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

Re: [regenrx] Simple Audio Amp

Expand Messages
  • Brian Carling
    That is s a areally nice simple design... thanks for sharing it!
    Message 1 of 18 , Sep 1, 2006
    • 0 Attachment
      That is s a areally nice simple design... thanks for sharing it!

      > I saw that the venerable TL082 has a voltage swing up to 13V into 600
      > ohms, so I decided to try one as an audio amp. To get maximum output,
      > I set it up as a push-pull amp by using each side as inverting and
      > noninverting amps, respectively, then routing the resulting antiphase
      > outputs into an inexpensive Radio Shack 1000:8 ohm transformer.
      >
      > Just now having cobbled it up, I've attached it to my HI-FI infinite
      > impedance detector. I'm pleased with the results. This amp is just
      > about as loud as my 386 amp, and seems to have less distortion. I
      > think it sounds better. Downside is that it uses two batteries,
      > although that can probably be remedied with a virtual ground.
      >
      > Because this amplifier consists ONLY of the IC, four resistors, and
      > the transformer, it can be built in a few minutes. I'd like to see
      > others try it and share their thoughts.
      >
      > Some applications might need a 1uF capacitor at the input, but it made
      > no difference in my application.
      >
      > Here's a link to the schematic:
      >
      > http://www.beecavewoods.com/hobbies/TL082amp.jpg
      >
      > Macrohenry
      >
      >
      >
      >
      >
      >
      >
      >
      > "In the interest of Regenerating radio's lost art of Regeneration"
      > Yahoo! Groups Links
      >
      >
      >
      >
      >
      >
      >
    • macrohenry
      ... http://www.beecavewoods.com/hobbies/TL082amp.jpg You re welcome! Now I have a question: I ve just successfully created a virtual ground by connecting all
      Message 2 of 18 , Sep 1, 2006
      • 0 Attachment
        --- In regenrx@yahoogroups.com, "Brian Carling" <bcarling@...> wrote:
        >
        > That is s a areally nice simple design... thanks for sharing it!

        http://www.beecavewoods.com/hobbies/TL082amp.jpg

        You're welcome!

        Now I have a question: I've just successfully created a virtual ground
        by connecting all ground connections to the junction of two 2.2k
        resistors, whose other ends are tied to V+ and V- respectively. With a
        single 9V battery there is still plenty of volume and only 2 mA
        quiescent current.

        My question is, to even better preserve battery drain, how do I
        compute how high a value to give these two resistors? The output is
        not referenced to ground, but instead consists of the two opposite
        potentials of the opamp outputs. Therefore, it seems that the two
        virtual ground biasing resistors are relevant only to the input. Is
        that the case, or will it lose output power if the values of those
        resisitors is increased?

        Another question: The opamp's input impedance is 10^12 ohms. Since
        the input to the inverting side is through a 10K resistor, and the
        input to the noninverting side is a straight wire, isn't the input
        impedance for all practical purposes 10^12 ohms/2? So if I want an
        input impedance of say 47K ohms, then I would simply connect a 47K
        resistor from Vin to ground, right?

        Thanks!

        Mac
      • n1tev
        Hello, Interesting Project! Here are a few of my ramblings (sorry). It s lots of fun to design and build your own audio amps, either using descrete transistors
        Message 3 of 18 , Sep 6, 2006
        • 0 Attachment
          Hello,
          Interesting Project!

          Here are a few of my ramblings (sorry).

          It's lots of fun to design and build your own audio amps, either
          using descrete transistors or using ICs. However, for all this
          effort and all these parts I think using an LM380 (1 Watt) or LM384
          (5 Watt) IC is far easier. Plus these are internally biased so
          their outputs are at vs/2. Of course, you will learn more by making
          your own amplifier and maybe that's more important.

          Concerning your questions: Since their inputs are referenced at Vs/2
          the outputs of BOTH ICs will be centered at Vs/2 also. If you
          can "float" your speaker or headphones, you could simply DC couple
          out to them but this is dangerous as the outputs might contact
          ground on the receiver case etc and short out the IC. AC Coupling
          (or a transformer) is safer.

          Having said that, I will stick my neck out and state that in
          general, transformer coupling almost always seriously degrades audio
          quality compared with RC coupling. Of course, a transformer is an
          impedance matching device, so maximum power is delivered (more
          current, less voltage or visa-versa). Other coupling methods are
          less efficent. There are hi-fi transformers of course, but these
          are expensive. Most vacuum tube regens use an output transformer
          (usually hi-fi or fairly decent) but interstage transformers
          (between detector and amplifier) that I have tried all sounded
          pretty bad. This cancels out one of the really important benefits
          of a regen (or any TRF receiver): hi-fi LOW distortion audio. Food
          for thought.

          Assuming that input bias currents are fairly low, you can (and
          should) use 100k Resistors to bias your inputs at mid supply. NOTE
          HOWEVER, that there is virtually NO power supply rejection now: any
          noise or AC signals etc can come in from the supply lines into the
          ICs and cause Hum or "motorboating". Therefore it is essential that
          your virtual ground be bypassed to AC common using a LARGE
          capacitor: 100 microFarads is a decent value to try (1/2 pie RC is
          pole frequency so choose maybe 3X or 10X lower frequency than your
          circuit BW when deciding on cap value).

          As for input impedance, the higher this is the lower the value of AC
          coupling caps required for a particular BW so why not just leave as
          is and use the smallest AC input coupling caps you can?

