10

Do we really need tube amp for power amp? I understand that the tone change is done in the preamp. And preamp with tube sounds a lot better than solid state amp as it's more rich and full in tone. However, as solid state amp can easier generate enough power for the amp for the output, if it's just for driving power for the sound output, how important to have the amps with tubes in both preamp and power amp? There are a few options for having tubes in preamp and solid state for power amp. Do they perform good as well?

6
  • 1
    The answers will likely be very opinionated, and it all depends on the sounds you're trying to create (especially overdrive tone) – PeteCon Jan 1 at 16:16
  • 1
    I don't think this warrants a full answer, but as a Front-of-House Audio Engineer, I can't stress enough the benefits of having a quiet stage, even for a loud concert. It sounds a ton better to the audience if I don't have to mix around a bunch of stage spill that doesn't match my PA's dispersion pattern. And your Monitor Engineer will love you too, if you have one, because he doesn't have to make me mad for the same reason, just for you to hear someone else. (for smaller gigs, the FOH and ME are the same person on the same console at the FOH position) – AaronD Jan 2 at 8:28
  • That said though, I also want to capture every bit of the guitarists' well-crafted sound, and part of that does indeed come from a tube power stage running flat out or perhaps slightly more. I've seen several solutions to get both "the complete sound" into the PA, and a quiet stage, and my favorite is to replace the speaker with a dummy resistor load and a speaker-level DI box. Then the amp can drive hard into its rated load, and I get that signal silently into my gear, within the variation that mic choice and placement would have anyway on a live cab. – AaronD Jan 2 at 8:28
  • 1
    Just don't ever forget to always have SOME kind of a load on a tube amp! You can break it by running it without anything. The reason can be a bit unintuitive, but as long as you never have it turned on without a load of some kind - speaker, power resistor, whatever - you'll be fine. – AaronD Jan 2 at 8:37
  • 1
    The reason has to do with the output transformer that most of them have. When it's loaded correctly, the load current on the secondary winding counteracts the primary winding, so as to reduce the internal magnetic field. When it's unloaded, the primary current is not counteracted, and so the magnetic field goes way up. When it gets too high, the iron core essentially clips, and any input beyond that effectively sees a short circuit. Now the amp itself is getting shorted at high power...because you removed the external load! (I told you it was unintuitive, but it's true) – AaronD Jan 2 at 8:38
13

It's an oversimplification to claim that "the tone change is done in the preamp". It's the complete signal chain that results in the final sound, including the power amp and the speaker cabinet. Note that it's not only the tubes that make the difference between a solid state and a tube power amp, but also the output transformer, which is necessary for tube amps but not for solid state amps. The output transformer's response is frequency dependent and non-linear, so it definitely influences the final sound coming out of the speakers. Of course, the influence of a tube power amp and of the output transformer becomes more pronounced at stage volumes, not so much in your bedroom.

Concerning the choice of tubes or solid state circuits in the pre-amp and the power-amp stages, I would even claim that nowadays many setups are the opposite of your suggestion: a solid state (mainly modelling) pre-amp driving a tube power amp, or driving a clean tube amp, including a pre-amp, but without pre-amp overdrive.

I more often encounter simple solid state overdrive pedals driving a good tube amp, than much more expensive tube-preamps driving a solid state (power) amp.

After all it's about taste and the style of music that one plays. It's important to try several combinations and to be open to everything instead of relying on what one thinks should be the "right" combination.

