Tube Circuits

Tubes, Tone and Topologies <--







































































































































































































































































Tubes, Tone and Topologies

What makes amps sound unique? Why does a Marshall sound different from a Fender, which sounds different from a Vox, which sounds different from an Ampeg, and so forth?

Many musicians tend to zoom in on the tube complement of a particularly unique-sounding model and assume that's the source and secret of the correponding amp's 'personality'. While certain tube types do contribute an individualistic nuance, they are neither the only, nor indeed even the most important element in an amp's tonal signature.

Consider, for example, that many recent models from Fender, a company traditionally associated with 6L6 or 6V6 output tubes, utilize EL84s, which most players associate with Vox amps. Similarly, when Jim Marshall switched from the fat-bottle KT66 to the EL34, his amps still sounded like Marshalls.

So, if not the output tubes, what exactly is responsible for an amp's personality and tone?

To understand the origins of tone, we need to examine the overall circuit topologies that make our favorite iconic amps unique. Understanding these design differences will allow even non-technical musicians to look at a schematic and quickly get a sense of how that circuit will sound in an amp.

Since they are arguably the first production amps built, it makes sense to start with Fender. Of course, Fenders used various designs for different models, and went through many variations over time. For our current purposes, it will suffice to examine four basic topologies:

Fender Champ 5C1

The earliest Fender Champ models were about as simple as an amplifier can get, as shown in the following block diagram:

(click to enlarge)

This circuit uses only two tubes: A high gain small signal pentode driving a single power pentode in Class-A, single-ended configuration. In between is a volume control potentiometer. This design uses no global negative feedback (more on this later). Single-ended amps typically have a rich, nuanced tone with lots of harmonics. The single high-gain preamp tube will tend to be raucious when overdriven (as opposed to the very smooth overdriven tone of amps that use many smaller gain stages in series, such as some Mesa Boogies.) Having one tube responsible for all the preamp duties can also result in screechy microphonics when turned up high, particularly when using a small-signal pentode.

Fender Champ 5C1

(click to enlarge)

Although this evolution of the venerable Fender Champ still uses only two tubes, the small-signal pentode of the original model is replaced by a dual triode, which consists of two separate tubes in a single bottle. Having two stages share the amplification duties means that each one can be run at more conservative levels, allowing for a smoother response. Also, having the volume control between the two preamp stages allows any intrinsic tonal loss from that part of the circuit to be compensated for in the second stage. The output remains a single power pentode, however, a negative feedback loop has been added. This takes a small amount of the signal going out to the speaker and mixes it back into an earlier part of the circuit. The reason this is called negative feedback is because it is 180-degrees out of phase with the signal at the point where it is re-introduced. If these signals were identical in magnitude across the entire frequency spectrum, the result would be . . . dead silence! The signals would completely cancel each other out. This, however, is not the case. Without going into a lot of combersome mathematics, suffice it to say that the cancellation is only partial, and varies with frequency, with the ultimate result that the overall output is only marginally lower, but is much tighter and more punchy. (This is not so say 'better'—just different.) Feedback of this type is actually rather uncommon in single-ended and cathode biased amps (discussed later), but is by no means unique. In my own designs, I usually include negative feedback and a presence control wired to allow those circuit elements to be completely bypassed by the user.

Fender Princeton Reverb AA1164

(click to enlarge)

The early Princeton Reverb represented a step up in both power and complexity from the Champ. For our current purposes, I'm ignorring the reverb and the tremolo circuits, so as to more accurately compare 'apples to apples'.

Like the Champ, this design uses a two-stage preamp, but also adds a bass/treble 'tone stack' to the control complement, as shown, situated between the first preamp stage and the volume control.

The output section uses two power pentodes in what is known as a push-pull configuration. The effect of this, from a twenty-thousand foot level, is that each tube only worke half the time, with each one idling while the other is doing its work. This allows the tubes to create considerably more power than in single-ended use, since each one can 'rest' half the time.

Like the Champ, the Princeton's output stage is Cathode Biased. Cathode biased amps tend to have a more tactile, compressed tone than the alternative (fixed bias), whereas fixed bias amps tend to be more punchy and 'in your face'. In a nutshell, the Cathode Biased amp will continuously adjust its bias point in response to increased current flow. What this means is that the amp's idle bias can be considerably hotter than is the case in a fixed-bias design, since the overall power will tend to self-regulate. (That's an over-simplification of course, but should suffice to provide an intuitive sense of what's going on.)

Many players and builders (mistakenly or deliberately) confuse cathode biasing with 'Class-A.' Operating class is a far more technical topic than is appropriate here, but suffice it to say that the term is in no way an indicator of superiority. A lot of people seem to think the Vox AC-30, for example, is a Class-A circuit because it is cathode biased. It is not. It is a cathode-biased Class-A/B design.

