Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 04:25:40 PM

Login with username, password and session length

Recent Posts

 

Basic circuit paramaters

Started by floodpud, September 24, 2012, 11:38:55 AM

Previous topic - Next topic

floodpud

Hi!
I have been creating a few effects (mostly distortion) from schematics and decided I would like to build my own circuit from scratch. I am currently in an electrical engineering program so i understand how transistors work, and I can design a basic amplifier (common emitter). Where I am having problems is with modifying the circuit for specific results. Here are some of my questions:

What are the parameters that I need to be designing my circuit around?
What effect does changing those parameters have?
Why use negative feedback vs. reducing input signal?
Am I looking for voltage gain or current gain or both?
What does input and output impedance mean? (I understand the concept of impedance but how does it apply to amplifiers?)
How can I calculate frequency response of a low or highpass filter? (again I am familiar with f=1/2piXC but how do I apply it?)

Thats a lot of ground to cover but any help would be appreciated.
Some of my questions may be odd because my knowledge is geared towards power & utilities but all answers are welcomed.

J M Fahey

Go step by step:
Start by
1) analyzing classic commercial designs, such as a MXR Dist+ and
2) understanding, not only "how" it works, but
3) what happens if you modify some value and then
4) "what was the designer trying to do?" and then
5) what if *I* wanted something different? (more/less gain/bass/treble/distortion/noise/battery consumption/cost/etc.)

Apply this to many and after a few, you'll start looking at things "like a designer".

Roly

{Grorp. :o }  Do what I did, finish Power and do Electronics.

Quote
>What are the parameters that I need to be designing my circuit around?
>What effect does changing those parameters have?

Too general.  As @JM said, start with building a simple known design, then futz with it changing component values and observing the effect.

In a straight amplifier stage we generally first try and set the DC bias conditions or Q-point so that the transistor collector or output node rests at about half the supply voltage so as to obtain the greatest possible voltage swing up and down before clipping occurs.  This may not be the case however in stomps/Fx.

Quote
>Why use negative feedback vs. reducing input signal?

Because NFB doesn't just reduce gain, that's incidental.  NFB has a number of effects, it reduces distortion, increases bandwidth, and reduces output impedance.  In guitar amps NFB is generally only applied around the power amplifier stages, not the preamp, and in valve/tube amps it may not even be applied there.  An exception is when op-amps are used when NFB is almost always applied locally around the op-amp.

http://en.wikipedia.org/wiki/Negative_feedback#Electronic_amplifiers
http://en.wikipedia.org/wiki/Negative_feedback_amplifier
http://en.wikipedia.org/wiki/Nyquist_stability_criterion

Quote
>Am I looking for voltage gain or current gain or both?

That depends on what you are trying to do.  In preamp stages (and stomps) you are normally looking for voltage gain.  In the main or power amplifier you are generally looking for both because you need more voltage than the typical 1 volt that comes out of your preamp, and you need current to drive the low impedance loudspeaker, and of course the product of voltage and current is power, P = E * I.

Quote
>What does input and output impedance mean? (I understand the concept of impedance but how does it apply to amplifiers?)

In audio amplifiers you can generally simplify and approximate impedance (a complex number, R+jX) as simple resistance.

The input impedance (or resistance) of an amplifier is a measure of the loading it presents to its source.  In the case of passive guitars the source (or output) impedance is moderately high and therefore easily loaded, so it is desirable that the guitar amp input has a high input impedance, generally at least ten times the source impedance.  In practical terms this means an input impedance of at least 500k ohms and better 1 megohm.  In the case of acoustic guitars with (naked, i.e. no inbuilt preamp) piezo pickups the source impedance is considerably higher so an input impedance of around 2.7 megs, or even higher, is the norm.

Conversely when you want to feed a signal out of your preamp it is desirable to have a low impedance (or source resistance) so that the output won't be excessively loaded by the following power amplifier input impedance, say typically 47k but could be much lower, say only 10k.

Where a long cable has to be driven the stray shunt capacitance of the cable is important, and a low impedance drive will overcome high frequency losses due to this capacitance (typical guitar cable capacitance will normally be in the range 500pF to 2nF).

In the case of a power amplifier a very low output impedance is desirable because we want to develop power in the load (typically 8 ohms).  The output impedance of a solid state power amplifier is typically a fraction of an ohm so most of the power developed by the output current is in the load, not the amplifier (unless of course you short-circuit it then most of the power is developed in the output transistors for a few hundred milliseconds before the magic smoke gets out).  The output impedance of a valve/tube amplifier is  generally much higher, in the order of ohms.

Quote
>How can I calculate frequency response of a low or highpass filter? (again I am familiar with f=1/2piXC but how do I apply it?)

The combination of a series resistance followed by a shunt capacitance forms a single pole low pass (LP) network.  The -3dB or "corner" frequency occurs when the reactance Xc of the shunt capacitance is equal to the source resistance.  Past this frequency the response falls by -6dB per octave or -20dB per decade.

The converse is also true, a series capacitance and shunt resistance forms a high pass (HP) network and the -3dB point is also when the capacitive reactance is equal to the load resistance.

If these L-sections are cascaded they become 2, 3 or more pole LP (or "zero" HP) networks and the loss at the "corner" frequency goes up by -3dB per section, -6, -9, and so on, and the ultimate rolloff also goes up by -6dB/octave or -20dB/decade to -12, -18dB/oct or -40, -60dB/dec.  It is uncommon to encounter networks in guitar amps that are more than one section.

Application.

If we assume a preamp output with a source impedance of 1k driving a guitar cable of 1nF, what will be the corner frequency?

By definition the corner will be when Xc = Rs, or 1k ohm.

Xc = 1 / (2 * Pi * f * C) ;reactance Xc of a capacitor C at frequency f, in ohms

Multiply BS by f;

f * Xc = 1 / (2 * Pi * C)

Divide BS by Xc;

f = 1 / (2 * Pi * C * Xc) ; Hertz, Farads, Ohms

f = 1 / (2 * Pi * 1*10^-9 * 1*10^3) = 159154.9430919

...or 160kHz, well above the audio range and thus presenting no problem.

But if the source impedance were 1 megohm we would get;

f = 1 / (2 * Pi * 1*10^-9 * 1*10^6) = 159.1549430 or 160Hz

...which would only be useful for amplifying a kick drum (if that).

See also;

http://www.ozvalveamps.org/maths.htm#Filters

Two very useful tools are Duncan's Tone Stack Calculator, and one of the Spice variants such as LTSpice which greatly ease the calculations, however it is vitally important that you develop an intuitive feel for what responses are expected from given component values to avoid massive decimal point errors and such.  This is best done, as JM says, by analyzing existing circuits, and at the bench trying actual components and observing the results.

You will find a wealth of amp and stomp circuits here;

http://www.freeinfosociety.com/electronics/schempage.php?cat=1

...but you must always remain aware that no circuit is ever fully complete, nor entirely error-free.

Note also that I have spoken in general, and that there are many odd arrangements you will encounter, but the basic theory always applies - nobody has yet broken Ohm's Law.  ;)

HTH

If you say theory and practice don't agree you haven't applied enough theory.

polo16mi