Tonestack Components Values

[POTL]

Hello
I found the information that if you reduce the resistor values ​​by 10 times and increase the capacitor values ​​by a factor of 10, the noise in the circuit decreases.
Tell me how much this information is correct?
Does noise only come from resistors or potentiometers too?
Is there any meaning or difference that will not be noticeable?

[mth5044]

Download Duncan's tonestack calculator and give it a shot!

I'm sure there are tons of qualifiers, but I have heard that said about about the typical fender tonestack.

[GibsonGM]

I don't know if "10x" is good, as you are messing with impedances...but I have heard that doubling the caps and halving resistors in COUPLING networks reduces noise quite a bit...

[POTL]

Download Duncan's tonestack calculator and give it a shot!

I'm sure there are tons of qualifiers, but I have heard that said about about the typical fender tonestack.

I have this program.
I do not remember that it showed the level of thermal noise of the resistors =)
The frequency shows, but not more.

[POTL]

I don't know if "10x" is good, as you are messing with impedances...but I have heard that doubling the caps and halving resistors in COUPLING networks reduces noise quite a bit...

I did not try to change the values ​​2 times, 10 times the calculator shows identical values.
With impedances, I figured out, now I'm not hammering this head =)
The question is only in the noise, is there any sense in doing this.
I podgelyal this idea from Brian Wampler

[Rob Strand]

capacitor values ​​by a factor of 10, the noise in the circuit decreases.
Tell me how much this information is correct?

Generally that is true *for that part of the circuit*, it decreases the noise by factor of sqrt(10).
If that part of the isn't the main contributor to noise then you will not notice any reduction in practice.

Does noise only come from resistors or potentiometers too?

Both.

Is there any meaning or difference that will not be noticeable?

Apart from what I mentioned above, you can get unwanted sided effects by scaling the values.
For example if the circuit driving the resistor and capacitor network has an impedance comparable to the
the size of the parts network then scaling down the network will start to load-down the other parts of the circuit
and this will modify the frequency response.

A classical example is the tone control driven by a collector output.  If you keep scaling down the tone control values it will eventually change the tone control frequency response and reduce the output level.   Some care is required in determining how low you can go.

One work around is to decrease the collector resistor, but that will reduce the gain of the stage.  So now you need to make more changes to that stage to fix the gain and then it might change how it sounds.

In short there is a ripple effect which needs to be considered.

[antonis]

1k & 1M resistors are of the same cost..
1μ & 1n capacitors ...??

[amptramp]

Duncan's Tone Stack Calculator:

http://www.duncanamps.com/tsc/

You are correct in saying that noise decreases as resistance decreases and this applies to pots as well as resistors.  You have to consider whether you can drive a lower resistance tone stack and the noise drops with the square root of the resistance drop.  Cut the resistance by a factor of nine and the noise drops by a factor of three.  But you have to increase the capacitor values by the same factor (in this example, nine) that you reduce the resistance values to maintain the frequency response.  If you are driving the tone stack from a low-impedance source like an op amp, there should be a negligible change in response.  If you have the FET stages with drain resistors, the drain resistor is the output impedance of the driver and changing that may produce a noticeable change in response.

The resistive thermal noise is not usually a limiting factor in a guitar amp.  Most guitar effects can make use of a TL071 as an input buffer and this has a noise level of 18 nv/(Hz)^0.5 and 0.01 pA/(Hz)^0.5 and this is adequate with a 1 megohm resistance to ground or Vcc/2.  If you have low-level signal in the millivolts, resistive noise is usually swamped by ground noise and hum from the amplifier and the rest of the circuit grounding scheme.

[PRR]

> reduce the resistor values ... the noise in the circuit decreases.

Sorry if this blows your mind, but: NO.

The hiss power is the same in every resistor.

The hiss voltage drops with lower resistors, but the hiss current rises.

And the hiss voltage drops as square-root of resistance. Which means a 10X drop is only 3X less hiss.

Whereas a 10X gain in front of the tone network directly gives nearly 10X better signal to hiss ratio.

The first circuit shown in #4 has huge gain in front of the tone network. Lowering those impedances may not be needed. It does also have a ton of gain after the tone network. A detailed study of the gain-structure would be needed to optimize that. Since it is surely some kind of "distortion", lowest hiss may not be possible with any gain or tone choices.

The second picture in #4 is Fender's classic. Note that in the original there is a gain of *50* in front of this. To move this in front of the gain=50, with the same signal/hiss ratio, would require resistors near 0.02 Ohms, a dead short on the guitar and an unlikely load for any buffer. Gain structure is far more important than tone parts values.

BTW, a TL072's hiss voltage is similar to a 20K resistor. Yes, you can use a 1Meg bias resistor, but the bias resistor should never dominate a low-hiss input circuit. We'd normally use a 1Meg with a guitar of 50K-100K impedance. The TL072's 20K EIN is sufficiently lower than 50K-100K that it is generally "low hiss" in guitar inputs.