Connect with us

Copying Op Amps to Make Mic Amps

Discussion in 'Electronic Design' started by D from BC, Mar 17, 2007.

Scroll to continue with content
  1. D from BC

    D from BC Guest

    Wow ...3 threads at same time for mic amps.
    "Improved mic amp posted on Rapidshare"
    "A Better Microphone Amplifier Yet"
    "Is S.E.D. actually sci.electronics.dummies"

    I'm surprised mic amps are a lively topic on here.
    I thought it would be a beaten to death topic like audio power
    amplifier design.
    Also, by now shouldn't the ultimate microphone amp be a well known
    cookbook circuit?.
    Ever since "Mary had a little lamb" (history) there's been plenty of
    time to master mic amps.
    Is mic amp design still in fuzzy land?

    Copying Op Amps to Make Mic Amps
    I'm sure the best op amp topologies have been thought out by
    How about just (partially) copying a very good op amp internal design
    with discrete low noise transistors for the front end mic circuit?
    I recall somebody posted that it's the IC process that creates noisy
    transistors in op amps.
    D from BC
  2. Joerg

    Joerg Guest

    Nah, the cool thing to do is to take a few ECC85 tubes and copy the
    opamp with those. Then give the glass of the tubes a spit shine for even
    lower noise :)))
  3. yep it's a fun subject
    But why copy opamp circuits in the first place, they just seem to have
    copied the discrete version

    FWIW this is a newish guy on the block

  4. Eeyore

    Eeyore Guest

    So am I actually. I'm stunned that the self-appointed 'experts' on here don't
    understand the fundamentals of what makes them good for their intended
    application. Even the simple versions, never mind the hot-rod one I've posted in

    LOL ! You want me to discuss that too ?

    I recall my lecturer at college saying that pole-zero compensation was "too
    difficult to analyse". Kinda got my interest that did !

    You're making the flawed assumption that those who put together 'cookbooks' are
    competent in this field.

    There's nothing fuzzy about it at all. All that's needed is competent design. A
    lack of which is seriously evident even here.

    A mic amp is a form of instrumentation amplifier. Op-amp topologies are not

    ICs have some difficulty replicating the kind of really low-noise discrete
    transistors that are possible. In fact they've never ever been able to
    realistically compete.

  5. Jim Thompson

    Jim Thompson Guest

    You forgot to mount it inside a glass case, with a brass name plate

    ...Jim Thompson
  6. Eeyore

    Eeyore Guest

    Yup, it's an integrated copy of an Australian discrete design IIRC.

  7. and it took Thatcorp a *very* long time to get it right

  8. Joerg

    Joerg Guest

    Ah, you haven't seen the real audio freaks yet. Glass cases are frowned
    upon. It has to be a polished base and the tubes must stick out. Also,
    preferably there should be a vintage Gibson casually sitting next to the
    amplifier even if you can't play the guitar. A lava lamp can't hurt
    either. A 1960's VW bus with a peace sign in the driveway would also be
    a nice touch.

    Glass is decidedly not cool. Smoked glass would be the worst, as that is
    identified with the previous generation. I remember a writing on a wall
    in one of those freak homes: "We are the people our parents always
    warned us about".
  9. D from BC

    D from BC Guest

    lol :)
    D from BC
  10. D from BC

    D from BC Guest

    Ooooo :)
    THAT corp. 1510 @ 1nV/root hz input noise (60dB gain)

    I don't need a mic amp but now I feel like making one :)
    D from BC
  11. Rich Grise

    Rich Grise Guest

    Here's the one I used:
    with appropriate resistor and cap values, of course. :)

  12. Rich Grise

    Rich Grise Guest

    Pfeh - ECC85 is a 6AQ8 - RF local osc. and mixer. Use ECC83 == 12AX7,
    optimized for low-noise audio. :) (the ECC81 (12AT7) and ECC82 (12AU7)
    are optimized for switching and RF, but I can't remember which is which.)

  13. Joerg

    Joerg Guest

    Guess you are right about that. I didn't do much audio stuff with tubes
    and if so I took what was there for free from old TV sets. My area was
    RF and there I used ECC81 and UHF tubes from tuners (EC86?).

    Oh, I did one audio project. It later became a big RF amp for ham radio
    but first I turned it into a Gorilla of an audio amp. Mostly because of
    a bet. Not about money, someone said it would never work and I just
    could not let that sit. That thing was awesome. But you could not crank
    it up and then play a strong chord on the E-guitar because the mains
    breaker would trip. Other than that it really made the rafters shake.
  14. Tim Wescott

    Tim Wescott Guest

    Don't forget to dip them in LN2 -- it "aligns the crystalline structure"
    for better sound.


    Tim Wescott
    Wescott Design Services

    Posting from Google? See

    "Applied Control Theory for Embedded Systems" came out in April.
    See details at
  15. Jim Thompson

    Jim Thompson Guest

    Oh, John! You're giving away my secrets ;-)

    ...Jim Thompson
  16. Ian Bell

    Ian Bell Guest

    Mic pre design is a complex trade off. There is no perfect design, that's
    why there are so many.

  17. John  Larkin

    John Larkin Guest

    Analog ICs typically use lots of transistors and few resistors or
    capacitors, the reverse of most discrete design. And they can match
    the transistors for offset and beta orders of magnitude better than
    you can with discretes. That's why a simple-looking analog integrated
    circuit may have hundreds of transistors. So the designs are usually
    quite different.

    Some analog processes can use transistors as good as anything you can
    buy as discretes. Opamps like LT1028 have noise below 1 nv/rt hz.

  18. John  Larkin

    John Larkin Guest

    Yeah, but you don't have integrated ferrite beads!

  19. Eeyore

    Eeyore Guest

    Where's your 500pV/sqrt Hz input deice ?

  20. John  Larkin

    John Larkin Guest

    Transistors are all made out of the same silicon, integrated or
    discrete, so the physics is the same. The transistors in the front end
    of an LT1028 or INA103 are about that noise level. There will always
    be niches where you can squeeze a little more performance out of a
    discrete, but with the exception of really big jfets, any advantage is
    usually small. The IC boys can just throw so much more resources at a
    design... gain-bandwidth, matching, tempco, bias current cancellation,
    current sources galore, superbeta, and enough sales volume to pay for
    a *lot* of engineering.

    As the old stuff gets integrated, the only thing to do is move up the
    abstraction stack. As a longtime circuit designer, I sort of regret
    that, but that's life. You can't keep designing the same circuit all
    your life.

Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day