Tube Amps versus Solid State

Cafe Society
41

When it comes to guitar amps there are no absolutes, but a few things can be worked through.

Class B means, in the context of a guitar amplifier, a design that has a symmetric “push pull” output stage. This requires a different design for the output transformer as well, its input (primary) winding is two winding joined in series (aka a centre tapped winding). When the signal goes positive, one tube conducts current through one side of the output transformer’s primary winding, when the signal goes negative the other tube conducts current through the other side of the output transformer primary.

A Class A guitar amplifier has a single tube that is conducting half of its full forward current when there is zero signal. When the signal goes positive it conducts more, when the signal goes negative it conducts less. Again it controls the current passing through the output transformer primary. The downside is that even at idle the tube is conducting half full current, and thus runs much hotter than a class B.

Either way the output transformer is driven with varying current in the primary which is reflected in the output at a low enough impedance that is can drive the speaker.

This is all there is from a basic point of view. However when it comes to guitar amps we are usually looking to overdrive the output stage. At this point the difference between a symmetric and asymmetric output stages becomes important. To a good first approximation the symmetric stage will clip the waveform in a similar manner on the positive waveform and the negative. This is because both the positive and negative sides drive the tube to higher conduction as the signal rises in amplitude, so the final waveform is the combination of mirror images. A Class A output stage does not clip symmetrically. The transfer function of a tube looks quite different at the two extremes - driven fully versus cutting off. So the output waveform when driven hard is not symmetric.

The mathematics here contains the core of the sound. You can always decompose a waveform into two parts, the half wave symmetric and the half wave anti-symmetric parts. The Class B amplifier has an output that has very low half wave anti-symmetric components, whereas the Class A amplifier output has a mix of both.
Half wave symmetric waveforms contain no even harmonics. Half wave anti-symetric waveforms contain both odd and even harmonics.
The bottom line is that an overdriven class A amplifier will have an output with significant even harmonics, and an overdrive Class B will be predominantly odd order. Odd order harmonics generally sound harsher, as the higher order ones clash harmonically. Low order odd harmonics are however good. After all the third harmonic is the fifth up an octave. Even harmonics are more rounded in sound. The second harmonic is the octave, the fourth up two octaves, and the sixth is the fifth up two octaves.

But the preamp and general gain layout of the amp is going to dominate the sensitivity of the amp. Also there are other aspects of the output stage that govern its sound and response. The use of negative feedback is critical. There are huge difference here between different amp designs. So it is essentially impossible to generalise about the difference between A and B designs based only on that one aspect of the design.

42

Fuckin’ A; that’s a helluva post, Francis! Thanks!

43

Wow. Great post, thanks. My brain is clipping.

44

Add me to the Likes, Francis Vaughn - great stuff.

As you say, the bottom line is:

[QUOTE=FV]
The bottom line is that an overdriven class A amplifier will have an output with significant even harmonics, and an overdrive Class B will be predominantly odd order. Odd order harmonics generally sound harsher, as the higher order ones clash harmonically. Low order odd harmonics are however good. After all the third harmonic is the fifth up an octave. Even harmonics are more rounded in sound. The second harmonic is the octave, the fourth up two octaves, and the sixth is the fifth up two octaves.

[/QUOTE]

In many threads, I try to articulate how what is most important is what frequencies get dampened by that particular instrument/guitar + amp. Generated sound - by a vibrating reed, guitar or piano string, etc. - starts out with a broader set of frequencies - a lot of noise in the signal. What matters is when the body of an instrument and/or the guts of an amp dampen some of those frequencies, allowing the remaining ones to be boosted by whatever power stage is in your rig (could be your lungs blowing, could be the power tubes of an amp).

As **Francis Vaughn **summarizes, Class A amp is more likely to dampen the bad stuff - odd harmonics - and leave the good stuff - even harmonics - in place to be boosted. In its way, this is no different than a Stradivarius or a Prewar Martin, which for various reasons are great at dampening unpleasing harmonics and leaving/boosting pleasing ones (other instruments do this as well, but when people praise Strads and Prewars, they are praising this effect).

I suppose one thing that is interesting is the number of “dampening gates” an electric guitar rig has in it.

  • The body of a guitar dampens some frequencies. We can argue about electrics and the materials used in solidbodies and semi-hollows and hollows, but it is clear that the body of the instrument affects tone.
  • the Pickup - Gibson PAF’s, brass-plate-undergirded Tele bridge pickups, sealed EMG metal pickups - all are forms of dampeners first. Meaning, anything they ADD to the tone is really secondary to their role in the dampening chain.
  • The cord - longer cords and wireless rigs dampen tones. Hendrix liked curly cables because they cut highs.
  • The amp - per above.

Cool.

45

One thing no one has discussed is speaker break up. You have to remember that an electric guitar/amp combo is basically a distortion generator and to get that famous growl at high volume quite often the speaker is slightly overdriven as well which is why even with an attenuator or a load box the sound isn’t quite the same.
FV, great post, saves me some writing!!

46

Yeah, speaker breakup is a very complex question, and a big part of the sound.

And of course why there is so much choice in voicing speakers. Some of the issues are similar to amplifier distortion, in that half wave symmetric and antisymmetric components also form part of the analysis, but there is so much more going on as well.

Speakers have a who host of distortion mechanisms.

First up, the voice coil heats up and its resistance changes. This acts a bit like a slow acting compressor, can actually distort deep bass notes, and changes the frequency response of the system slightly. (Tube amps can have remarkably high output impedances* - up to 30 ohms, and so changes in speaker resistance can do odd things.)

