Tube and solid state efficiency

Started by Venusblue, September 05, 2010, 06:42:31 PM

Previous topic - Next topic

Venusblue

I had a bit of a question that I don't really even know how to phrase.

If you were to compare the output of both, how do they compare? If your constant is the same guitar, pickups, 4x12 cab and all that, and you were only changing the circuit (amp heads, i suppose) between a 50w solid state head and a 50w amp running 6l6's?

Could you theoretically come up with a decibel to watt ratio for each? I realize that this wouldn't be exact or carry over through every circuit and amp, but I guess I was curious if such a thing would even be worthwhile to research. Is there too much of a difference between every circuit, or would the numbers still have some form of practical application? I don't really even know the term for phrasing this question.
I love the smell of baked tubes in the morning.

Labaris

As far as I know (and don't know very much) the tubes response varies with power (number of watts delivered to the speakers) while transistors have a more lineal response. So, depending on the power range you could have some differences in the efficiency, being the transistors the more efficient.

But I've always heard from guitarists that the tube-amp-watts are more watts than the transistor ones (wich can't be truth).

I think that an important thing to consider is the response of human ear, wich is totally non-linear and might make us perceive an amp louder than the other, even if there's no difference in the total energy they deliver.

That's all I can tell, sorry for not having a complete answer.
A long way is the sum of small steps.

petemoore

 A good hard question.
  RMS and Peak power...very different.
  Tube and SS, same-ish [when both are operating within thier linear volume range], very different when pushed past 'themselves', the SS amp clipped can spike current which is capable of popping the coil exits cone syndrome.
  Tube pushed past 'themselves' transition into and out of their linear range much more gracefully and simply cannot exceed their rated output, 'out of linear range' equates to 'distorting'.
  So..for speaker life past the first big peak, use a larger than needed solid state amplifier that it should never clip.
  For large clean power, SS seems to be the 'effecient' one, inexpensively done right is what I'd call it. Good, perfectly usable clean power is 'cheep SS'.
  Good clean tube power can be nearly free, or get very expensive, depending on the sources: Curbs, Garage Sales, Bench Gifts, or Boutique Tube Amplifier builder/seller...hugest variances in prices.
  For a 4x12''  in cabinet, to get cabinet sound, 100w of SS, 50w of tube isn't very good thing to say, not without trying or at least including whether RMS or peak power are the rating-index for both.
  So since I can't say a darn thing that is solid, statement like this seems appropriate...I had a 70w tube DYna amp, and we all agreed it was much more powerful sounding than the 250 and 300w solid state power amplifiers before their little red lights started to blink.
 
 
 
Convention creates following, following creates convention.

Labaris

#3
Quote from: petemoore on September 05, 2010, 08:45:57 PM
A good hard question.
 RMS and Peak power...very different.
 Tube and SS, same-ish [when both are operating within thier linear volume range], very different when pushed past 'themselves', the SS amp clipped can spike current which is capable of popping the coil exits cone syndrome.
 Tube pushed past 'themselves' transition into and out of their linear range much more gracefully and simply cannot exceed their rated output, 'out of linear range' equates to 'distorting'.
 So..for speaker life past the first big peak, use a larger than needed solid state amplifier that it should never clip.
 For large clean power, SS seems to be the 'effecient' one, inexpensively done right is what I'd call it. Good, perfectly usable clean power is 'cheep SS'.
 Good clean tube power can be nearly free, or get very expensive, depending on the sources: Curbs, Garage Sales, Bench Gifts, or Boutique Tube Amplifier builder/seller...hugest variances in prices.
 For a 4x12''  in cabinet, to get cabinet sound, 100w of SS, 50w of tube isn't very good thing to say, not without trying or at least including whether RMS or peak power are the rating-index for both.
 So since I can't say a darn thing that is solid, statement like this seems appropriate...I had a 70w tube DYna amp, and we all agreed it was much more powerful sounding than the 250 and 300w solid state power amplifiers before their little red lights started to blink.
 

So, when out of their linear range tube amps start to compress while solid state ones start to crash...
That compression (of tubes) means "loudness", as the RMS power goes up.

I'm I right?
Sorry Venusblue (for posting here), but I'm curious about this ;)
A long way is the sum of small steps.

R.G.