6
  • 1
    Note that it would be perfectly possible to build a solid-state amp that uses an output transformer. It would just be rather pointless. (And, the transformer is actually much more linear than both the tubes themselves and the speaker.) About transistor-based distortion etc. pedals – yes, they provide most of the gain&clipping in many guitarists' sounds, however that doesn't mean they don't make use of some tube overdrive. The key is often to get some broadband harmonic spectrum from harsh transistor clipping, the smoothen it up again with some filtering & softer tube clipping. – leftaroundabout Jan 1 at 14:49
  • @leftaroundabout Transistor amps with output transformers were extremely common in the 1960's. Part of the reason was that the engineers were used to that design from making tube amps. It's not entirely pointless! Suppose the output transistors are running on, say +25/-25 power rails. That limits the output voltage and ultimately power to nominally around 40W RMS into an 8 Ohm load. A transformer could step that voltage up to drive more power: say double the voltage for theoretically quadruple the power. – Kaz Jan 3 at 9:33
  • @Kaz: Center-tapped output transformers could also make it easier to design symmetric push-pull output stages using two transistors rather than four, and can help protect amplifiers from a variety of things that could otherwise go wrong. – supercat Jan 5 at 23:33
  • @supercat Ah! You jogged my memory about something. Vacuum tubes do not come in complementary pair devices. The push and pull circuits use a phase inverter, and the same polarity. The correctly assembled signal comes out at the output due the way the two halves of the stage are wired to the transformer. And, so, that's exactly what some transistor amps were doing in the early days. Those designs didn't use matched PNP/NPN complementary devices, but the same devices, with inverted phase, so "push" and "pull" were the same direction voltage swing. The output transformer was then essential. – Kaz Jan 5 at 23:58
  • @Kaz: Yup. When power transistors first came on the scene, they were relatively expensive compared to passive devices like transformers, so using two transistors and a transformer would often have been cheaper than using four transistors. – supercat Jan 6 at 0:02
6

Tube power stages only have a notable effect when driven at high signal levels, so they don't just linearly amplify the signal but soft-clip it somewhat. This is in effect also what high preamp-gain accomplishes, but it is controlled with the master volume pot – which really works the same way as the gain pot, except it's placed after the preamp(s) and right before the power stage. That means it's only possible to overdrive a power stage by also taking the amp to its loudness limits. (Though it is possible to “burn” some of the energy with resistors, to avoid the loudness going too crazy.)

Many guitarists insist that power stage overdrive sounds inherently different/better compared to mere preamp overdrive. I'm not completely convinced about this, but what's for sure is that the standard preamp designs clip in a different way from the standard power stage designs. There are mainly two differences:

  • Power stages tend to be built in a “push-pull” configuration, which means both polarities of the signal clip in a symmetric manner. As a result, the overtones introduced by the overdrive are all odd-numbered harmonics, whereas assymmetric preamp overdrive excites both even and odd harmonics. I find it a bit dubious that this would make a big difference, because if you send in even harmonics into the power stage, even harmonics will also come out, even though the overdrive was symmetric.
  • Power stages sit after all tone-shaping circuit elements. Therefore, the waveform will not get warped anymore before it reaches the speaker (except a little by the output transformer).
    More importantly (but note this depends on the concrete amp design), if the tone shaping sits between the preamp gain stage and the power amp, this will affect how much each of the partials are driven into the power overdrive but not into the preamp overdrive. So having both pre- and post gain controls gives you more flexibility about how the overdrive reacts. However, this phenomenon doesn't really require the second overdrive to be a power stage – you could as well achieve this by cascading two preamps with an EQ between them.
2
  • 'Power stages sit behind' - sounds a little ambiguous. Would after be better? – Tim Jan 1 at 10:22
  • I missed that one. Meant the one in the final para. – Tim Jan 1 at 15:07
3

A solid-state amp has some decent features that are worth considering when buying an amp. For high-gain sounds I have found that solid-state makes the controlling of the gain more precise, a solid-state amp just responds to gain in a more predictable manner.

For a long time it was rather hard to get truly hard to get high-gain sounds out of tubes. I don't know if that is the case anymore though. I do remember Dimebag Darrel playing solid-state amps for many years because he was not happy with tube's gain sound.

I just think the stigma that surrounds solid-state amps is not entirely warranted by the technology. Solid-state amps where originally designed to be affordable alternative to tube amps. They where also designed to be much more practical and have less parts that can break.

This made it so that for many years high-end solid-state amps sounded like an oxymoron, but really nowadays there really is no reason to choose anyone over the other, even modelling amps have become really great.

I can honestly say that the Marshall code 25 amazes me every time I play it. For the breadth of tones at the price, I don't see you really getting anything better.

So in the old days there may have been issues regarding quality, but these days like most things in guitar tone, it just boils down to personal preference.

