Question on non-inverting gain stage

[bufferz]

In looking at non-inverting gain stages I have noticed some designs will tie this capacitor to ground and others to bias voltage - i have illustrated this as a switch below. can someone please help me understand explain the logic behind this variance? Is one method superior to the other?

[PBE6]

This confused me a lot too.

One important item that isn't included in your diagram is the bias resistor. The bias resistor provides a DC voltage for the AC audio to swing around. This DC voltage is present in the signal in the feedback loop as well. In single supply circuit like a battery-powered stompbox, Vbias is usually about 4.5V while ground is at 0V. If you connect the feedback loop to ground through a resistor, the 4.5V and 0V potentials fight each other and things get ugly very quickly. Adding a capacitor to the ground leg ensures that the DC bias in the feedback loop is decoupled from ground and the opamp can function as expected using Vbias as its reference.

Further to the above, the capacitor will decouple the feedback loop from any other DC voltage too, so it doesn't make any difference if you feed the ground leg to 4.5V (Vbias) or 0V (ground). Having a capacitor in the ground leg makes the connection much simpler. If you leave out the capacitor, you must connect the ground leg to Vbias.

So how does this affect the AC signal? Well, luckily AC doesn't care about DC potentials or large capacitors, so it finds its own way to ground through the capacitor (and actually through the + terminal of the battery too!). Connecting the ground leg to Vbias or ground doesn't affect the AC signal either way, other than through the high-pass shelving effect that the resistor/capacitor combination in the ground leg provides.

[bufferz]

Hey thanks - great answer this makes total sense to me now. Cheers

[PRR]

> capacitor to ground and others to bias voltage

If bias voltage is solid (large cap), it makes no difference. Whichever is easier (closer).

[bufferz]

Thanks, I do have another observation/curiosity - I have noticed that on inverting buffers, summing amps, or even a bax eq stage some people tie the positive input of opamp to bias voltage directly and others will have it ted to bias voltage through a 47k resistor...whats the point of that resistor?

[PBE6]

It's to reduce voltage offset caused by input bias current, although it's rarely necessary:
https://e2e.ti.com/blogs_/archives/b/thesignal/archive/2012/04/11/input-bias-current-cancelation-resistors-do-you-really-need-them

[PRR]

> whats the point of that resistor?

Old-old texts told you to do this. It is almost always a bad idea in Audio.