Quote from: Venusblue on September 05, 2010, 06:42:31 PM
Could you theoretically come up with a decibel to watt ratio for each? I realize that this wouldn't be exact or carry over through every circuit and amp, but I guess I was curious if such a thing would even be worthwhile to research. Is there too much of a difference between every circuit, or would the numbers still have some form of practical application? I don't really even know the term for phrasing this question.
Very insightful question. It has to be subdivided to make sense, but it's feasible.

There is a short answer. The short answer is, given speakers with the same efficiency at making electrical signal into sound loudness, the tube amp would require much more electrical power input from the wall power to make the same loudness as the solid state amp. This is a gross generalization, as there are inefficient and more efficient examples of each, but if you make an attempt to use the same biasing class for each, the solid state amp is more efficient every time.

That is because tubes have a high internal impedance and so a lot of power that goes through them gets converted to heat inside them. Solid state power devices have a lower internal impedance and lower losses per watt transferred through them.

To do a fair test, you have to use equivalent biasing classes. Class A biasing it the least efficient, and tubes suffer smaller disadvantages here. Class AB and Class B biasing is much more efficient in terms of wasted power in the amplifier for both tubes and solid state; but tubes have other losses not related to biasing, so they will be less efficient.

First, tubes have to have a heater. This is purely wasted heat, as it contributes nothing to the sound signal, it just has to be expended to get the tubes to do anything at all. Next, tubes work by dragging electrons through a significant distance through a vacuum, and then collecting them up as they hit a metal plate at the end. This process wastes the energy to pull the electrons over, and the energy they lose getting into the plate. This is summed up in tube terms as plate resistance. Finally, tubes really can't "saturate" like transistors can, conducting a lot of current and having the voltage across them drop very low. Power pentodes and power beam tubes (e.g. 6L6) really can't conduct enough current to get the voltage across the tube below 50V or so. Solid state devices can get down to a fraction of a volt in some cases, at much higher current. This comes out as heat.

All the common guitar amps are biased for Class AB. This reduces the standby power drain over class A by a lot. Almost all guitar amps over 15W output are Class AB, with a few notable exceptions. A 50W Class AB power amp may pull 150W at idle (no signal at all) and over 300W at 50W output. By contrast, a 50W solid state amp may pull less than 20W at idle and 120W at full output. These are very broad generalizations, but they're not unrealistic. The economic penalties for this are severe: The power transformer is almost always the most expensive single part in the amp, and the output transformer is usually second for low quality amps, and may swap places with the power trannie for careful, high quality designs. Then there's a choke. The solid state amp needs no output transformer or choke, and half the size of the power transformer.

The world was tired of the expense, weight, and wasted heat of tube amps when transistors came along. It is only because of the differences in sound quality, especially to musicians, that tube amps have survived in spite of being heavier, hotter, and costing more.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

edvard

+1
Transistors are MUCH more efficient at using the electrons fed to them.
What Pete Moore said about tubes being perceived as louder when they are pushed to the non-linear edges is also true.
Careful (and not-so-careful) use of compression and limiting during the recording process has resulted in the modern full-throttle-guitar-but-you-can-still-hear-the-vocals music so popular with the kids today.
Also, I've read that due to internal impedances, grid conduction, etc. the frequency response of a tube amp changes as it's driven harder, which may contribute to much of the mystique.

Does anybody know  if any solid state amp manufacturers have tried doing things like artificially inducing similar non-linear responses in the power section?
Might be a good idea for a DIY project...
All children left unattended will be given a mocha and a puppy

petemoore

  Does anybody know  if any solid state amp manufacturers have tried doing things like artificially inducing similar non-linear responses in the power section?
Might be a good idea for a DIY project...

  Aside from the 1,000's of approaches to this design goal, there's only one that I know of that seems to be novel or completely unconventional yet perfectly logical approach.
  Not sure if that's what's alluded to in the Valvetronix AD15VT preamp, no-one seems to have had a look inside to see what is meant by the phrase "Active Preamp Loading'' in this one.
 
 
 
Convention creates following, following creates convention.

teemuk

#7
Quote from: edvard on September 06, 2010, 10:42:10 AMDoes anybody know  if any solid state amp manufacturers have tried doing things like artificially inducing similar non-linear responses in the power section?

Yes. About all the biggest manufacturers have been for few decades introducing stuff to mimic them.