Since one tube in a push-pull circuit amplifies only the peaks of the signal, and the other, the valleys, there has to be some way to split the incoming signal into two copies, each 180-degrees out of phase with respect to the other. This task is accomplished in a circuit known as a phase inverter.

The particular kind of phase inverter used in the Princeton is uses only one single triuode, and is known as a Cathodyne or Concertina phase inverter. This design is best suited to smaller power tubes, and can tend to sound unpleasant when overdriven (although careful design can ameriolate this to some extent). Nevertheless, with all else being equal, the an amp with a Cathodyne wil typically be designed for cleaner tones, where power tube overdrive is not desirable (e.g., bass amps.)

Fender Super Reverb

(click to enlarge)

Most of the larger Fender designs (Super Reverb, Twin Reverb, Showman, Bandmaster, Bassman, etc.) share a common topology, typified by that illustrated above.

Like the Princeton Reverb discussed earlier, they utilize a push-pull output section, with global negative feedback. Preamp differences are relatively minor, largely consisting of the addition of a 'Middle' control to the tone stack. The most significant difference is in the Phase Inverter, which uses two triodes (usually in the same physical bottle) to implement what is known as a long-tailed pair (LTP). This is a more foregiving design than the Cathodyne, and is better able to drive larger tubes into distortion. Also, these larger Fender amps usually employ Fixed Biasing, which explains their tighter, punchier sounds.


(click to enlarge)

Although Marshall made many different amps, when we think of Marshall amps we tend to envision the mighty, 100W Super Lead stacks of the late 1960s and 1970s, as employed by such giants as Jimi Hendrix, Eddie Van Halen, and most early metal artists.

Since Jim Marshall began by modifying Fender tweed Bassman amps, we shouldn't be too surprised to find many elements common to both Fenders and Marshalls. Indeed, close examination will reveal that the most obvious difference is the location of the tone stack.

Passive tone contols are inherently 'lossy' circuits. Even with all the tone controls maxed out, there will still be significant signal degradation. There are two ways to remedy this.

In Fender's case, the answer is to simply put the tone controls between two gain stages, and adjust the second stage to make up for any resultant losses. This design decision has significant impact on the overall tonality. In essence, the tone adjustments can only affect the signal that came before them, so when there are one or more preamp stages following them, they affect the nature of what goes into them, but not what comes out. This is particularly significant when those later stages are overdriven, in which case the tone controls more directly affect the quality of the overdrive, rather than the tone, per se.

The other alternative is to precede the tone controls with a circuit element that is better suited to driving a lossy (low impedance) load. One common element well suited for this purpose is the Cathode Follower, which is essentially a zero-gain buffer.

Marshall uses this latter approach. The two gain stages have only a volume control between them. Following the second preamp stage is a cathode follower, which drives the tone stack. From there, both the Fender and The Marshall circuits are essentially identical.

This topological design configuration is the most important contributor in making a Fender sound like a Fender, and a Marshall like a Marshall.

Vox AC-30

(click to enlarge)

The model most closely associated with 'the Vox sound' is the top-boost AC-30, most notably used by The Beatles and Queen's Brian May.

The most important aspects of this design, as shown above, is the use of a single, high-gain small-signal pentode, cojupled with a cathode-biased output section utilizing EL-84 tubes. Once again, this is a cathode-biased, class A/B circuit, not a class-A, as many believe.

Small-signal pentodes, like the EF-86 used here, have a more complex, nuanced tone than triodes, and can be more touch-sensitive when positioned at the beginning of the signal chain, as is the case in the Vox. Unfortunately, when the EF-86 is the only preamp stage, it must be run at very high gain levels, and this, in turn, can lead to unwanted microphonic noises. as well as (to my ears) a tendency toward shrillness.


(click to enlarge)

Although never quite attaining the stature of Fender, Marshall and Vox, Ampeg made some fine, unique-sounding amps, with some interesting features.

Their preamp section resembles a Marshall up to the second gain stage, but then, rather than using a Cathode Follower to buffer the tone controls, Ampeg employed a low-loss tone-stack design known as a Baxandall. Unlike the Marshall, Fender and Vox tone configurations, which all introduce a mid-range notch at all settings, the Baxandall's response is almost dead flat with both (bass and treble) controls centered. Since lowering bass and treble in this design effectively boosts the mids, a middle control can be ommitted. Some players find this difficult to get used to, and it is certainly possible (and beneficial) to add a middle control in various ways. The biggest benefit, however, is the low overall loss, which allows the control to be placed after the last gain stage without the necessity of a buffer.

