Reading schematics: why stuff goes to ground

tonymcbony · Post by **tonymcbony** » 28 Jul 2011, 01:57

Alright, the great unveiling of my ineptitude.

I have a hell of a time visualising the flow of current or understanding why stuff that's sent to ground actually does anything.

For example, look at any tubescreamer or rat schematic where the clipping diodes go to ground.
I know that it's clipping the signal and that's where we're getting the distortion from but how the hell does it get from there to the output? How are we hearing anything if it's going to ground?

This applies to anytime something goes to ground: in my head, going to ground means just dumping it out of the circuit, getting rid of it.

I guess this is one of the main things that is stopping me from moving on from being a paint-by-numbers solder jockey.

DrNomis · Post by **DrNomis** » 28 Jul 2011, 03:59

tonymcbony wrote:Alright, the great unveiling of my ineptitude.

I have a hell of a time visualising the flow of current or understanding why stuff that's sent to ground actually does anything.

For example, look at any tubescreamer or rat schematic where the clipping diodes go to ground.
I know that it's clipping the signal and that's where we're getting the distortion from but how the hell does it get from there to the output? How are we hearing anything if it's going to ground?

This applies to anytime something goes to ground: in my head, going to ground means just dumping it out of the circuit, getting rid of it.

I guess this is one of the main things that is stopping me from moving on from being a paint-by-numbers solder jockey.

When the signal reaches a certain voltage level that's equal to the diode's forward voltage, the diode starts conducting current, because there are two diodes in the circuit which are arranged back-to-back, and in opposite directions, each diode will conduct on it's respective half of the waveform, clipping off the top and bottom peaks of the waveform, because the diodes go into and out of conduction very quickly, only a very small amount of current flows at any one time....

JudgeMingus · Post by **JudgeMingus** » 28 Jul 2011, 05:22

Visualised another way (just for variety of viewpoint - Dr Nomis' explanation is certainly accurate!) -

The diodes won't pass the signal until it reaches a certain level - this part of the signal is undiverted and goes to the output.
When the signal exceeds that level (the forward voltage of the diode), any excess is siphoned off to ground and does not appear in the output. Hence - clipped peaks on the waveforms.

This is (as I understand it...) the reason germanium diode clipping usually has less output - the forward voltage for germanium diodes is lower, so more of the signal gets dumped to ground.

tonymcbony · Post by **tonymcbony** » 28 Jul 2011, 06:50

Thanks folks, I need to clarify here.

Take a look at this http://www.aronnelson.com/gallery/main. ... ewsIndex=1

Now, in particular R6 and R7, we know that they're clipping for the reasons you've both excellently explained.
In my mind, signal is getting clipped by these diodes, and then dumped to ground, and that's where I'm having trouble.

If the clipped signal is being dumped to ground, then why can I hear the clipping?

DrNomis · Post by **DrNomis** » 28 Jul 2011, 07:09

tonymcbony wrote:Thanks folks, I need to clarify here.

Take a look at this http://www.aronnelson.com/gallery/main. ... ewsIndex=1

Now, in particular R6 and R7, we know that they're clipping for the reasons you've both excellently explained.
In my mind, signal is getting clipped by these diodes, and then dumped to ground, and that's where I'm having trouble.

If the clipped signal is being dumped to ground, then why can I hear the clipping?

That's becauser the clipping is introducing what's called "Harmonic Distortion", more specifically, it's introducing Odd-Order Harmonic Distortion in the signal, your ears hear it as a "Buzzy-ness" or "Raspiness" compared to the smooth sound of the guitar, it will also sound brighter and louder too, which is caused by alot of high-frequency content due to the signal distortion....

In the schematic, R6 and R7 are part of the "Negative Feedback System", R6 and R7 work in conjunction with the gain control pot to control how much gain the Op-Amp IC produces, remember that the Op-Amp has a + input and a - input, the Op[-Amp IC is known as a "Differential Op-Amp", that is, it "Amplifies" the "Difference" between the + and - inputs, to be able to do that, the - input has to be referenced to ground so that the Op-Amp can see a difference between the signals on the two inputs, normally the signal that we want to amplify is fed into the + input, the - input is used to control the gain by feeding a certain amount of the output signal into it, because the - input is an "inverting input" any signal fed into it will come out the output flipped upside down (180 degrees out-of-phase), this has the effect of reducing the output, the reason why negative feedback is applied to Op-Amps is that they are capable of very huge amounts of gain, up to about a million times the input signal, this amount of gain would make the output easily swing from one extreme to another, if you were to feed in a sinewave, the result would be a square wave with very sharp corners, so, to bring this huge amount of gain under control, we feed a certain amount of the output back into the - input, incidentally, the + and - inputs are at very high impedances and therefore don't need much current drive at all, the point where R6 and R7 connect to the - input can be thought of as a "virtual-earth"....

The maximum amount of gain that you will get out of the Op-Amp is determined by the ratio of R6,R7, and whatever the gain control pot happens to be, reduce the resistance of the gain control pot and the gain gets reduced, and vice versa....

Mr.Grumpy · Post by **Mr.Grumpy** » 28 Jul 2011, 16:02

Most simple schematics like fuzz circuits follow some basic rules...these are general rules and don't alway apply, but here goes:

1) The signal flows from left to right. Note that on almost all schematics, the input is on the left, the output on the right.
2) The DC (bias) currents flow vertically, from V+ (or V-) to ground.

Now, back to your original confusion about clipping diodes. Yes, if you replaced your clipping diodes with wires, then indeed all the signal would be grounded and your overdrive would make no sound. However, diodes don't act like a straight wire, even when they are forward biased. I hate to bring up the "water analogy" that compares water to electric current, but it helps a lot of people understand. Imagine that a diode is a sort of one way valve, and that the "flapper" is held closed with a spring. You can understand then that it would take a certain amount of water pressure (voltage) to "push the spring open" and allow the water (or current) in our case to flow. Once the pressure drops again, the flapper shuts and stops the flow. So the clipping diodes don't shunt all the current to ground, they just limit the amount of "pressure" (voltage) that can occur on the wire. Since there is always some 'back pressure' from the diode, the voltage does not go to zero. 0.3 or 0.7 volts may not sound like much, but it is plenty of voltage for the input of a guitar amplifier.

jonfoote · Post by **jonfoote** » 29 Jul 2011, 12:35

i think the answer your looking for is that an audio signal has both a + and - voltage and ground sits right in the middle. i.e. ground is not the lowest limit, if audio gets louder it gets more + and more - voltage.

http://www.google.co.uk/imgres?q=audio+ ... 9,r:7,s:15

if you look at that picture, ground is the line that goes through the middle

tonymcbony · Post by **tonymcbony** » 01 Aug 2011, 05:37

Thanks everyone, you've all helped hardcore!

Nocentelli · Post by **Nocentelli** » 01 Aug 2011, 22:36

Further to the picture above from jonfoote, if you look at a normal and a "clipped" waveform side-by-side, you can sort of see why the clipped one sounds more ragged and distorted, and that only sound of an amplitude above a certain threshold gets clipped:

The clipped version also appears slightly quieter than the normal version, although in a pedal circuit the overall amplitude of the signal can be boosted before or after the clipping stage to bring it back to unity: You can also see how this clipping/distortion/fuzz effect reduces the dynamic range of the sound.