5532 OpAmp distortion

[FireAce]

Greetings, having a bit of a problem with a headphone amplifier circuit (attached). I'm certain I've had this working in the past but cant remember.
Uses a NE5532 Op amp, I'm feeding it +_ 15.39 volts. The circuit is perfect for about 3 seconds, nice crisp sound, plenty of volume, no noise, then boom it starts distorting until theres no more volume period.

Any ideas?
Thanks

 

[audioguru]

What is the impedance of each ear of the headphones? If it is less than 200 ohms then the dual opamp will get too hot when playing loudly.
Measure the plus and minus 15V supply to see if it is dropping when playing loudly.
Double-check the value of the 33 ohm resistors.

 

[FireAce]

The opamp is actually not heating up or even showing signs of it. I've tried it with audio level as quiet as possible and it reacts the same. I've tried lowering the voltage, removing the 33 ohm resisters, etc.

Just popped in a tl082 in its place and it appears to be working fine. Might just have to settle for less.

 

[throbscottle]

Monitor the DC voltages at the op amp outputs as the volume goes - I bet you a smiley it floats away to either supply voltage.

 

[audioguru]

Monitor the DC voltages at the op amp outputs as the volume goes - I bet you a smiley it floats away to either supply voltage.

Yes, good observation.
I also notice that its volume control biases both opamps and the value of the volume control is very high at 250k ohms. The maximum input current is 800nA for each opamp so the input offset voltage could be 0.4V. Each opamp has a voltage gain of 11 which stupidly includes DC so the output offset voltage could be as high as 4.4V.

The TL082 has Jfet inputs with an extremely low input bias current so it does not have this problem.