          Happy homebrewing,
          N1TEV




          --- In regenrx@yahoogroups.com, "macrohenry" <macrohenry@...> wrote:
          >
          > --- In regenrx@yahoogroups.com, "Brian Carling" <bcarling@> wrote:
          > >
          > > That is s a areally nice simple design... thanks for sharing it!
          >
          > http://www.beecavewoods.com/hobbies/TL082amp.jpg
          >
          > You're welcome!
          >
          > Now I have a question: I've just successfully created a virtual
          ground
          > by connecting all ground connections to the junction of two 2.2k
          > resistors, whose other ends are tied to V+ and V- respectively.
          With a
          > single 9V battery there is still plenty of volume and only 2 mA
          > quiescent current.
          >
          > My question is, to even better preserve battery drain, how do I
          > compute how high a value to give these two resistors? The output
          is
          > not referenced to ground, but instead consists of the two opposite
          > potentials of the opamp outputs. Therefore, it seems that the two
          > virtual ground biasing resistors are relevant only to the input. Is
          > that the case, or will it lose output power if the values of those
          > resisitors is increased?
          >
          > Another question: The opamp's input impedance is 10^12 ohms. Since
          > the input to the inverting side is through a 10K resistor, and the
          > input to the noninverting side is a straight wire, isn't the input
          > impedance for all practical purposes 10^12 ohms/2? So if I want an
          > input impedance of say 47K ohms, then I would simply connect a 47K
          > resistor from Vin to ground, right?
          >
          > Thanks!
          >
          > Mac
          >
        • Brian Carling
          Set an LM380 amplifier aide by side with a 6V6 amplifier and I guarantee I can tell you which one is which. There are scientific reasons why a tube amplifier
          Message 4 of 18 , Sep 6, 2006
          • 0 Attachment
            Set an LM380 amplifier aide by side with a 6V6
            amplifier and I guarantee I can tell you which one
            is which.

            There are scientific reasons why a tube amplifier
            has superior audio. Ask any musician.

            > Hello,
            > Interesting Project!
            >
            > Here are a few of my ramblings (sorry).
            >
            > It's lots of fun to design and build your own audio amps, either
            > using descrete transistors or using ICs. However, for all this
            > effort and all these parts I think using an LM380 (1 Watt) or LM384
            > (5 Watt) IC is far easier. Plus these are internally biased so
            > their outputs are at vs/2. Of course, you will learn more by making
            > your own amplifier and maybe that's more important.
            >
            > Concerning your questions: Since their inputs are referenced at Vs/2
            > the outputs of BOTH ICs will be centered at Vs/2 also. If you
            > can "float" your speaker or headphones, you could simply DC couple
            > out to them but this is dangerous as the outputs might contact
            > ground on the receiver case etc and short out the IC. AC Coupling
            > (or a transformer) is safer.
            >
            > Having said that, I will stick my neck out and state that in
            > general, transformer coupling almost always seriously degrades audio
            > quality compared with RC coupling. Of course, a transformer is an
            > impedance matching device, so maximum power is delivered (more
            > current, less voltage or visa-versa). Other coupling methods are
            > less efficent. There are hi-fi transformers of course, but these
            > are expensive. Most vacuum tube regens use an output transformer
            > (usually hi-fi or fairly decent) but interstage transformers
            > (between detector and amplifier) that I have tried all sounded
            > pretty bad. This cancels out one of the really important benefits
            > of a regen (or any TRF receiver): hi-fi LOW distortion audio. Food
            > for thought.
            >
            > Assuming that input bias currents are fairly low, you can (and
            > should) use 100k Resistors to bias your inputs at mid supply. NOTE
            > HOWEVER, that there is virtually NO power supply rejection now: any
            > noise or AC signals etc can come in from the supply lines into the
            > ICs and cause Hum or "motorboating". Therefore it is essential that
            > your virtual ground be bypassed to AC common using a LARGE
            > capacitor: 100 microFarads is a decent value to try (1/2 pie RC is
            > pole frequency so choose maybe 3X or 10X lower frequency than your
            > circuit BW when deciding on cap value).
            >
            > As for input impedance, the higher this is the lower the value of AC
            > coupling caps required for a particular BW so why not just leave as
            > is and use the smallest AC input coupling caps you can?
            >
            > Happy homebrewing,
            > N1TEV
            >
            >
            >
            >
            > --- In regenrx@yahoogroups.com, "macrohenry" <macrohenry@...> wrote:
            > >
            > > --- In regenrx@yahoogroups.com, "Brian Carling" <bcarling@> wrote:
            > > >
            > > > That is s a areally nice simple design... thanks for sharing it!
            > >
            > > http://www.beecavewoods.com/hobbies/TL082amp.jpg
            > >
            > > You're welcome!
            > >
            > > Now I have a question: I've just successfully created a virtual
            > ground
            > > by connecting all ground connections to the junction of two 2.2k
            > > resistors, whose other ends are tied to V+ and V- respectively.
            > With a
            > > single 9V battery there is still plenty of volume and only 2 mA
            > > quiescent current.
            > >
            > > My question is, to even better preserve battery drain, how do I
            > > compute how high a value to give these two resistors? The output
            > is
            > > not referenced to ground, but instead consists of the two opposite
            > > potentials of the opamp outputs. Therefore, it seems that the two
            > > virtual ground biasing resistors are relevant only to the input. Is
            > > that the case, or will it lose output power if the values of those
            > > resisitors is increased?
            > >
            > > Another question: The opamp's input impedance is 10^12 ohms. Since
            > > the input to the inverting side is through a 10K resistor, and the
            > > input to the noninverting side is a straight wire, isn't the input
            > > impedance for all practical purposes 10^12 ohms/2? So if I want an
            > > input impedance of say 47K ohms, then I would simply connect a 47K
            > > resistor from Vin to ground, right?
            > >
            > > Thanks!
            > >
            > > Mac
            > >
            >
            >
            >
            >
            >
            >
            > "In the interest of Regenerating radio's lost art of Regeneration"
            > Yahoo! Groups Links
            >
            >
            >
            >
            >
            >
          • James Duffey
            Brian - In reply to your comments: There are scientific reasons why a tube amplifier has superior audio. Ask any musician. Well, I wouldn t ask a musician if
            Message 5 of 18 , Sep 7, 2006
            • 0 Attachment
              Brian - In reply to your comments:

              "There are scientific reasons why a tube amplifier
              has superior audio. Ask any musician."

              Well, I wouldn't ask a musician if I wanted a scientific reason. While
              there may be some scientists out there that are good musicians, most
              musicians are not good scientists. :^)= Most of the musicians I know
              don't really care that much about their sound systems. I think that you
              would draw blank stares from most musicians if you ask them "What are
              the scientific reasons a tube amplifier has superior audio? Some rock
              musicians may prefer tube guitar amps to solid state ones, primarily to
              duplicate the distortion present in classic rock guitar songs, like
              Jimi Hendrix.

              WHile you can certainly tell the difference between a 6V6 based
              amplifier and a LM380, most, if not all of this difference is due to
              different design/bias topology and the power levels being run. Most
              people would be hard pressed to tell the difference between a well
              designed 6V6 based amp and a well designed MOSFET based amp running the
              same power level and with the same topology, say Class A single ended.
              Class A amplifiers, driving a high efficiency load sound good whether
              made out of hollow state or solid state.

