OK, so I've seen on some audio sites that people are using constant current sources on the output of op-amps (but inside the feedback loop) tied to the V- rail. So it will be constantly driving current even with 0V output. They usually do this in headphone amplifiers to avoid any kind of crossover distortion due to the low signal levels.

Is there any merit to doing this at all? Is it really class A operation?

 

Those fools use an OPA2134 dual opamp that has 0.00008% distortion that is so low that it is completely inaudible and is extremely difficult to measure. Then they add a biased load to it so that only one of its output transistors operates in class-A instead of both output transistors operating in class-AB.
It might end up with more distortion. It might melt.

Attached Images

OPA2134 details.PNG (17.2 KB, 14 views)

__________________
Uncle $crooge

 

Really it was only ever of use on one specific opamp (or series of opamps), the LM324 if I remember correctly? - I've no doubt Audioguru will remember and correct me?.

To reduce power consumption the output stage was run in class B, so you got cross-over distortion - by forcing the output stage to be single ended, you got a low power, less distorted opamp. However, far easier to use a better opamp in the first place!, and probably better performance as well.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

The old LM324 quad opamp was the first low-power opamp. They reduced its supply current by removing the bias voltage between the output transistors so they operated in cutoff for the middle 1.2V of a signal. The crossover distortion is horrible. Negative feedback doesn't help much to reduce the distortion because the low power opamp has poor high frequency response.

Adding a DC load allows one output transistor to operate in class-A.
The distortion is 3% when it has 100% negative feedback reducing it.
The distortion drops to 0.0018% when the output operates in class-A:

Attached Images

LM324 crossover distortion2.PNG (24.7 KB, 10 views)

__________________
Uncle $crooge

 

Told you Audioguru would know!

And I got it right as well! - I thought it was the LM324.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

 

Of course you are correct and you have a good memory.

The old LM358 dual low-power opamp uses the same old opamps as the old LM324 quad.

__________________
Uncle $crooge

 

Thanks guys! I knew I was doing right asking here. I thought it was quite silly to do using something like the new Burr Brown or Analog Devices parts, but I'm really surprised it actually was legitimately used on an older part. At least now I know where they got the idea

My op amp stash consists entirely of OPA2132/2134/NE5532 chips, so I have zero need for this. I've been meaning to get some TL072's just to have, but they shouldn't need this either.

 

The TL072 is rated at 0.003% distortion. You might hear distortion if it is 67 times more (0.2%).

__________________
Uncle $crooge