Cmoy amp fine on centre tapped bench supply - but distorts on resistor virtual ground

Discussion in 'The Projects Forum' started by Circuitfreak101, Apr 13, 2016.

  1. Circuitfreak101

    Thread Starter New Member

    Apr 13, 2016
    11
    0
    Hi all

    I've been building a few of the famous Cmoy headphone amps as I needed a good quality low cost headphone amp inside my latest creation and the Cmoy gets a lot of praise and is cheap. However, every time I disconnect my centre tapped bench supply and connect the battery power config or a virtual centre tapped wall wart with 2 x 4.7K resistors the amps fine at very, very low volume but distorts as soon as you attempt to get normal listening volume in 32 Ohm headphones. I notice measuring pins 4 and 8 to virtual ground that they are out of balance when running from the battery tapped or wall wart tapped supply ... any ideas ?

    Thanks
     
  2. RichardO

    Well-Known Member

    May 4, 2013
    1,238
    385
    A schematic drawing of your circuit would help a lot to know what you are doing.

    But.. My guess is that you are trying to use the virtual ground as a power connection to the 32 Ohm headphones. As a rule, power supply splitters do not work very well at levels more than a few millwatts.

    Again, a guess, but I would drive the headphones from a bridge amp.
     
  3. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    A resistor virtual ground will only work properly if there is very little current flowing through the resistors (everything connected to the virtual ground must be high impedance).
    If that's not the case then you will need a lower impedance virtual ground, likely by using some active circuits to generate it.
    Post your circuit diagram and we can better advise you.
     
  4. Circuitfreak101

    Thread Starter New Member

    Apr 13, 2016
    11
    0
    Thanks for you help guys, here's the standard circuit

    http://tangentsoft.net/audio/cmoy/misc/cmoy-tangent-sch.pdf

    As I say if I use my centre tapped bench supply instead of the suggested battery supply the circuit is fine, if I try centre tapping a wall wart with 2 resistors or centre tapping 2 batteries in series the amp distorts at very low volume in 32 Ohm cans. Swapping out the op amp for an NE5532 doesn't make a difference.
     
  5. dannyf

    Well-Known Member

    Sep 13, 2015
    1,836
    367
    You will need to buffer the virtual ground or it floats with output signal.
     
  6. jjw

    Member

    Dec 24, 2013
    173
    31
    Have you tried to take the gnd between batteries without resistors?
    Also C1 should be taken away. Two capacitors in series kills the filtering of the power supply.
     
  7. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    Since you have two 9V batteries why are you not taking the ground from between the two batteries as jjw suggested? :confused:
     
  8. Circuitfreak101

    Thread Starter New Member

    Apr 13, 2016
    11
    0
    For the application I'm using the circuit for it's not possible to use batteries. I was hoping to use an 18V wall wart and obtain a centre tapped supply by using the 2 resistors as a potential devider, 0V being their centre meeting point.
     
  9. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    My LTspice simulation showed no significant distortion with your original circuit so I'm not sure exactly what's causing your observed distortion. :confused:
    Try the center-tapped circuit with your bench supply set to 18V and see if you still get the distortion.

    If you are seeing offset in the split voltages try reducing R1 and R2 from 4.7kΩ to 1kΩ.

    You might try add a 270μF capacitor in series with each output (+ to amp) to the headphones to block DC, and connect the common connection to -9V.
    That will prevent the high headphone current from going through your virtual ground circuit.
     
  10. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,073
    3,856
    You can use this for a virtual ground and eliminate the big capacitors. Power the op amp with your 18V wall adapter.

    image.jpg
     
  11. Circuitfreak101

    Thread Starter New Member

    Apr 13, 2016
    11
    0
    Thanks for your help guys, much appreciated. I have found though that an NE5532 seems happy enough to drive enough power through 32 Ohm cans in a standard inverting config with the non inverting input biased halfway of the supply rail so I've left the Cmoy design for now.
     
Loading...