Biasing between the two halves of an op-amp

[Fabei]

Hi all guys!
I have a little doubt! When using an op-amp (TL072) I usually bias the input on the first half (a fixed gain stage for example, or a buffer) and when from the first output the signal goes to the second half I bias it again.
I've checked many schematics and sometimes it's done this way and sometimes not. On my last build I didnt bias the second input, and not even used a cap between the two halves (buffer and gain stage) and it worked good.
Is there any pros or cons to do it one way or another?
thanks all!

[midwayfair]

Breadboard a dual op amp. Bias the first op amp and run a piece of wire from pin 1 to pin 5. What's the voltage on pin 5?

Now bias pin 5 separately. What's the voltage on pin 5?

If there is a difference, what effect do think that difference might have?

What effects does putting a capacitor in the signal path have? Jot a few of them down (whatever you can think of) and check the wikipedia capacitor article to see if you missed a few. Are they good or bad effects? Under what circumstances might a "good" one be bad?

What situations would you NOT be able to do the first method? Hint: Look at the input of any simple op-amp based pedal (a buffer is simplest). Breadboard the extra components and measure voltages everywhere if you need to.

[Fabei]

Breadboard a dual op amp. Bias the first op amp and run a piece of wire from pin 1 to pin 5. What's the voltage on pin 5?

Now bias pin 5 separately. What's the voltage on pin 5?

If there is a difference, what effect do think that difference might have?

What effects does putting a capacitor in the signal path have? Jot a few of them down (whatever you can think of) and check the wikipedia capacitor article to see if you missed a few. Are they good or bad effects? Under what circumstances might a "good" one be bad?

What situations would you NOT be able to do the first method? Hint: Look at the input of any simple op-amp based pedal (a buffer is simplest). Breadboard the extra components and measure voltages everywhere if you need to.

Possibly the best answer I could ever get!
I was missing two points here: one about biasing and one about empirically searching and discover on my own!

[madbean]

Just to elaborate a bit on Jon's excellent post - re-biasing depends partially on whether or not you are coupling different stages with a cap. When not coupling, DC offset might play a factor esp. when going between two different IC types. But it really depends on how it's being done. At the same time, there is a lot of forgiveness in pedal designs so sometimes you can get away with things an engineer would frown upon.

[Fabei]

Just to elaborate a bit on Jon's excellent post - re-biasing depends partially on whether or not you are coupling different stages with a cap. When not coupling, DC offset might play a factor esp. when going between two different IC types. But it really depends on how it's being done. At the same time, there is a lot of forgiveness in pedal designs so sometimes you can get away with things an engineer would frown upon.

Thanks for answering!

Yep I realized that if the first bias is setting a reference point from which the IC Output can swing up and down so if connected directly to the second I shouldn't need to bias again. If I put a cap in there I have to re-bias.

Regarding DC offset I think that going from a biased buffer stage to a gain stage on the same IC shouldn't be really a problem.
But DC offset and what may cause it is something I'm learning recently (I've recently decided to understand things before building it so I can possibly debug them), and the only thing I to prevent DC offset right now is that's almost mandatory to use a cap to GND on the voltage divider that provides the bias current