Boutique Amps

(click to enlarge)

So-called 'boutique' amps, by their very nature, vary widely in design. For the purposes of this analysis I'll look at one of the more popular models: The Matchless Chieftain.

The Chieftain is highly regarded as a clean, chimey amp, with a tight overdrive, highly decoupled tone controls, and a rather Vox-like chime.

Its massive clean headroom results from the use of dual parallel triodes in the preamp gain stages, as shown above. I personally like this approach a lot, particularly in the initial gain stage, because if you start out making distortion, there's no way you can ever get a loud clean tone when you want it. On the other hand, with the wide availability of overdrive, distortion, and fuzz pedals to suit every taste, overdriving any front end is just a matter of stepping on your favorite effect.

the other distinguishing feature of this Matchless design is the complete separation of the bass, middle, and treble controls. On most other designs, the controls interact, making it more difficult to dial in precisely the tone you seek. The matchless design effectively decouples them.

The output section of the Cheiftain uses a pair of EL34s, in cathode-biased configuration, with no global feedback.

Overall amp, this amp impresses me as a very elegant piece of engineering.

The Wattwerx Topology

(click to enlarge)

The core topology used in my designs is shown above. A parallel triode in the first preamp section ensures a clean powerful signal to feed the subsequent stages. Using a small-signal pentode (EF86), configured for moderate gain, in the second position, has several desirable results. First, by keeping the gain down to reasonable levels, the circuit can be used in physical situations (particularly, combo amps) where microphonics would be a problem if configured for higher gain. This position also allows this tube to be overdriven by the earlier stage, thereby bringing out its complex, nuanced tone. This is simply not possible when put in the first position, unless outboard effects are employed. Finally, small-signal pentodes are well suited for driving low-impedance loads, such as tone stacks. Hence, it obviates the necessity of an extra buffer stage.

My tone stack is a variation of the Baxandall, but with the addition of a parallel middle control derived from Steve Bench's inductor-based tone stack.

The phase inverter is a basic LTP, driving a cathode based 2XEL34 output section. This power amp is unusual among cathode-biased push-pull designs in two ways:

1. Rather than used a single, hard-wired resistor and capacitor to bias both tubes, I use two discrete resistors and capacitors, combined with a high-precision balance control. This allows tubes to be balanced exactly, even as they begin to wear. This ensures the best, most consistent tone and power output for the life of the tubes.

2. I incorporate negative feedback and a presence control, with the option of switching the entire loop completely out of the circuit, for more traditional cathode-biased sound and feel.

This basic circuit is used exactly that used in the Little Dragon amp.

Wattwerx Variations

An added advantage of this topology is the numerous ways it can be expanded, increasing versatility without the overhead of adding additional tubes, stages, etc. For example, the following two variations were initially included in earlier Little Dragon Prptotypes, and were extremely functional. Unfortunately, due to the limited space in that chassis, I ultimately decided to leave them out because the additional wiring would have resulted in a maintenance nightmair. I intend to include them both in subsequent models where space constraints are less acute.

US/UK Switching

(click to enlarge)

The US/UK switch, whose functionality is illustrated above, allows one miniature toggle switch to alternate between US (e.g., Fender) and UK (e.g., Marshall) topologies. This switch completely changes the entire character of the amp. The resultant circuits corresponding to these two choices are:

US Configuration

(click to enlarge)

The 'US' configuration puts the small-signal pentode in the first preamp position, as shown above. This drives the tone stack, followed by the volume control, and finally, the parallel-triode preamp section.

UK Configuration

(click to enlarge)

This is the configuration hard-wired in the Little Dragon amp. Here, the parallel-triode stage comes first, followed by the volume control, the small-signal pentode, and finally, the tone stack.

Parallel Preamp Mode Switching

(click to enlarge)

This variation utilizes a DP3T (on-on-on) miniature toggle switch to select between three different configurations within the parallel-triode preamp stage. Used in conjunction with the US/UK switch, it yields the following three effective topologies:

Positon 1 (low gain)

(click to enlarge)

This position results in a single active triode, configured for low gain, with no bypass capacitor.

Positon 2 (Medium gain)

(click to enlarge)

This position results in a single active triode, configured for medium gain, with a bypass capacitor included in the circuit.


Positon 3 (High Gain/Maximum Headroom)

(click to enlarge)

This position results in a parallel triode configuration, which results in maximum gain and maximum headroom. This is the configuration hard-wired into the Little Dragon amp.

Maximum Flexibility/Minimum Components

Combining these two easy-to understand switches provides six different, distinct, useful amplifier configurations. It is used as the basis of my all amp designs where not precluded by chassis dimensions.