              Most of the preferences that people have for tube amplifiers have been
              shown to be preferences for 2nd order distortion over 3rd order
              distortion, slew rate deficiencies in early bipolar amplifiers, and a
              bit of hum.

              There is nothing inherently better sounding in tubes than in solid
              state devices. They can be designed to sound the same. The quality of
              an amp's sound can be quantified in terms of frequency response,
              distortion, slew rate, and transfer function. Bill Carver has
              demonstrated solid state amplifiers that can be configured to sound
              like a tube amplifier and in double blind tests he has been proven
              correct, even with subjects that maintain they can tell the dfference.

              Personal preferences for tubes over transistors are fine, but to
              ascribe them to scientific reasons is incorrect.

              What this has to do with regenerative receivers is beyond me, other
              than my solid state Grundig Yachtboy sounds lots better than the old
              tube Knight Space Spanner I built 40 some odd years ago. :^)= - Dr.
              Megacycle KK6MC/5

              --
              James Duffey KK6MC/5
              Cedar Crest NM 87008
              DM65
            • macrohenry
              ... Hi, Charles, thanks for the detailed help. Yes, learning and the thrill of discovery are important to me. ... The transformer coupling is intriguing
              Message 6 of 18 , Sep 7, 2006
              • 0 Attachment
                --- In regenrx@yahoogroups.com, "n1tev" <n1tev@...> wrote:

                > Of course, you will learn more by making
                > your own amplifier and maybe that's more important.

                Hi, Charles, thanks for the detailed help. Yes, learning and the
                thrill of discovery are important to me.

                > Having said that, I will stick my neck out and state that in
                > general, transformer coupling almost always seriously degrades audio
                > quality compared with RC coupling.

                The transformer coupling is intriguing because of a stereo tube amp I
                built that sounds more accurate and full of life than any of my higher
                end solid state equipment. I'll let a skeleton out of the closet
                here: I never replaced my prototype's test setup output transformers
                with the proper Hammond 1608s I ordered. So the amp still uses a pair
                of split bobbin power transformers for output transformers and just
                sounds wonderful. That experience got me on the quest for other
                inexpensive and unusual choices that sound good.

                > Assuming that input bias currents are fairly low, you can (and
                > should) use 100k Resistors to bias your inputs at mid supply. NOTE
                > HOWEVER, that there is virtually NO power supply rejection now:

                Do you mean in addition to my current virtual ground using 2.2K
                resistors? I think they only affect the input, as the outputs are
                floating at the transformer. What difference would it make to the
                performance if I used 100K resistors instead of the 2.2K? Im a bit
                confused here, since mid supply is virtual ground and I'd be just
                grounding the inputs.

                > it is essential that
                > your virtual ground be bypassed to AC common using a LARGE
                > capacitor: 100 microFarads is a decent value to try (1/2 pie RC is
                > pole frequency so choose maybe 3X or 10X lower frequency than your
                > circuit BW when deciding on cap value).

                Will do!
                >
                > As for input impedance, the higher this is the lower the value of AC
                > coupling caps required for a particular BW so why not just leave as
                > is and use the smallest AC input coupling caps you can?

                You mean with the 2.2K virtual ground resistors, right?

                As for input impedance, is it still 10^12 ohms (TL082's specification)
                since the inputs are respectively through a straight wire and through
                a 10K resistor?

                When you mention AC coupling caps, can you comment on any necessity
                for putting one in each input, or will one do just ahead of where the
                inputs are tied together?

                For your convenience, here's the circuit again (without virtual ground:
                http://www.beecavewoods.com/hobbies/TL082amp.jpg

                Thanks for your help!

                Mac
              • Hue Miller
                I popped the back off an extra pair of computer speakers here. I saw a simple circuit on about a 1 inch square board and a small power transformer mounted
                Message 7 of 18 , Sep 7, 2006
                • 0 Attachment
                  I popped the back off an extra pair of computer speakers here.
                  I saw a simple circuit on about a 1 inch square board and a
                  small power transformer mounted separately. The online
                  data sheets say 2 - 15 volts operating voltage and even at
                  2 volts supply, in "bridged" (mono, not stereo or dual channel)
                  the output is still 80 mW. Since despite what the computer
                  sellers say, computer speakers are in no way "Hi Fi", due to
                  tiny speakers and plastics cabinets, they aren't highly valued
                  and i always see a selection at the Goodwill (type) stores. This
                  might be a reasonable source for the audio section of your project.
                  I don't know what other chips you might run in to, but probably
                  the idea is similar, and the price is right. You could probably just
                  harvest the whole thing, transplant it to your project, use only
                  one channel if you're unsure how to bridge it, and just feed your
                  receiver audio to the "top" of the audio pot (volume control).
                  Possibly just directly bidging the 2 channels would work, i don't
                  know, for mono. -Hue Miller
                • Brian Carling
                  ... The two comments were not necessarily connected. I didn t say the ask a musician FOR a scientific reason. ... Again I never said to do that. ... Very
                  Message 8 of 18 , Sep 8, 2006
                  • 0 Attachment
                    > Brian - In reply to your comments:
                    >
                    > "There are scientific reasons why a tube amplifier
                    > has superior audio. Ask any musician."
                    >
                    > Well, I wouldn't ask a musician if I wanted a scientific reason.

                    The two comments were not necessarily connected.
                    I didn't say the ask a musician FOR a scientific reason.

                    >While
                    > there may be some scientists out there that are good musicians, most
                    > musicians are not good scientists. :^)= Most of the musicians I know
                    > don't really care that much about their sound systems. I think that you
                    > would draw blank stares from most musicians if you ask them "What are
                    > the scientific reasons a tube amplifier has superior audio?

                    Again I never said to do that.

                    >Some rock
                    > musicians may prefer tube guitar amps to solid state ones, primarily to
                    > duplicate the distortion present in classic rock guitar songs, like
                    > Jimi Hendrix.

                    Very funny, but what the heck does this have to do with
                    regenerative receivers? I must accepot some responsibility
                    for the rabbnit trail too, and we should end it here, lest we incurr the
                    wrath of the moderators.

