19 June 2012

Guitar Amp: How it works - Part 2

The Preamp section

We discussed in the PT section that, even that a transformer could be used to raise the original voltage up to 220V (or 120V), this helps nothing, since the available power (generated by coils) is just so ridiculous that we will have no current enough to move a speaker.

We already saw that we will need to pass from about 0.3V to 220V and, that means an increase of more than 733 times the original signal (a lot more when single vintage alike coils are being used).
This cannot be achieved in a single step so, there are two sections in the amp responsible for amplify the original signal: the preamp and the power amp.

The preamp increases the power of the original signal up to a level that the more powerful power amp section will raise up to the levels needed by the speaker.


The input jack

We can discuss for long the impact of guitar and patch cables, effects and everything that is between our guitar and the amp but, this is another history.
Imagine that we are directly plugged to the amp, with a high quality cable, without any loss of signal strength or tone. The first physical element that our signal will see is just the jack's plug.
This is just a connector, OK, but you are warned to maintain that connector always clean and shinny, to provide a good contact for our guitar cable plug (which should be maintained clean and shinny, as well).
This is where the signal starts its travel within the amp so, we need a clear an strong signal there.

Usually, between ground and that jack, we will find a 1MOhm resistor, which function is to avoid white noise when nothing is plugged in that jack.

Following component in the signal path will be a Bias Resistor and then, the signal will enter in the first amplification stage.


Preamplification Stage

We are talking here about Tube Amps but, changing some names, the way as amplification takes place is quite similar. While Solid State Amp use transistors (among other integrated circuits based on transistors), Tube Amps use vacuum tube valves for same task.

First diodes and triodes were made with tubes. Tubes made the design of the electronic gear really big and, dissipated great amount of heat and very prone to failure. When Guitar Amplifiers started to be built, the only option was the tube so, every guitar amp that participated in a record during the very first decades of '60s and '70s were based on tubes.
When the transistor was finally commercialized, everything in electronics moved from its tube designs to solid state designs. Benefits are clear: smaller devices, more reliable, low level of maintenance, way cheaper, etc.

The fact is that transistors seems to be more accurate when amplifying the original signal but, as always in music, tubes added its own distortion that created music in a more attractive way to the human ear.
Anyway, if the first amplifiers were made with transistors and then, used to record the great hits that every guitarist remembers from that age, maybe today we would find tube amps as noisy, distorted and, no so attractive.

Diodes and Triodes

Diodes were the concept that drove to triodes and, triodes are the basis of Tube and Transistor designs so, we will talk just a bit about both.

A Diode consists basically in two parts: a cathode and an anode, separated by a physical gap.
Cathode can be a filament (similar to those in electric bulbs) or some kind of plate (heated with a filament, named heater), highly resistive, that when a current travels it, heats at great temperature. When a certain temperature is reached, the cathode
starts to emit electrons that are being catch with the positive anode (usually named Plate).
The diode, also, conducts current in a single way, from cathode to anode.
With this simple electronic device we achieve some kind of electronic door, that closes or opens the circuit but, there is no control: or there is no current or there is all the current that the cathode can generate during its emission. White or black, no grey zones.

So, someone way smarter than myself thought: what about if we use some other thing to regulate the beam of electrons that reach the plate?. And then, the design of the Triode came.
The triode is just a diode with a third element called Grid. The grid is a conductor wrapped around an structure that surrounds the cathode or filament used to regulate the flow of current between the cathode and anode, like a faucet valve.
To achieve this, the grid takes values between, practically the same as the cathode to zero (or even positive in some cases). When the grid is more negative than the cathode, the electrons beam is rejected against the cathode and, the flow to plate is stopped.
When grid is less negative than the cathode, some of the electrons can reach the plate. When the grid has zero value, all the beam can reach the plate.

In the case of tube valves, the parts of a triode usually have the names we talked above (cathode or filament, plate and grid or control grid).
In the case of transistors, the cathode is named Emitter, the anode Collector and the grid Base but, its principle of operation is quite well the same.


