boost pedal

Discussion in 'The Projects Forum' started by babb27, Jul 19, 2011.

  1. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    hey i am trying to design a boost pedal for my guitar.

    here are the specifications i designed it to:

    gain = 4 or 12dB
    single supply 9V
    input impedance of 100k
    output impedance of 20k

    when i run the simulation i come across some interesting results.

    it does not have the correct biasing i chose ( 4.5V )
    and the signal is clipping.

    attached is the schematic and simulation results.

    please help or give any advice.
     
  2. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    i should clarify about the gain. i mean Av = 4 which is about 12dB

    i realize now this maybe too much and a gain of 6dB will do the job. is this where my problem in the design arose from? the trade offs of gain/input impedance/voltage headroom?
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I'm wondering why you didn't use an operational amplifier?
    The input impedance would be very high, output would be very low, and your distortion would be very low.
     
    babb27 likes this.
  4. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    that is a good idea lol. keeping it simple for sure. i will go ahead and try that out. thank you
     
  5. Adjuster

    Well-Known Member

    Dec 26, 2010
    2,147
    300
    Your circuit has an emitter resistance of 760KΩ, and ignoring base current loading, the base is biased to half the 9V supply.
    That works out at an emitter current of 3.9V/760KΩ or around 5μA. With such a tiny emitter current, the possible undistorted output swing into the 24KΩ collector load will be less than 120mV peak.

    The fact that you are getting any significant collector swing at all may be due to pulses of current charging the huge (1 Farad!) emitter bypass capacitor - if you ran the simulation long enough these might cease.

    You really need to look up a decent text on the four-resistor bias method, and note that to get a voltage gain as low as 6dB you would normally leave all or part of the emitter resistance un-bypassed, rather than selecting a minuscule emitter current. The un-bypassed emitter resistance will also make it easier to achieve a large input impedance.
     
    babb27 likes this.
  6. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    the reason i used the 1F capacitors in the simulation was because this was the practice in my electronics class last semester. any bypass capacitors were assumed to be infinite for the ideal case and we used the 1F to simulate that.

    in practice i will obviously have to choose a value that will be suitable for the frequency range of the guitar.

    thanks for the help, and i will try again tomorrow.
     
  7. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    thanks Adjuster and SgtWookie, i got the pedal designed and will be building it sometime this week
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Why don't you post your designs' schematic so that we can give it a design review?

    It's easier to fix any problems discovered before it's built than afterwards.
     
    babb27 likes this.
  9. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    good idea. so attached is the schematic and simulation with an input of 100mV. when changed to 1V input it is similar with a gain of 2.

    EDIT: the resistance value of the collector is 5100
     
  10. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    An AC analysis from 100Hz to 20kHz shows a roughly 3.317dB gain, and roughly 180° phase shift at low frequency, and 179° at high frequency.

    Your plots are very coarse. You must be allowing quite a bit in your timestep. Cut that down considerably, 1/10th to 1/20th the size it is now.

    You're still using 1 Farad caps, which isn't realistic - you're not going to use caps that large in a real preamp. Try changing C14 to between 10uF and 47uF, and twice that for C15.

    You don't have any feedback path, so your gain will be inconsistent, and will vary widely over temperature.
     
    babb27 likes this.
  11. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    looks like i have a lot more to consider than i thought.

    i will try out your idea with the op amp. and start reading about feedback and designing so insensitive to beta and temperature changes.

    thanks for the help
     
  12. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    hey wookie, what simulation program are you using? i ran an AC analysis with the capacitors changed to your suggested values and my results are attached.

    the gain looks pretty consistent over the frequency range of the guitar and the 20k audio range. I am confused by what you mean by inconsistent gain. do you mean inconsistent when the amplitude of the input signal changes?
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I am using Linear Technology's very good and absolutely free LTSpice.
    There is plenty of support for it on the Yahoo! LTSpice Users' Group.

    Inconsistent over temperature and output load. That might not be a problem for you.

    You currently have 100k for an output load. Typical line-level impedance is 10k Ohms. Since your circuit has no feedback, it will not be able to compensate for variances in output impedance.

    If you used a dual opamp, you could use one of the channels as a voltage follower to buffer the input, providing a very high impedance, and use the other channel to provide whatever gain you want (within reason), and a rather low-impedance output.

    You would not have to worry about stability over temperature, varying output loadings, etc. - as both amps would have feedback, so would automatically correct the signal.
     
    babb27 likes this.
  14. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Here, I threw a simulation together for you using an MC33078 dual opamp.

    Have a look at the attached.

    C1 passes the AC signal, blocks the DC level.
    R1/R2 establish Vcc/2, a "virtual ground".
    R3 keeps U1a+'s level averaging at Vcc/2.
    U1a provides a high impedance input and low impedance output buffer.
    U1b with R4/R5 provides a gain of 2.
    C3 passes the AC signal, blocks the DC level; V(out) averages 0v.
    The value of R6, as long as it doesn't go below ~2k, won't affect the amplified signal level.
     
    babb27 likes this.
  15. babb27

    Thread Starter New Member

    Mar 16, 2011
    21
    1
    sorry i should have been more specific in my first post. the 100k will be a potentiometer that will be used as a volume knob.

    i really like that design. it is significantly more robust than mine. also i may switch over to this LTSpice from Pspice. Does that program allow for schematics with more than 64 connections? Anytime I try to make bigger schematics in Pspice it says I exceed the connection limit.

    also your output is in phase. for audio applications do i need to be concerned about a phase shift of 180? would anything less or more cause a delay?
     
    Last edited: Jul 21, 2011
  16. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I see.

    LTSpice does not have any limitations on connections or components.
    It comes with a pretty decent library of their own components, but if you want to model other manufacturer's parts (IC's in particular), you'll need to add them yourself.
    There are a number of diodes, caps, transistors and MOSFETs that come with it.
    You can find many other models at the Yahoo! group, or by using Google, or by visiting manufacturer's websites. SPICE models are not available for every component you might wish to simulate, particularly in digital IC's. LTSpice is much more oriented towards analog than digital. There are some 74HC and CD4000 series logic IC's available on the Yahoo! group, but it is by no means complete.
    IBIS models are not supported.
     
    babb27 likes this.
Loading...