                    Your Jii Hendrix remark shows that you know very little
                    about the difference between tube and solid state audio
                    amplifiers, so I pretty much see where you are coming from.

                    > WHile you can certainly tell the difference between a 6V6 based
                    > amplifier and a LM380, most, if not all of this difference is due to
                    > different design/bias topology and the power levels being run.

                    Not a scientific answer, although two points for using fancy
                    words like "topology." I know that a lot of military contractors
                    who charge our government $50 extra for using that word, LOL.

                    > Most
                    > people would be hard pressed to tell the difference between a well
                    > designed 6V6 based amp and a well designed MOSFET based amp running the
                    > same power level and with the same topology, say Class A single ended.
                    > Class A amplifiers, driving a high efficiency load sound good whether
                    > made out of hollow state or solid state.

                    No, you are mistaken James. It has to do with coloration
                    and harmonics. I have worked in the industry for over 30 years.
                    Others will back me up on this. You CAN hear the difference.
                    Look at the transfer curves for a tube and the characteristics
                    of a transistor some time. It's only subtle when you look
                    at the graphs, but the sound is noticeably different.

                    > Most of the preferences that people have for tube amplifiers have been
                    > shown to be preferences for 2nd order distortion over 3rd order
                    > distortion, slew rate deficiencies in early bipolar amplifiers, and a
                    > bit of hum.

                    Oh yes, we prefert the hum. Puhleez!

                    > There is nothing inherently better sounding in tubes than in solid
                    > state devices.

                    In your opinion perhaps.

                    >They can be designed to sound the same.

                    For what cost though? And please explain why 40 years
                    after transistors became commonplace, and 15 years after
                    DSP enabled the triclks you describe, we still have the
                    vast number of guitar players prefering tube amplifiers
                    and manufacturers frequently coming out with multiple
                    new models of tube amplifiers, especially in the past
                    ten years! Add to that the VAST hordes of tube-only
                    audiophiles. The result is an empirical one, but it flies
                    in the face of your theory.

                    > an amp's sound can be quantified in terms of frequency response,
                    > distortion, slew rate, and transfer function.

                    Yes, but that leaves out the human ear.

                    >Bill Carver has
                    > demonstrated solid state amplifiers that can be configured to sound
                    > like a tube amplifier and in double blind tests he has been proven
                    > correct,

                    OK fine, but can he sell them for $300 each to me?
                    My wallet can tell the difference.

                    > Personal preferences for tubes over transistors are fine, but to
                    > ascribe them to scientific reasons is incorrect.

                    Whatever.

                    > What this has to do with regenerative receivers is beyond me,

                    So why must you insist on commenting at such great length on it
                    then?

                    Dr. Megacycle KK6MC/5

                    Ah, as I suspected, another phd.
                  • Brian Carling
                    Let s cut through the smoke and mirrors here a little, James old chap. My comment was that in a regenerative receiver a 6V6 audio amplifier is going to sound
                    Message 9 of 18 , Sep 8, 2006
                    • 0 Attachment
                      Let's cut through the smoke and mirrors here a little, James
                      old chap.

                      My comment was that in a regenerative receiver
                      a 6V6 audio amplifier is going to sound better than
                      an LM380 audio output stage. Let's try to stick to the
                      subject here instead of wandering down the rabbit trail
                      to Mr. Carvin's unrelated experiment. We are not talking
                      about adding any fancy electronics or digital processing.
                      We are not running the LM380 at 10 mW of Class A.
                      We are comparing head to head with ordinary simple
                      components. The 6V6 will be louder and sound more
                      pleasant. Write multiple paragraphs to contradict it all
                      you want, but if I were building a regen set that is
                      what I would use. Now if YOU prefer to use squalid
                      state rubbish, then that is your prerogative.
                      If your ears are unable to tell the difference, then fine.
                      Enjoy your regen, but please don't try to take these
                      extreme pseudotechnical trails to criticize mine.

                      > Brian - In reply to your comments:

                      > Well, I wouldn't ask a musician if I wanted a scientific reason. While
                      > there may be some scientists out there that are good musicians, most
                      > musicians are not good scientists. :^)= Most of the musicians I know
                      > don't really care that much about their sound systems. I think that you
                      > would draw blank stares from most musicians if you ask them "What are
                      > the scientific reasons a tube amplifier has superior audio? Some rock
                      > musicians may prefer tube guitar amps to solid state ones, primarily to
                      > duplicate the distortion present in classic rock guitar songs, like
                      > Jimi Hendrix.
                    • n1tev
                      Hi Mac, Well...I simply read your post without realizing there was a circuit diagram. Now I understand what you are trying to do here. If I were you, I would
                      Message 10 of 18 , Sep 8, 2006
                      • 0 Attachment
                        Hi Mac,
                        Well...I simply read your post without realizing there was a circuit
                        diagram. Now I understand what you are trying to do here.

                        If I were you, I would redraw this circuit to show the op-amps
                        inside the chip...it's hard to understand unless you do this.

                        And...I would continue to use two 9V batteries as this gives you 18V
                        (+9V,-9V swing) rather than a only a +4.5, -4.5V swing. It will
                        probably sound better using dual supplies too.

                        But let's go ahead with thinking about single supply conversion
                        anyway.

                        Normally, a virtual ground is created (to allow single supply
                        operation) by biasing the non inverting op-amp input at Vs/2 using
                        two large value (such as 100k) resistors. If this is suitably
                        bypassed to ground using a BIG cap then this will work well. The op-
                        amp's output is centered at Vs/2, as is it's two inputs. AC
                        coupling needs to be used for the inverting input and for the output.

                        I wrote an application note on this subject. You can download it
                        FREE at:
                        http://www.analog.com/UploadedFiles/Application_Notes/42621535592205A
                        N581.pdf

                        Now you are more or less doing the same thing here with your 2k Ohm
                        resistors: you are setting "ground" at Vs/2 (mid supply). But there
                        are problems. First you MUST AC COUPLE your inputs. You need this
                        because you have set the two inputs at Vs/2 (grounding pin 3 and
                        grounding the other 10k from pin 6 does this). So the inputs are at
                        Vs/2 and so are the outputs. You could do the same thing by
                        connecting two 100k resistors to your Vin point. So, you MUST put a
                        cap in series between the pin 5/R1 tie point and the input source
                        and another one between the other 10k (the 10k tied to pin 6) and
                        ground.