3
  • 5
    Please stop using “gain” as a synonym for overdrive/clipping/distortion. They're not the same; you can in principle have lots of gain without any nonlinearity whatsoever. – leftaroundabout Jan 1 at 15:07
  • @lefaroundabout In a guitar setting, it's perfectly clear (or not any less unclear than usual). You're asking dogs not to bark. :) No guitarists says "I need more gain" meaning more volume for clean jazz chords. – Kaz Jan 3 at 9:41
  • > For a long time it was rather hard to get truly hard to get high-gain sounds out of tubes. That must have been before Deep Purple, Black Sabbath, Hendrix, ... the problem of getting distortion out of a tube amp was solved when Dimebag Darrel was climbing monkey bars in a schoolyard. – Kaz Jan 3 at 9:43
3

Although there is nothing "magical" about vacuum tubes that couldn't be emulated with a transistor-based circuit in a power amp, most transistor-based amps are designed to behave as thought they have a source impedance that is very near zero ohms. Thus will cause them to put half as much power into an 8 ohm load as they would put into a four ohm load. Many tube amplifier designs, by contrast, will behave as though they have a significant output impedance. They use a transformer to reduce the output impedance enough to usefully drive 8-ohm or (depending upon the transformer and the windings chosen, 4, 16, or 32 ohm) speakers, but the impedance will still typically be higher than that of a solid state amp. Some tube amps had an output impedance that was on the order of hundreds or thousands of ohms, but tube amps can use feedback to substantially reduce output impedance. I have no idea to what extent different kinds of tube amps did so, however(*).

Because the impedance of a speaker can vary in ways that depend upon the drive frequency and the speaker's position relative to things like walls or corners, the amount of power a speaker will draw from an amp which is fed a constant signal will vary as a result of such factors in ways that depend upon the effective output impedance of the amp. For many purposes, it is considered desirable to minimize output impedance, but doing so is easier with solid-state designs than with tube-based ones. Thus, many transistor designs have an effective output impedance that is lower than would be typical of tube-based amps. One wouldn't need to use vacuum tubes to recreate such effects (using a more powerful amp than one would otherwise need and putting a resistor in series with the load could have a similar effect), but such effects would represent a difference between common tube amps and common transistor amps.

(*)A power amplifier using bipolar transistors in an emitter-follower configuration can very easily achieve a very low output impedance with linear response, and will be inherently stable. It's possible for a tube amplifier to use voltage feedback to counteract the inherently-high output impedance of the tubes, but it requires precisely-balanced compensation circuits. Applying two little compensation will leave the amplifier with some residual output impedance. Applying too much will cause the amplifier to oscillate or otherwise malfunction.

8
  • Interesting point that I had not considered. But it would be nice to quantify how much influence the impedance really has. — If it's actually a significant effect, then one consequence would be that resistive power attenuators are actually not neutral to the power stage behaviour. – leftaroundabout Jan 2 at 0:29
  • Correction: a voltage amplifier (exemplified by a good solid-state) will put half as much power into 8 ohms as into 4 for the same signal at the same amplification level. A tube amplifier will still put less into 8 ohms, but it could have a separate tap on its output transformer to compensate for that. Current drive can be obtained in a transistor amp using a tiny resistor in series with the output, which used to obtain a feedback signal to mix with or supplement the voltage feedback. – Kaz Jan 5 at 16:26
  • Some solid state guitar amps use current feedback it. On a Marshall 8008, it is switchable. Early Roland JC-120's had current feedback: by 1981-ish, it was gone! (I compared the schematics at one point). I experimented for years with a current feedback amp mod. I don't run that amp in current mode any more; I just EQ it. I keep thinking about undoing the mod, but never get around to it. – Kaz Jan 5 at 16:29
  • @Kaz: You're right about power being inversely proportional to impedance. Mea culpa. While tube amps do indeed generally have separate transformer taps for different speaker impedance, all frequencies will generally have to go through the same transformer taps. – supercat Jan 5 at 17:02
  • @leftaroundabout: From a random google search, education.lenardaudio.com/en/05_speakers_3.html would suggest that speaker impedance would increase by a factor of about 3-6 fold near the resonance peak, and also start to increase again past a certain point, so a current-based output would drive more power at those frequencies than a voltage-based one. – supercat Jan 5 at 17:36
2

Do we really need tube amp for power amp? I understand that the tone change is done in the preamp.

The situation is nuanced. For instance, the late Eddie Van Halen's legendary sound is almost automatically associated with tubes, and nothing but tubes, right? But, when playing live, he used a technique of of capturing the output of an entire tube amplifier (power section and all), putting that through some processing, like delays, and feeding the output into several powerful solid-state amps, such as the HH Electronic V800: a 1979-dated solid-state amplifier that can drive 800W into 8 Ohms, when bridged to mono, or else 250W per channel in stereo. Still, though making use of solid state power amplifiers, he did use the entire tube amplifier for creating tone.

