Tube amp power watts equivalent to Solid State?

There is an issue here that has not yet been discussed. It has to do with something called 'space charge effect' that often occurs in tubes and does not occur in transistors.

First- space charge: this is where a tube is conducting and some of the electrons bounce off of the plate and gather in its vicinity. This usually happens more near the point that the tube is about to saturate, depending on the tube. Pentodes BTW are designed to minimize space charge effect.

The space charge is thus an excess of electrons near the plate. This has the effect of reducing the tube's ability to conduct and makes it harder to completely saturate.

In practice, the result is that the tube will not hard-clip like a transistor will. This means that a tube amplifier will have a clipping characteristic that can be quite 'soft' if just barely clipping; the amp will enter saturation in a gradual or perhaps even graceful manner, with less of the odd ordered harmonics that are caused when clipping onset is immediate as in transistors.

Because some odd orders and in general higher ordered harmonics are present, which the ear uses as loudness cues, the amp will just tend to sound louder at this point, but without breakup that accompanies hard clipping.

But even with hard clipping, tube amps do not make as much odd ordered harmonic distortion due to the space charge in the power tubes. The result is they sound smoother to the human ear.

This is why guitar players tend to prefer tube guitar amps BTW.

So the bottom line is the reason tube amps often seem to operate with more power than they should has everything to do with how they overload; specifically the reduced amount of odd orders present at clipping. If you get rid of this 'soft clipping' characteristic you often need a lot more power to seem to do the same job. Makes sense now?

Al, Ralph......we are all waiting. Try to be as accurate as Unsound was please. ;^)

If we actually graphed the amplifier's distortion response when subjected to a standardized non-repetitive waveform, we could probably put numbers to that. Until then its a subjective experiment.

I can tell you this- if distortion is present it can cause the sound to seem considerably louder than one might expect. So given two amps, one lacking distortion and the other with distortion, the more distorted one is likely to sound considerably louder.

You know how SETs, when reviewed, frequently have the comment (paraphrasing):'...this amp sounded considerably more dynamic than it had any business being on account of its low power...' ?

This is entirely due to the distortion of the amp, which is unmeasurable at lower power levels. But if the power goes over about 20-25% of full power, the higher ordered harmonics come into play, which are likely only on the transients of the music. So you have the loudness cues on the transients and considerably lower distortion in between- the result will be that the amp sounds really dynamic. But its actually an interaction between how we perceive sound and the behavior of the electronics.

So to answer the question, I would say it depends on the amp and to no small degree the load (which must be benign). But the tube amps may seem to have anywhere from a 3 to 6 db advantage over transistor amps (of the same power) in this regard, depending on the topology of the amp.

I can break that down further- SETs will be more likely to seem to behave as if they have a lot more power, followed by push-pull triode amps. On the other end of the scale, a push-pull pentode amp employing feedback will seem to have the least 'overhead' in this regard, as they will have a sharper clipping onset.

The better tube amps will often seem to compress slightly before outright clipping occurs. This has a lot to do with improved power supplies, as a robust supply will reduce IMD in the amp.

Atmasphere, am I to understand that you're suggesting that as long as an amp avoids actual clipping that it will sound more powerful the more it distorts?

Yes, to a certain degree. It will sound **louder** more specifically.

This distortion thing can be a big deal. To give you an example, we try hard to make sure our amps don't make much in the way of higher ordered harmonics (even orders are canceled not just in the load but in each stage of the amp, and with only one stage of gain, triode operation etc., odd orders are minimized). What we find is that a person that might have a 7 watt SET will turn the volume down before the amp is outright clipping "because it was loud enough', whereas the same customer using our 60-watt amp will drive the amp to nearly full power on the same speaker.

We've actually had customers call up and ask us why the amp does not seem to make any power; that changes if they actually bring a sound pressure meter into the room and see how loud they are playing.

IOW, tube amps make 'loudness cues' that interact with the human ear/brain system in a way that transistor amps do not. IOW its all about distortion.

IMO/IME experience you do indeed want the amp to be as undistorted as possible and lacking colorations (the ear interprets lower ordered harmonics, the 2nd, 3rd and 4th as 'lushness', 'warmth', etc.) because its all about the amp being part of musical reproduction rather than musical generation, if you get my drift. Doing that is a bit of a trick! If you add loop feedback to reduce distortion, you will actually add odd ordered harmonics (to which the ear is quite sensitive, and also finds to be irritating). Norman Crowhurst described this in his writings a good 50 years ago; not much progress has been made in that area since. That is why we avoid feedback when we can. But then you pay a different price for that, see
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php