                        If you were not using a transformer and you were trying to draw
                        current through your phantom ground you would be in trouble, as the
                        current would cause uneven voltage drops across your two resistors
                        and you would need really small resistors and large bypass caps.
                        This would "eat-up" a lot of supply current.

                        But...you are using a transformer, so now current flow is between
                        the two amplifiers and the battery but not through "ground".

                        The input impedance of your circuit is 10k because R1 is 10k and the
                        op-amp's "virtual ground" forces pin two to be the same as pin 3
                        which is "ground".

                        Find the input coupling capacitor values using F= 1/2 pie RC (also
                        c= 1/2 pie rf). For F of about 10Hz and a 10k input Z (R1), the cap
                        should be about 1.59 microFarad. So use a 2.2 microFarad FOR BOTH
                        CAPACITORS. The + should go towards R1 since it is at Vs/2 and the
                        + towards the other 10k that is grounded. Does this make sense?

                        Other problems: There is a gain mismatch because one amplifier
                        operates at g=100 and the other is at g=101 (gain of A1 is R1/R2
                        while gain of A2 is 1 + R1/R2) but your resistor mismatch will be
                        more than this anyway. With 5% resistors, you could have an 11%
                        (worse case) gain error. I would use 1% resistors here to avoid
                        problems.

                        A 1meg Ohm value for R2 seems really high and some amplifiers with
                        high input bias currents could have very large voltage offsets but I
                        guess you need to keep input impedance at 10K..and you are using a
                        FET amplifier, so you might get by OK. But if you can go lower (if
                        you can drive) say 1K then it would be better to use 100k R2 and 1k
                        R1. If you decide to use 1k's and 100k's then increase input
                        coupling caps to 22 MicroFarads.

                        Now...since no current is flowing through your virtual ground
                        resistors, make them BIG so that you keep supply current as low as
                        possible. Using two 2K resistors you consume 2mA just making your
                        ground (9V/4k). So, use two 100k resistors.

                        I would also add two BIG bybass capacitors: maybe 50 MicroFarads or
                        more, one between each side of the battery and your phantom ground.
                        Be sure to observe polarity. This will assure that your phantom
                        ground is a true AC ground.

                        One final thought: the output transformer is "open loop"..no
                        feedback between transformer secondary and the amplifier input.
                        This can be tricky to do but it's worth trying. Adding feedback
                        will greatly reduce transformer distortion, which might be really
                        high. After everything else is working OK, try grounding one side
                        of the transformer secondary and then connecting a series resistor
                        between the other side of the secondary and the input to the
                        amplifier (BEFORE the input coupling cap). You have a 50/50 chance
                        of creating an oscillator. So expect feedback. Ground which every
                        side of the transformer secondary that gives you stable operation.
                        I'm guessing that maybe a 100k resistor would be good here but
                        experient and see.

                        Well...that's more than enough from me.

                        Happy homebrewing!






                        --- In regenrx@yahoogroups.com, "macrohenry" <macrohenry@...> wrote:
                        >
                        > --- In regenrx@yahoogroups.com, "n1tev" <n1tev@> wrote:
                        >
                        > > Of course, you will learn more by making
                        > > your own amplifier and maybe that's more important.
                        >
                        > Hi, Charles, thanks for the detailed help. Yes, learning and the
                        > thrill of discovery are important to me.
                        >
                        > > Having said that, I will stick my neck out and state that in
                        > > general, transformer coupling almost always seriously degrades
                        audio
                        > > quality compared with RC coupling.
                        >
                        > The transformer coupling is intriguing because of a stereo tube
                        amp I
                        > built that sounds more accurate and full of life than any of my
                        higher
                        > end solid state equipment. I'll let a skeleton out of the closet
                        > here: I never replaced my prototype's test setup output
                        transformers
                        > with the proper Hammond 1608s I ordered. So the amp still uses a
                        pair
                        > of split bobbin power transformers for output transformers and just
                        > sounds wonderful. That experience got me on the quest for other
                        > inexpensive and unusual choices that sound good.
                        >
                        > > Assuming that input bias currents are fairly low, you can (and
                        > > should) use 100k Resistors to bias your inputs at mid supply.
                        NOTE
                        > > HOWEVER, that there is virtually NO power supply rejection now:
                        >
                        > Do you mean in addition to my current virtual ground using 2.2K
                        > resistors? I think they only affect the input, as the outputs are
                        > floating at the transformer. What difference would it make to the
                        > performance if I used 100K resistors instead of the 2.2K? Im a bit
                        > confused here, since mid supply is virtual ground and I'd be just
                        > grounding the inputs.
                        >
                        > > it is essential that
                        > > your virtual ground be bypassed to AC common using a LARGE
                        > > capacitor: 100 microFarads is a decent value to try (1/2 pie RC
                        is
                        > > pole frequency so choose maybe 3X or 10X lower frequency than
                        your
                        > > circuit BW when deciding on cap value).
                        >
                        > Will do!
                        > >
                        > > As for input impedance, the higher this is the lower the value
                        of AC
                        > > coupling caps required for a particular BW so why not just leave
                        as
                        > > is and use the smallest AC input coupling caps you can?
                        >
                        > You mean with the 2.2K virtual ground resistors, right?
                        >
                        > As for input impedance, is it still 10^12 ohms (TL082's
                        specification)
                        > since the inputs are respectively through a straight wire and
                        through
                        > a 10K resistor?
                        >
                        > When you mention AC coupling caps, can you comment on any necessity
                        > for putting one in each input, or will one do just ahead of where
                        the
                        > inputs are tied together?
                        >
                        > For your convenience, here's the circuit again (without virtual
                        ground:
                        > http://www.beecavewoods.com/hobbies/TL082amp.jpg
                        >
                        > Thanks for your help!
                        >
                        > Mac
                        >
                      • Ned Keating
                        I do wish Brian would temper and shorten his commentaries. He s fast becoming a condescending bore. Maybe he should stick to selling batteries. I m sure many
                        Message 11 of 18 , Sep 8, 2006
                        • 0 Attachment
                          I do wish Brian would temper and shorten his commentaries. He's fast
                          becoming a condescending bore.
                          Maybe he should stick to selling batteries. I'm sure many more agree.
                          Ned Keating
                        • Brian Carling
                          Ned - I tried to end this thread, but YOU had to whine to the whole world. YOu wanted a reaction, you got it: Ned - I don t believe your opinions are very
                          Message 12 of 18 , Sep 8, 2006
                          • 0 Attachment
                            Ned - I tried to end this thread, but YOU had to
                            whine to the whole world. YOu wanted a reaction, you got it:

                            Ned - I don't believe your opinions are very important
                            to this group, and you just created a lot of
                            pathetic, macho, off-topic noise here.
                            I think you should be on the TV evening news.
                            At least there you could censor everyone else,
                            but viewers would have the option of switching you off.
                            He he - in fact I just did that. You are being kill-filtered.
                            Future e-mails from you will not be read or reacted
                            to here. I apologize to te group for this response.
                            Nobody in their right mind wants to read this
                            pre-K level of stupidity.