Next, the current in the voice coil will modulate the magnetic field strength of the speaker. How much it can do this is partly dependant upon the magnet material, which is why magnet type can affect the sound.

But of course the big sonic component is when you get the cone to break up. And that is why we get so many different cone designs. Ribs allow the designer to control the breakup when it occurs, so a heavily ribbed cone tends to break-up in a much less brutal manner than one with fewer ribs, or no ribs at all. The thing about break-up is that it is very non-linear. Once a cone has started to break up, it will stay that way at levels lower than needed to start the breakup in the first place. Which is part of the manner cone design directly affects the feel of the entire system. More ribs tends to break up earlier as well. So a very smooth cone may have a very sharp and brutal feel and sound, whereas a heavily ribbed cone will be more smooth in its feel, and not so brutal in sound. And as always, there is no right answer.

I look at a large number of musical instruments as a system of exciters and harmonic generation mechanisms with additional frequency shaping. Which is essentially what WordMan is alluding to. Bowed instruments slam the harmonics into the system right at the start, as the catch and release of the bow on the string creates sharp sawtooth waves. After that it is up to the rest of the instrument to craft the sound with a complex mess of resonators with varying frequencies and dampening coefficients. Reeded instruments (clarinet, oboe, saxophone etc) also start with a harsh very harmonically rich source - the vibrating reed that is also very sawtooth in nature. Then the rest of the instrument sculpts the sound.

An electric guitar is rather special in that the principle harmonic generation component can be much later in the system. Run clean an amp is mostly just sculpting the signal from the pickups, and the mix of harmonics is a product of pickup location design, an the whole host of resonant elements of the guitar’s makeup. But once you overdrive the amp you have a very strong harmonic generator that is more akin to the sawtooth of a bowed or reeded instrument. Except that the waveform is clipped and closer to a square wave. So it has a very strong harmonic content, and lots of odd order harmonics. So the shaping of the signal both before and after the overdriven tube in the amp is critical to the sound. Indeed it is really the key determinant of the amp’s sound.Also triodes have a fundamentally different transfer function than pentodes which is a key part of why power amp distortion is different to, and prized, relative to preamp distortion. And even the output transformer matters. Indeed many highly regarded amps have quite cheap transformers, as their poorer frequency response and characteristics as the core becomes overloaded also form part of the sound. It is a glorious mess. :smiley:

  • A common misconception is that the output connections on a tube amp that are labelled with different impedances are setting the output impedance of the amp to “match” the speaker. They don’t. They are chosen to present the desired load line to the output tubes given a load that matches the label. The actual output impedance could be anything.
47

Oh man, such fine geekery. Thanks, Francis Vaughn.

Speakers - what you are saying about cone breakup is cool. There are so many speaker variations out there. I think of the various slits and poked holes and such that players added to their speakers, too. And please factor in Open backed vs Closed back cabinets! With the added spill out of sound from the rear, coming at a bit of a delay vs the main sound, it thickens things up vs. the percussiveness of a closed back cab.

Electrics - yeah, the line is that we really are playing the amp. The guitar is more of an amp-triggering device ;). It’s true, especially with a tube amp set at its sweet spot, where lighter playing is clean, but a sharper strum crunches up just so.

Transformers - and the fact that cheap ones add tone. That’s a theme throughout electric rigs. Older components are more inefficient, but that inefficiency is musical. Tubes. Old “bumblebee” capacitors and old-style pots. Curly cords that cut highs which is great for a Stratocaster with no flippin Tone control on the bridge pickup, etc.

One of the reasons it has taken decades to figure out Digital is that it’s too efficient. Waaaaay too much bandwidth let through. Our ears love good inefficiency.

48

Another excellent post, Francis. And some fine additional comments from WordMan as well! I haven’t had conversations like this in over 20 years; thanks everyone!

In the GOG Guitar thread I mentioned that I “play the sound” when I play guitar, by which I meant that all the different aspects of the rig, however big or small it may be, cause me to play differently. An acoustic guitar with no amplification gets picked, strummed and fingered differently than an electric with lots of distortion, for instance. Understanding the waveforms that are being created and how they can best be manipulated is a huge part of how I, personally, approach music and has been since I was a teenager.

In college I took a class on Electronic Music and it helped me understand and be able to verbalize (in some small, inefficient way) what I was hearing and why I liked it. And of course, it gave me some tools to be able to try my hand at it (well, until they changed the lock combination on the lab the next year, anyway). I can’t even estimate how many nights I spent in that lab, easily dozens, playing with various wave generators and manipulators, patching ¼" cables all over the place. Good times!

In later years, when I knew every band in town and worked with quite a few (doing audio, mostly) I enjoyed many hours shooting the shit about this kind of thing and learned even more from a bunch of very smart, very experienced guys and a whole gaggle of younger guys like myself who were eager to know everything there was to know about gear and how to use it and why.

This thread has been a fantastic read, with both new information and a bit of nostalgia for me.

Thanks folks!

49

Yeah, and that’s part of the reason when a guitarist tells me “direct boxes are fine”, I tell them, “ok, then you use it, and I’ll use your mic on my cab”. Nobody is willing to trade a cab for a direct box. If they use the direct out on my amp, it’s better because through the preamp I have some control of the amount of gain and e.q., but it doesn’t sound anywhere near as good as a mic’d cabinet. PA speakers aren’t made to break up and distort pleasantly, so it ends up being a comparatively thin, weak sound.