Tetraodes and Pentodes

Tube valves have some issues. We know that when we have two separated surfaces polarized we have a condenser or capacitor. So, there is some capacitance effect in tubes. Remember that a capacitor is part of a passive filter that catches some range of frequencies from the signal so, there is a potential loose of some of the frequential content of the original signal.

To control some of this issues, the Tetraode was designed. Tetraode is just a triode with one more element, a second control grid that solves some of the issues inherent to tube triodes.

But, man, it creates new issues and, therefore, Pentode was designed. Pentode is just a Tetraode with one more element of control, one more grid. This design seems to solve the most of issues related to tube amplification.

Despite of very few cases, the most of tubes that make a preamp section are all triodes and, despite of very few cases, the most of tubes that make a power amp section are all pentodes.

A Pentode can be working in triode mode, just deactivating the two additional control grids. This lowers tube's power but, creates a warmer sound. You will see a lot of amps with a switch to change operation mode of power amp from Pentode to Triode. Now, you know what is happening there.


Preamp tubes

Usually, tubes used for preamp section in Guitar Amplifiers are part of the family 12AX7. There are some special cases, for sure but, we will focus in the spreadest one. We will talk deeper about tubes in other tube related articles. For this article, we just need a pinch of information.

Preamp tubes have a pair of triodes.
We know that each triode is an amplification stage so, a tube of the family 12AX7 has two independent amplification stages within the same flash.

Preamp tubes are usually named V1 (for Valve number 1), V2, etc., when looking to some amp's diagram but, you will usually see a letter at the end of each tube name, by example: V1A and V1B. This means, triode A of tube 1 and triode B of tube 1, by example.
That's why, the same tube can have two different functions, because of the independence of its pair of triodes. By example, tube 1 can have triode A as the first gain stage and triode B as the driver for the reverberation can.

Family 12AX7 includes several tube models with different characteristics. Pure 12AX7 model has a gain factor of 100, while a 5751 tube has a gain factor of 70, a 12AX7-WA a factor of 60, etc.
What means to have a gain factor of X?

You know, the coil is generating the signal in an alternate way, changing the phase from positive to negative and then, generating voltage and current values of different levels.
Imagine that, when the coil is quiet, voltage is zero and than, when it reaches its higher power, the signal takes a value of +/- 0.3V. The range of the original signal can then be between -0.3V and +0.3V, that means an absolute range of 0.6V.

If the tube has a factor of 100, that means that the range of voltage variance at its output divided by the range of voltage variance at its input will be equal to 100.
So: output range / input range = 100.
We know the input range (0.6V) so, the output range will be input range * 100, that is 60V.
That means that a difference of 0.6V in the input of the tube will generate a difference of 60V in the output of the tube. If we set the plate voltage to 100 (in idle state), the output of the tube will vary between 70V (100 - 60/2) and 130V (100 + 60/2).
So, an small variation of voltage (range 0.6V) creates a higher variation of voltage (60V) and therefore, there is a power amplification of the signal.

But, we saw that the speaker works at 220V so, we need a range of 440V to be able to fully drive the speaker (from -220V to +220V). This final step will be done by the Power amp section, later.

Some amps are using just one preamp triode to amplify the original signal, sending it to the power amp for final amplification. Those usually have the purest tone and are very kind with the ear.

Other amps, use more sophisticated preamp sections. By example, Hi Gain amps (like Mesa Boogie's Rectifiers or Soldano SLO 100, etc), are using the output of a triode as input to a second triode then, the output of the second triode into a third triode, etc. This way to stack triodes to amplify the original signal is named cascaded gain.
In those kind of amps, you will see several switches that voice the channel in different way. Most of those switches are just selecting some triodes to add to the cascaded chain.

The preamp section usually have what is named as a tone stack (because how it looks in amp design diagrams). Tone stack can be formed by a single Tone control or more (Treble, Middle, Bass). Sometimes, the signal, after crossing that tone stack becomes significantly lower and therefore, a triode of a preamp tube can be used to recover the signal level after the tone stack.