                            Now, hopefully you can stop with the boring, immature
                            personal attacks and we will get back to discussing
                            things related to regenerative receivers, like audio
                            stages.

                            Go get some breath mints, then crawl back under your
                            rock with the other trolls.

                            Be well.

                            > I do wish Brian would temper and shorten his commentaries. He's fast
                            > becoming a condescending bore.
                            > Maybe he should stick to selling batteries. I'm sure many more agree.
                            > Ned Keating
                            >
                            >
                            > "In the interest of Regenerating radio's lost art of Regeneration"
                            > Yahoo! Groups Links
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                            >
                          • Ned Ely
                            Brian, Besides, he is giving Neds a bad rap! hehehe... I have lost the address of your site where we can put some of our creations. Thanks for making the
                            Message 13 of 18 , Sep 8, 2006
                            • 0 Attachment
                              Brian,

                              Besides, he is giving "Neds" a bad rap! hehehe...

                              I have lost the address of your site where we can put some of our creations.
                              Thanks for making the space available.

                              Ned - kd5yqh

                              > Ned - I tried to end this thread, but YOU had to
                              > whine to the whole world. YOu wanted a reaction, you got it:
                              >
                              > Ned - I don't believe your opinions are very important
                              > to this group, and you just created a lot of
                              > pathetic, macho, off-topic noise here.

                              <SNIP>
                            • macrohenry
                              ... Excellent, thanks, I got it and will read it. Perhaps the forum member who emailed me with this question will read it, too! Everything you say makes
                              Message 14 of 18 , Sep 8, 2006
                              • 0 Attachment
                                --- In regenrx@yahoogroups.com, "n1tev" <n1tev@...> wrote:

                                > I wrote an application note on this subject.

                                Excellent, thanks, I got it and will read it. Perhaps the forum
                                member who emailed me with this question will read it, too!

                                Everything you say makes sense, and this is exactly the knowledge I'm
                                seeking. Thank you. I have a couple of questions in response to your
                                comments below:
                                >

                                > Other problems: There is a gain mismatch because one amplifier
                                > operates at g=100 and the other is at g=101 (gain of A1 is R1/R2
                                > while gain of A2 is 1 + R1/R2) but your resistor mismatch will be
                                > more than this anyway. With 5% resistors, you could have an 11%
                                > (worse case) gain error. I would use 1% resistors here to avoid
                                > problems.

                                I was aware of the gain mismatch, and figured it would be swamped out
                                by the component's tolerance. But you make a good point. Since the
                                transformer's primary isn't referenced to ground, what would be the
                                sonic effect of a gain mismatch?

                                > A 1meg Ohm value for R2 seems really high and some amplifiers with
                                > high input bias currents could have very large voltage offsets but I
                                > guess you need to keep input impedance at 10K..and you are using a
                                > FET amplifier, so you might get by OK. But if you can go lower (if
                                > you can drive) say 1K then it would be better to use 100k R2 and 1k
                                > R1.

                                Yes, this was a concern to me; currently to get the volume I want, I'm
                                using 10k and 2 Meg. I've added a 10K volume control ahead of the
                                input and can crank it up 3/4 before it starts overdriving.

                                I don't know the output specs for the infinite impedance detector I'm
                                using to drive the amp. Here are links to the ckt and the article for
                                those who are interested. It really brings out the best in AM fidelity.

                                http://sound.westhost.com/articles/am-f6.gif
                                http://sound.westhost.com/articles/am-radio.htm

                                The point here is that I may be able to use 1K as input without
                                loading the detector too much. I do have a concern about loading this
                                detector. With 10K it can cleanly resolve 10 kHz adjacent medium
                                strength stations. I'll just tack on some 2.2K and 220K to see what
                                happens here.
                                >
                                >
                                > One final thought: the output transformer is "open loop"..no
                                > feedback between transformer secondary and the amplifier input.
                                > This can be tricky to do but it's worth trying. Adding feedback
                                > will greatly reduce transformer distortion, which might be really
                                > high. After everything else is working OK, try grounding one side
                                > of the transformer secondary and then connecting a series resistor
                                > between the other side of the secondary and the input to the
                                > amplifier (BEFORE the input coupling cap). You have a 50/50 chance
                                > of creating an oscillator. So expect feedback. Ground which every
                                > side of the transformer secondary that gives you stable operation.
                                > I'm guessing that maybe a 100k resistor would be good here but
                                > experient and see.

                                I've actually been working on this, as it was pointed out to me from
                                an engineer on another forum. He commented that the open loop
                                characteristic of the IC could cause such feedback not to be worth
                                while. I learned that in the ckt as drawn, I could connect the
                                feedback resistor and ground to EITHER end of the transformer output
                                and still get only positive feedback.

                                So now I'm using a 70 volt line transformer with C, 4, 8, 16 ohm
                                outputs with the ground at the 8. Now I can reverse the C and 16 taps
                                and get positive and negative feedback, respectively. I sure hope the
                                8 ohm tap is truly in the middle. I've not measured it. Feedback
                                resistor is a 500K pot, and yes I can get it to oscillate easily.

                                I'll try your suggestions and send a report. Thanks again.

