Can someone tell me if I've figured this out?

[Samuel]

Opamp biasing: am I understanding this correctly? The 9V source becomes the V+ for the opamp. In order to bias correctly, 4.5V is generated from the 9V source, and this is coupled to the AC signal (input from the guitar). This 4.5V acts as virtual ground to the opamp, causing actual 0V ground to appear to the OA as -4.5V. Am I getting this?

[smoguzbenjamin]

Yup :mrgreen: That's one way of doing it.

[Mark Hammer]

Here's another way to think of it.

AC signals fluctuate from above (positive) to below (negative) ground, right?  Ground is the *reference point* of that fluctuating signal such that it goes negative or positive relative to that reference point.

Okay, if you power an op-amp with a single-ended supply using a 9v battery, where is the reference point for that AC signal?  There certainly isn't one provided by the power-supply scheme since there IS no negative supply.

Obviously, one has to invent or fake such a point of reference, which is why you see these invented grounds referred to as "Vref", Vr", "floating ground", or "bias".  This is a purely artificial point of reference that AC fluctuates around.

What should Vref *be*?  Keep in mind that op-amps cannot always swing from one end of the power supply rails to the other.  An "ideal" op-amp would be able to provide a +/-4.5 (i.e., 9vac peak to peak) output, but not all are so ideal.  In many instances, op-amps are only able to swing within a volt or so of the power-supply extremes.  With a 9v supply that means a total possible voltage swing of 7v.  That limits how much clean headroom we have.  Although there *are* instances where you deliberately want ot have more possible swing in one direction than the other, in most applications one wants the two half-cycles of the AC signal to share the range of voltage swing equitably.  That means setting up a Vref which divides the entire range of voltage swing in half (or close to it) so each half has a reasonable amount of headroom.  This is why you always see Vrefs coming from the junction of a pair of equal-value resistors: they divide the supply voltage in half evenly for each half cycle to have reasonable headroom above and below that reference point.

[Samuel]

Sweet. I was actually thinking correctly about this. Yay. Plus I get great info re: rail-to-rail swing in OAs. Little by little I creep towards confidence in this stuff...

Thanks so much, Ben, Mark, it is very much appreciated.

[EdJ]

Mark,if you ever find the time,can you write a book about electronics and their use in stompboxes?Please.
You are makeing whole chapters clear in one sentence!
Greetings,Ed

[aron]

In case anyone is wondering. YES, I have tried to alter the biasing on a op amp by making VREF alterable.

NO, nothing amazing came out of it.

Oh well. Try it yourself  :roll:

[smoguzbenjamin]

Hmmm. I'm tiring of my TL072s. They clip way too fast for my taste.

[Transmogrifox]

If you use an op amp that sounds good clipped, you could make an adjustable symmetry with a pot by making it able to vary the virtual ground reference.  To make a better clipper out of it, regulate the opamp supply voltage to about 5 or 6 volts, then make your central virtual ground reference at 1/2 the opamp supply. Put in a pot so you can adjust the symmetry here, and now you have a single "vaccuum tube model".  Feed this stage into an identical inverting stage (drive both in the inverting configuration) and you have something similar to the average tube sim overdrive unit.  With the proper frequency adjustment, one could have a fine OD pedal out of this.  Use a CMOS op amp like a TLC272 or so...

[smoguzbenjamin]

Hmm yeah but I was using them in an acoustic simulator which needs as much clean headroom as possible :mrgreen:

Thank god for 4558's. Somehow they're just better. Or it might be that i bashed my 072 a little hard. Electron damage