Peavey's "T-Dynamics" power amps, Korg's "Valve Reactor", Eric Pritchard's designs, and Hughes-Kettner's "Dynavalve" are some of the most recent examples.
Older stuff is e.g. TubeWorks' "MOSValve" power amp designs, Trace Elliot's "Reactor" power amp, ADA's, Ampeg and Randall's preamp tube-driven hybrids, and various soft clipping power amp designs from e.g. GMT, Yamaha, Sunn, Gallien-Krueger and Warwick.

Here's examples of some of my own little experiments:
http://www.ssguitar.com/index.php?topic=1372.msg11443;topicseen

The earliest examples of replicating the non-linear response resulting from driving a reactive speaker load with high-impedance output amp date to 1965 (Triumph's Silicon 100 amps). Today it's a pretty standard feature in instrument amps.

Earthscum

I've looked at D-Class amps off and on in schematics. Seems like the D-Class driver stage would work well with a tube pre, or a tube emulator pre. I've been contemplating digging into the D-Class for awhile now, and had planned on using complementary pairs in the preamp (in place of CD4049 inverters... those things are hissy!) because I figured the decreased leading and falling edge times would lead to a larger perceived loudness. If I'm not mistaken, R.G. just confirmed this, correct?
Give a man Fuzz, and he'll jam for a day... teach a man how to make a Fuzz and he'll never jam again!

http://www.facebook.com/Earthscum

earthtonesaudio

Quote from: Earthscum on September 06, 2010, 11:09:55 AM
I've looked at D-Class amps off and on in schematics. Seems like the D-Class driver stage would work well with a tube pre, or a tube emulator pre. I've been contemplating digging into the D-Class for awhile now, and had planned on using complementary pairs in the preamp (in place of CD4049 inverters... those things are hissy!) because I figured the decreased leading and falling edge times would lead to a larger perceived loudness. If I'm not mistaken, R.G. just confirmed this, correct?

If it's good enough for a Bassman...

I don't really understand what you're saying with regard to edge time and loudness.  I don't think of those two concepts as having any generic relationship at all.

teemuk

Quote from: earthtonesaudio on September 06, 2010, 12:08:48 PMIf it's good enough for a Bassman...

I like how the "12AX7 tube preamp" is actually just a preamp featuring a single 12AX7 tube - rest of the stuff being solid state. Kinda like most things marketed as "tube preamps" (e.g. ADA MP-1, Marshall JMP-1 etc).  The most difference people still can tell is that it says "tube" in it so it must sound awesome. :icon_lol:

alanlan


R.G.

Quote from: Earthscum on September 06, 2010, 11:09:55 AM
... because I figured the decreased leading and falling edge times would lead to a larger perceived loudness. If I'm not mistaken, R.G. just confirmed this, correct?
Huh? I don't normally think of decreased edge times and loudness as related either. What did I say that implies that?

'Course, it's always possible that I don't understand everything I know about this, too...  :icon_lol:
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Earthscum

#13
compression. The signal edges are steeper, your speaker pushes out faster and has a more rounded top to the signal. I guess I can't really explain it in words... it's the difference between a car with more torque vs horsepower, if that analogy makes sense. A car with more torque is going to be able to leave the starting line faster. If your signal rises and falls, the speaker will push the air into more dense waves, and when those waves expand out, they have more energy contained. Basically it seems to me that tube amp compression simply makes the speakers move faster (not frequency-wise). Basically more like an air cannon  than a (hand) fan.
Give a man Fuzz, and he'll jam for a day... teach a man how to make a Fuzz and he'll never jam again!

http://www.facebook.com/Earthscum

Incubusguy

#14
It's my understanding that the ear doesn't really react in that way to steeper edges. It responds more to average levels of sound energy. It seems to me that the human body in general responds more to averages than to transients. Perhaps this is some evolutionary response where constantly responding to transients would be a waste of time and energy, whereas responding to averages gives a better idea of how the environment is changing through time?

Quoting from here (http://en.wikipedia.org/wiki/Absolute_threshold_of_hearing): 'The ear operates as an energy detector that samples the amount of energy present within a certain time frame.'

Compression works because it increases the number of peaks which reach a given dB level in a given time. The brain perceives this as an increase in loudness.

I think that the extra energy contained in the signal having steeper edges is, to an extent, negligible compared to the sum of the total energy in the signal.

Again, I think it may be a case of the speakers moving more with compression and not faster, per se.