                                Mac



                                >
                                > --- In regenrx@yahoogroups.com, "macrohenry" <macrohenry@> wrote:
                                > >
                                > > --- In regenrx@yahoogroups.com, "n1tev" <n1tev@> wrote:
                                > >
                                > > > Of course, you will learn more by making
                                > > > your own amplifier and maybe that's more important.
                                > >
                                > > Hi, Charles, thanks for the detailed help. Yes, learning and the
                                > > thrill of discovery are important to me.
                                > >
                                > > > Having said that, I will stick my neck out and state that in
                                > > > general, transformer coupling almost always seriously degrades
                                > audio
                                > > > quality compared with RC coupling.
                                > >
                                > > The transformer coupling is intriguing because of a stereo tube
                                > amp I
                                > > built that sounds more accurate and full of life than any of my
                                > higher
                                > > end solid state equipment. I'll let a skeleton out of the closet
                                > > here: I never replaced my prototype's test setup output
                                > transformers
                                > > with the proper Hammond 1608s I ordered. So the amp still uses a
                                > pair
                                > > of split bobbin power transformers for output transformers and just
                                > > sounds wonderful. That experience got me on the quest for other
                                > > inexpensive and unusual choices that sound good.
                                > >
                                > > > Assuming that input bias currents are fairly low, you can (and
                                > > > should) use 100k Resistors to bias your inputs at mid supply.
                                > NOTE
                                > > > HOWEVER, that there is virtually NO power supply rejection now:
                                > >
                                > > Do you mean in addition to my current virtual ground using 2.2K
                                > > resistors? I think they only affect the input, as the outputs are
                                > > floating at the transformer. What difference would it make to the
                                > > performance if I used 100K resistors instead of the 2.2K? Im a bit
                                > > confused here, since mid supply is virtual ground and I'd be just
                                > > grounding the inputs.
                                > >
                                > > > it is essential that
                                > > > your virtual ground be bypassed to AC common using a LARGE
                                > > > capacitor: 100 microFarads is a decent value to try (1/2 pie RC
                                > is
                                > > > pole frequency so choose maybe 3X or 10X lower frequency than
                                > your
                                > > > circuit BW when deciding on cap value).
                                > >
                                > > Will do!
                                > > >
                                > > > As for input impedance, the higher this is the lower the value
                                > of AC
                                > > > coupling caps required for a particular BW so why not just leave
                                > as
                                > > > is and use the smallest AC input coupling caps you can?
                                > >
                                > > You mean with the 2.2K virtual ground resistors, right?
                                > >
                                > > As for input impedance, is it still 10^12 ohms (TL082's
                                > specification)
                                > > since the inputs are respectively through a straight wire and
                                > through
                                > > a 10K resistor?
                                > >
                                > > When you mention AC coupling caps, can you comment on any necessity
                                > > for putting one in each input, or will one do just ahead of where
                                > the
                                > > inputs are tied together?
                                > >
                                > > For your convenience, here's the circuit again (without virtual
                                > ground:
                                > > http://www.beecavewoods.com/hobbies/TL082amp.jpg
                                > >
                                > > Thanks for your help!
                                > >
                                > > Mac
                                > >
                                >
                              • James Duffey
                                Brian - I apologize to you and the list for not making my point on tube and solid state amplifiers clearer and by cluttering it up with things that were
                                Message 15 of 18 , Sep 10, 2006
                                • 0 Attachment
                                  Brian - I apologize to you and the list for not making my point on tube
                                  and solid state amplifiers clearer and by cluttering it up with things
                                  that were somewhat unrelated to the original point. The point that I
                                  was trying to make is that the configuration of the amplifier, how it
                                  is biased, the type of feedback used, the load it is driving, the
                                  source driving it, the noise floor, and how much headroom is available
                                  at the output and input is much more important in designing amplifiers
                                  that accurately reproduce the input signal at the output than is the
                                  type of active device used. N1TEV made the point that I wanted to, only
                                  better. Thanks Charles.

                                  You requested scientific answers. This prompted me to dig through my
                                  old files. There is an excellent paper on the subject we were
                                  discussing:

                                  W. Bussey and R. Haigler (1981); "Tubes versus transistors in electric
                                  guitar amplifiers"; IEEE International Conference on Acoustics, Speech,
                                  and Signal Processing, V. 6 pg. 800

                                  I believe this is available on the web at:

                                  <http://milbert.com/articles/TvsT/tvtiega.pdf>

                                  although I am not sure of the copyright implications.

                                  This is a nice little study performed by CBS Music Instruments, whom I
                                  believe owned Fender at the time, which conducted double blind A/B
                                  tests on a tube and solid state guitar amplifier using "professional
                                  and semi professional guitar musicians" as subjects. While the whole
                                  study is worth reading, a short summary of it is that as long as
                                  neither amp was driven into clipping, only one out of 12 subjects
                                  could reliably discern any difference between the two amps. When the
                                  amps were driven into clipping, 2 out of 6 subjects could reliably
                                  discern the difference.

                                  One of the notable differences in the tube amp was ripple
                                  intermodulation distortion, a phenomena where the 120 Hz ripple present
                                  on the power supply modulates the desired tone, resulting in sidebands
                                  at multiples of 120 Hz from the intended signal. Upon training, the
                                  subjects could identify the tube amplifier using this distortion.
                                  Beefing up the power supply reduced this distortion as one would
                                  expect.

                                  There is more to the study and it is well worth reading. It is
                                  scientific and systematic. It is not perfect, but most scientific
                                  studies are not.

                                  I understand some people prefer tubes to solid state devices. I have
                                  gear with both. My preference is for accurate reproduction of signals,
                                  regardless of the active devices involved. In my experience this is
                                  most easily and economically accomplished with modern solid state
                                  devices as the active component. "Harmonics and coloration" in the
                                  output of an amplifier, if not present in the original signal to be
                                  reproduced, are distortion if introduced by the active device.

                                  This will be my last post on the subject. I do not wish to prolong this
                                  discussion. I agree to disagree. You may have the last word.

                                  This post was subject to the 24 hour rule.

                                  Back to regeneration. - Mr. Megacycle KK6MC/5
                                • macrohenry
                                  ... Done! After reading your application notes and your posting, I made some changes in both the ckt and the drawing. Latest revision is here:
                                  Message 16 of 18 , Sep 10, 2006
                                  • 0 Attachment
                                    --- In regenrx@yahoogroups.com, "n1tev" <n1tev@...> wrote:
                                    >
                                    > Hi Mac,

                                    > If I were you, I would redraw this circuit to show the op-amps
                                    > inside the chip...it's hard to understand unless you do this.