One triode can be used to drive the reverberation can and, one more triode used to recover the weak signal coming back from the reverberation can and mix it with the signal amplified coming from the tone stack.
One triode can be used to make stronger the signal sent to the FX loop and, one more to recover the signal that comes back from such a loop.
In fact, depending on the amp's design there are several tasks that can be delegated to some tubes or triodes of different tubes.

The last tube of the preamp section is usually the Driver or Phase Inverter (PI). That tube feeds the signal to the power amp section, in a way that depends on the operation class of the amp (usually, class A or class A/B, in guitar amps).

Usually, Gain controls are potentiometers that are situated between a couple of gain triodes to regulate the amount of signal of triode 1 that is being amplified by triode 2. Sometimes, they can be called differently but, when you see an amp with volume and gain controls, gain is usually related to preamp section, while volume is related to power amp section.
Volume is usually a potentiometer just before the Driver or PI to control the amount of the preamp signal that is being amplified by the power amp section.
Master Volume is usually a potentiometer after power amp that regulates the amount of signal that reaches the speaker.

As a curiosity, every time that the signal travels across a triode, its electrical phase is swapped (so, from positive to negative, and viceversa). To obtain same phase as in the original signal, we need an even number of triodes.

Gain, Distortion, Microphonics and Noise

To increase the gain of a signal is similar to use a zoom tool in a PC Monitor.
Imagine that we have a very small typo N on the center of our PC's Monitor and we want to see it in a bigger detail. Imagine also that we have an infinite zooming tool so, we can make that N as big as we wanted.
While we are zooming in, as far as we can see the complete shape of that N, we are achieving amplification without distortion but, there will be a point where the size of that typo exceeds the available space on the monitor so, we start to loose the upper and lower strokes of that N. Reached this point, if we continue zooming in, we will have difficulty to distinguish a typo N, maybe all what we will see is just the backslash (\) stroke of that N in our screen.
So we can have a high fidelity amplified image of the original signal amplifying it up to certain level, further than that, we start to have a distorted amplified image.

We have other ways to obtain a distorted amplified image. Following the example of that N. Imagine that we took a picture of that typo, originally very small. If we start to zoom in the picture, there will be a point were the image starts to pixelate. If that typo was painted with several textures of color, the excessive zoomed picture will be even more distorted.
Now, imagine that our monitor has an issue and the image flicks, like on the old TV sets, then the strokes of that typo will be twisted. Or imagine that we have white noise in our Monitor, then the strokes will appear washed.
You can imagine probably a bunch of different ways to distort the original image by yourself so, you will realize that is easier to achieve a distorted image but, quite well impossible to obtain a High Fidelity image of the original signal.

Well, that's what happens inside the amp when we are amplifying the original guitar' signal across several amplifying stages. Tubes (also Transistors but in a different way) add several types of distortion to the original signal. Rest of components, including the Speaker, will add their own distortions.
While in Hi-Fi, we are trying to minimize at the lowest level the distortion of the amplified signal, in Guitar Amps there are some kind of distortions that are considered as unwanted, while other type of distortions are a wanted behaviour.

Pre-amp tubes introduce some harmonics of certain order, while power amp tubes introduce harmonic of another order so, the inherent characteristics of their distortion is different and therefore, results in sound are different.
Vintage amps were trying to keep the preamp stage as Hi-Fi as possible, minimizing the distortion of the preamp tubes. The rich distortion of the power amps (and speakers) added just that touch and feel to the sound that we all love.
Hi-Gain amps are cascading preamp stages so, they can achieve a big amount of distortion before it is finally amplified in the power amp stage. The sound is a thick, massive wall of distorted sound and, in some way cooler and edgier than the distortion achieved on power amp.

As we already mentioned, sometimes we have one (or more) gain control that allow us the decrease the signal level (already amplified and eventually distorted) that outputs from triode A as an input for triode B. In that way, we can regulate the amount of distortion in the several cascaded gain stages.

