
                    > You can think of the sound card as if it's a voltmeter - you don't
                    > want
                    > to match the volt meter's high input impedance to the voltage
                    > source you
                    > are measuring. If you don't need the voltage step up, a 1:1
                    > transformer
                    > ratio is fine. Could be 600:600 or 10K:10K. I doubt very much that
                    > you
                    > will find any benefit in a ratio other than 1:1 with an E MU 0202.
                    >
                    > Of course there are some practical issues related to stray
                    > capacitance
                    > in the transformer windings and how they alter the transformer' s
                    > high
                    > frequency response into various loads. You can see some of this in my
                    > various transformer pages. Hence, I would stick with the 600:600
                    > versions and deviate from that only if you know what you are doing.
                    >
                    > Jack
                    >
                    > dan edwards wrote:
                    > >
                    > >
                    > > thanks, jack
                    > >
                    > > i have looked at your site. great work. guess i overlooked this,
                    > though.
                    > >
                    > > so, quick easy answer is "No great advantage with the triad over
                    > the
                    > > r.s. cheapie" ??
                    > >
                    > > even with a passing attempt at 'better' impedance matching? i
                    > find a
                    > > variety
                    > > of z ratio triads are available. is my read of the tech specs of
                    > the
                    > > 0202 daughter
                    > > card input z of 10k ohms correct?
                    > >
                    > > 73, w5xz, dan
                    > >
                    > >
                    > > --- On *Thu, 7/2/09, Jack Smith /<Jack.Smith@cox. net
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=Jack.Smith%40cox.net>>/*
                    > wrote:
                    > >
                    > >
                    > > From: Jack Smith <Jack.Smith@cox. net
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=Jack.Smith%40cox.net>>
                    > > Subject: Re: [softrock40] audio isolation transfomers
                    > > To: softrock40@yahoogro ups.com
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=softrock40%40yahoogroups.com>
                    > > Date: Thursday, July 2, 2009, 10:12 AM
                    > >
                    > > Comparisons of the Radio Shack transformer and Triad SP-70 at
                    > > http://www.cliftonl <http://www.cliftonl/> aboratories.
                    > com/softrock_ lite_6_2. htm
                    > > <http://www.cliftonl aboratories. com/softrock_ lite_6_2. htm
                    > <http://www.cliftonlaboratories.com/softrock_lite_6_2.htm>>
                    > >
                    > > Jack K8ZOA
                    > >
                    > > w5xz@...
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=w5xz%40att.net>
                    > > <http://us.mc1802 mail.yahoo. com/mc/compose? to=w5xz%40att.
                    > net
                    > <http://us.mc1802.mail.yahoo.com/mc/compose?to=w5xz%40att.net>> wrote:
                    > > >
                    > > >
                    > > > i have a nasty ground loop with my emu1212; since it HAS
                    > > > balanced line inputs, i put a couple of cheapie radio
                    > > > shack isolation transformers (p/n 271-1374) in, and
                    > > > re-wired to the balanced inputs, with twisted pairs only.
                    > > > at $3 each, it was too cheap to pass up.
                    > > >
                    > > > now, there is a narrow but distinct notch around 0 Hz.
                    > > > about 20 db or so. beats the heck out of that noise hump,
                    > though...
                    > > >
                    > > > i suppose this is low freq roll-off? pleasantly surprised to
                    > > > see plenty of signal many many khz away from 0 hz, though.
                    > > > plus & minus 75 khz, plus.......sampling at 192k with the emu.
                    > > >
                    > > > overall, rx sensitivity seems better, also. with no particular
                    > > > attempt at impedance matching either; 1:1 transformers, supposedly
                    > > > i'm guessing the rxtx6.3 is about 50 ohms out, and i think
                    > > > the emu is 15k ohms in........
                    > > >
                    > > > do the good triad isolation transfomers do this also?
                    > > >
                    > > > 73, w5xz, dan
                    > > >
                    > > >
                    > >
                    > >
                    >
                    >
                  • Andy
                    ... This may be a really stupid and ignorant question on my part, but as long as the emu 1212 has balanced inputs, can t you use it with its balanced inputs
                    Message 9 of 10 , Jul 2 2:44 PM
                    • 0 Attachment
                      > i have a nasty ground loop with my emu1212; since it HAS
                      > balanced line inputs, i put a couple of cheapie radio
                      > shack isolation transformers (p/n 271-1374) in, and
                      > re-wired to the balanced inputs, with twisted pairs only.
                      > at $3 each, it was too cheap to pass up.

                      This may be a really stupid and ignorant question on my part, but as long as
                      the emu 1212 has balanced inputs, can't you use it with its balanced inputs
                      without the transformers? If they are good electronically balanced inputs
                      they mimic what a transformer does and break ground loops.

                      Usually you can go from unbalanced to balanced OK; it's the other direction
                      that may not work.

                      Andy
                    Your message has been successfully submitted and would be delivered to recipients shortly.