                                    Done! After reading your application notes and your posting, I made
                                    some changes in both the ckt and the drawing. Latest revision is here:

                                    http://www.beecavewoods.com/hobbies/TL082amp2006-9-9-750.jpg

                                    I added a volume control, added input capacitors, beefed up the output
                                    transformer, and added feedback from it. Having had success with a 70
                                    volt line transformer that is now harder to find, I went to Radio
                                    Shack and tried a 12.6 VAC power transformer. It was about $6.00 and
                                    works just fine. They have another for about $4.00 that would
                                    probably work as well. The impedance transformation is about 100:1,
                                    so the TL082 sees about 800 ohms.

                                    > And...I would continue to use two 9V batteries as this gives you 18V
                                    > (+9V,-9V swing) rather than a only a +4.5, -4.5V swing. It will
                                    > probably sound better using dual supplies too.

                                    I'm pretty amazed at how good and how loud a TL082 can sound. Using FM
                                    radio as an input, this ckt has better bass and sounds smoother than
                                    the LM386, which admittedly has more highs, but compared to the TL082,
                                    the 386 sounds sort of raucous and "in your face," perhaps in part due
                                    to using a single 9V battery.

                                    Of course, these last comments have to do with personal preference:
                                    Even using the center tap on the transformer makes a difference in the
                                    balance of the sound, hard to tell what I like better.

                                    I then followed Charles's suggestions for making a virtual ground. It
                                    works fine with single battery operation, though the volume is much
                                    lower, as one would expect. The added components for single supply
                                    operation are shown in red at the following URL:

                                    http://www.beecavewoods.com/hobbies/TL082amp2006-9-9%20singlesupply750.jpg

                                    That's all I have to report on this project. It will be a part of the
                                    low parts count AM-FM console I'm planning. Thanks to all who helped!

                                    Macrohenry
                                  • wd4nka
                                    Just to help steer the subject of audio amps back into the List topic, i have been considering using an old Australian Hi Fidelity design which Ross Hull s big
                                    Message 17 of 18 , Sep 11, 2006
                                    • 0 Attachment
                                      Just to help steer the subject of audio amps back into the List topic, i have been considering using an old Australian Hi Fidelity design which Ross Hull's big brother introduced to the ARRL during his 1936 visit to their headquarters. Seems that the Aussies had been pairing up high fidelity amps to their regens for years prior to the Hi Fi bug biting the broadcast officianadoes here on our shores.

                                      The design took advantage of the newly introduced 6L6. It has been used in music amps for years afterward. The original design called for a 6C5 driver, the cathode of which was tapped to drive the grid of one 6L6, the plate tapped to drive the other 6L6. No actual drive transformer was used. They called it a Kathodyne. The circuit and photos of a breadboard version was published in one of the '36 QSTs. If i can get a free afternoon, i'll go out to the mess i call an office and wade thru my 30's QSTs and see if i can actually find it. What is so attractive to me about it is the extremely low parts count and use of only one transformer, the output.

                                      Bet that'll perk up the ol' 'genny. :>)

                                      gary // wd4nka


                                      ----- Original Message -----
                                      From: James Duffey
                                      To: bcarling@...
                                      Cc: regenrx@yahoogroups.com ; James Duffey
                                      Sent: Sunday, September 10, 2006 10:49 AM
                                      Subject: [regenrx] Re: Simple Audio Amp


                                      Brian - I apologize to you and the list for not making my point on tube
                                      and solid state amplifiers clearer and by cluttering it up with things
                                      that were somewhat unrelated to the original point. The point that I
                                      was trying to make is that the configuration of the amplifier, how it
                                      is biased, the type of feedback used, the load it is driving, the
                                      source driving it, the noise floor, and how much headroom is available
                                      at the output and input is much more important in designing amplifiers
                                      that accurately reproduce the input signal at the output than is the
                                      type of active device used. N1TEV made the point that I wanted to, only
                                      better. Thanks Charles.

                                      You requested scientific answers. This prompted me to dig through my
                                      old files. There is an excellent paper on the subject we were
                                      discussing:

                                      W. Bussey and R. Haigler (1981); "Tubes versus transistors in electric
                                      guitar amplifiers"; IEEE International Conference on Acoustics, Speech,
                                      and Signal Processing, V. 6 pg. 800

                                      I believe this is available on the web at:

                                      <http://milbert.com/articles/TvsT/tvtiega.pdf>

                                      although I am not sure of the copyright implications.

                                      This is a nice little study performed by CBS Music Instruments, whom I
                                      believe owned Fender at the time, which conducted double blind A/B
                                      tests on a tube and solid state guitar amplifier using "professional
                                      and semi professional guitar musicians" as subjects. While the whole
                                      study is worth reading, a short summary of it is that as long as
                                      neither amp was driven into clipping, only one out of 12 subjects
                                      could reliably discern any difference between the two amps. When the
                                      amps were driven into clipping, 2 out of 6 subjects could reliably
                                      discern the difference.

                                      One of the notable differences in the tube amp was ripple
                                      intermodulation distortion, a phenomena where the 120 Hz ripple present
                                      on the power supply modulates the desired tone, resulting in sidebands
                                      at multiples of 120 Hz from the intended signal. Upon training, the
                                      subjects could identify the tube amplifier using this distortion.
                                      Beefing up the power supply reduced this distortion as one would
                                      expect.

                                      There is more to the study and it is well worth reading. It is
                                      scientific and systematic. It is not perfect, but most scientific
                                      studies are not.

                                      I understand some people prefer tubes to solid state devices. I have
                                      gear with both. My preference is for accurate reproduction of signals,
                                      regardless of the active devices involved. In my experience this is
                                      most easily and economically accomplished with modern solid state
                                      devices as the active component. "Harmonics and coloration" in the
                                      output of an amplifier, if not present in the original signal to be
                                      reproduced, are distortion if introduced by the active device.

                                      This will be my last post on the subject. I do not wish to prolong this
                                      discussion. I agree to disagree. You may have the last word.

                                      This post was subject to the 24 hour rule.

                                      Back to regeneration. - Mr. Megacycle KK6MC/5






                                      [Non-text portions of this message have been removed]
                                    Your message has been successfully submitted and would be delivered to recipients shortly.