Imagine that every triode has a gain factor of 10. Running the signal across 2 triodes, we will have an increase in gain of 100 (10 * 10). Across 3 triodes, the increase will be about 1000 (10 * 10 * 10). The individual gain factor of each stage is properly controlled by amp's designer to achieve its signature sound.

Process of amplifying has side effects, as well. Our guitar's signal has some floor noise. The cable itself can add (catch) some extra noise. Every electronic device in the path will introduce its own noise and, any pedal effect that operates on signal gain (EQ, Compressor, Overdrive, Booster, Distortion, Fuzz, etc), will increase the original signal, including the original floor noise.
So, by amplifying the signal we are also amplifying the loudness of the floor noise. If the difference between the level of the signal and the level of the floor noise is high then, we will be able to a higher amplification range with minimal impact in noise but, if the signal level is way closer to the floor noise level, we will have a nice amount of noise as soon as we start to amplify guitar' signal.

Bad news is that Tubes introduce their own operating noise that is being added to the floor noise already amplified. The most critical valve in a Guitar Amplifier is the one having the first amplifier stage (triode), where guitar' signal comes first, to start amplification.
For this position is prioritary to have a nice tube, with very low operating noise and a high quality representation of the original signal. You can understand that any error introduced in this first amplification stage will be propagated and amplified on following stages.

Due to mechanical degradation of the plates and filaments of the tube, due to high stress operation, tubes can have additional noises, as pops and cracks that occurs sporadically. For the first stage we want a tube with the lowest presence of those.

And, what usually happens when an electronic device has some metallic parts that move, the tube can be prone to microphonics issues. Once more, we want a tube in the first stage with as low microphonics as possible.

Tubes need a certain time to be operating at their best. Usually, after half to one hour, tubes are delivering their best sound (if in a good condition).
What we are usually looking in tubes is that sweet spot where the signal begins to be slightly distorted but it's fully recognisable, where that sweet spot is found depends on every valve (and amp's design) and it's commonly called tube's break up. At this point, we will have some harmonics generated by the tube itself added to our signal and a very light distortion, very useful for Blues works, by example. This sweet spot is usually found around the 70% to 80% of the total gain of the tube (imagine, dial numbers 7-8 in a gain or volume control with10 numbers).

Most of this issues disappear with Solid State amps so, why Tube Amps are the way to go?.
As we said, Solid State amps have a more hi-fi image of the sound and, the way as they distort the signal isn't so pleasant to the ear as tubes do it. In short, tubes generate a sound that is warmer, musicaler and just right to the ear. There are really good Solid State amps anyway, and they have their own use (mainly for clean, jazzy works). Magic of tubes is in the fact that they are imperfect and even unpredictable.

So, preamp section is where our sound is originally forged and, has a great impact on our final tone. To swap tube models in our preamp section will have a clear impact on how the amp sounds and feels but, up to a certain point. In the same way that you cannot make a Les Paul to sound as a Stratocaster (or viceversa), there is no way that a complete tube swap can change the foundational tone of our amp. If it was voiced vintage alike, it will continue to sound vintage alike. If it was voiced to be a hi-gain monster, it will continue to sound as a hi-gain monster. But, tube swap can add, subtract or modify certain textures, harmonics and distortion nature and grain so, it worth the time to experimenting with them.
If we needed a greater clean headroom in our amp, what we need to do is to lower the gain factor of our first preamp tube (move down from 12AX7 to 5751 or 12AY7 or even 12AU7) will decrease the possibility that the signal starts to distort while being amplified.
If we needed an earlier break-up point in our power section, we can choose another tube for the PI or Driver with that earlier break-up spot.

2 comments:

  1. this is really interesting. I have been playing a long time and really didnt know anything about my amp. now I do! great post keep up the hard work. check these out IStillGotMyGuitar

    ReplyDelete
    Replies
    1. Thanks for your kind words!.
      Stay tuned, I planned to post two more parts.

      Delete

Please, feel free to add your comments.