The mistake people make is thinking that a solid state amplifier rated at 50W or 100W will "do" the same kinds of things as a tube amplifier with the same numbers. Without getting too much into details, the nutshell of it is that

  1. You don't want to drive a solid-state power amplifier into clipping, so you need more headroom; and

  2. Tube amplifiers often have multiple outputs for speakers of different impedances (like 4, 8, 16 Ohms) so they can drive those different speakers with the same power. This is because they are transformer-coupled: there is a transformer between the output tubes and the speaker, and output transformers can have multiple taps for different numbers of output windings. Solid-state amplifiers do not have this feature: they drive less power into higher impedance speakers. A solid-state amplifier rated at 100W into 4 Ohms will not put out 100W into 8 Ohms: it could be 75 or as little as 50. A tube amplifier rated at 100W into 4 Ohms can still put it into 8 Ohms if it has dedicated output jack for 4 Ohms and 8 Ohms.

Both of these points indicate the need for getting a decent amount of wattage in a solid state amplifier.

Another mistake is buying a solid-state guitar amplifier from a company/brand that is famous for tube amplifiers. Such gear is always going to be entry-level fare whose upgrade path is that maker's tube product line. That tube product line is where they make the revenue, and that business is bolstered by a customer base that believes in tube amplifier myths. Of course, they are not going to produce myth-busting solid-state amplifiers.

If you want great sound, get a powerful professional audio power amplifier, in a wattage that is at least four times more than the tube amplifier you would otherwise use (chances are it will still weigh less than the tube amp, and fit into two "U" spaces of a 19" rack, and cost less).

Look for something that can be bridged to mono for more power: for guitar, you don't always need stereo.

Lastly, I want to mention something noteworthy: tube amplifiers face speakers with a relatively high output impedance, whereas the output impedance of solid-state amplifiers is low. This creates a different frequency response: the way tube amps drive a speaker naturally creates a "mid scooped" response that works with guitar. This is absent in a solid-state amplifier unless it is designed for guitar and compensates for it in some way. The tone controls of a guitar preamp might go only so far; it helps to have an equalizer.

(I'm not writing about digital modeling situation here, but using a bona fide analog pre-amplifier with tubes and a traditional tone control stack, going through perhaps some effect loop, and into a power amp. It's useful to have an EQ: graphic or parametric.)

P.S. It's worth mentioning a tube amplifier effect that doesn't easily reproduce in off-the-shelf solid state amplifiers: power supply sag. Tube amplifiers (which are totally tube, and thus use a tube-based rectifier for converting AC to DC instead of a silicon diode-based one) can exhibit a "sag" in their internal power supply voltage when used at high volume, during transient peaks. This creates an interactive, dynamic quality in the note attack, like compression. Someone who is specifically seeking this effect will probably not be satisfied by any combination of tube pre-amp and solid state power-amp. It may be obtainable via modeling or some amplifier mods to give the power supply apparent internal resistance or some such thing. So for instance, someone using Eddie Van Halen's aforementioned trick of using an entire tube amplifier as a pre-amplifier could obtain power supply sag, using the right amplifier used in the right way. Then that effect would simply be reproduced in the subsequent solid-state part of the signal chain.

P.P.S. Another tube amplifier effect is the fact that the output signal passes through an output transformer, which introduces nonlinearities, particularly when the transformer's iron core is saturated with magnetic flux. For a time, transistor amplifiers were built with output transformers. Modern transistor amplifiers aren't and so the output transformer effects are absent.

1

Benefits can be cost or convenience in using the design as an amp. Also in reliability and other NFRs such as perceived sound differences. Like @Neil Meyer said it is really up to personal preference to decide.

0

Don't ignore the subjective effects.

A channel that I find fascinating, regularly does blindfold tests to see if people can really tell the difference between, say, an expensive guitar and a cheap copy; a 'boutique' effects pedal and a mass-produced one, etc. In a lot of cases the differences are illusory. Even experienced experts can be fooled.

Andertons do lots of comparisons like this and the results are fascinating. I'm pretty sure they've compared valves with solid state several times over the years. Here are a couple of examples from 6 years ago. Modelling has improved enormously since then.

I'll let you draw your own conclusions.

Blindfold Clean/Blues Amp Test - Modelling vs Solid State vs Tube

Blindfold Challenge - Driven Guitar Amps - Modelling vs Solid State vs Valve

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.