[TWIFOSP]

So I was reading up a bit on tube amps, and specifically how they work ect.

I came across some information on why tube amps sound differently:

--snip--
For example SET's have a very well-known pattern of distortion that produces what is known as a second harmonic distortion. Second harmonic distortion is musically equivalent to adding the same tone one octave higher, to form a chord. In this case the added tone is at a lower level (not as loud) but the effect is to "fatten" the sound. This second harmonic tends to be very pleasing to the human ear, thus making SETs extremely pleasant to listen to for long periods of time.
--snip--
(http://stereos.about.com/cs/gtamplif.../tube_amps.htm)

So I was curious, are there are solid state amps that have attempted to reproduce this so-called second harmonic distortion? The purist in me would want a switch that would turn it on and off. But having just got into tube amps myself, (and still very newbie) I can't argue that they do sound very pleasing. I find myself not carying about the analytic side of the presentation. I still analyze music, detail ect just for kicks, but when I want the goosebump feeling, man I'm tube all the way.

Could we have the best of both worlds by artificially introducing this second harmonic chord in a solid state amp? Has this ever been done?

 

[Wodgy]

Most people can't hear second harmonic distortion, even at relatively high levels (<8%). If you injected just second harmonic distortion into the signal it wouldn't sound any different. It's more the overall distortion profile that accounts for the tube sound.

Class A, zero feedback solid state amps using JFETs (or MOSFETs) sound most like tube amps.

 

[JohnFerrier]

"May the Force be with you."

My recommendation is to use solid state devices that most closely match the distortion pattern of tubes.


Quote:

Originally Posted by Wodgy Class A, zero feedback solid state amps using JFETs (or MOSFETs) sound most like tube amps.

Well, look at that...


JF

 

[TWIFOSP]

What kind of opamps do they use?

If we can't hear the second harmonic, then why do tube amps sound different and more full? The overall distortion patterns... hrm...

So I guess this is old hat and I just need to figure out which solid state op amps have those?

 

[Earwax]

Quote:

Originally Posted by Wodgy Most people can't hear second harmonic distortion, even at relatively high levels (<8%). If you injected just second harmonic distortion into the signal it wouldn't sound any different. It's more the overall distortion profile that accounts for the tube sound. Class A, zero feedback solid state amps using JFETs (or MOSFETs) sound most like tube amps.

Such as a Szekeres?

 

[Wodgy]

Quote:

Originally Posted by TWIFOSP What kind of opamps do they use? If we can't hear the second harmonic, then why do tube amps sound different and more full? The overall distortion patterns... hrm... So I guess this is old hat and I just need to figure out which solid state op amps have those?

There are many JFET input op-amps (e.g. the AD8620) but op-amps also use colossal amounts of feedback.

You're probably looking at fully discrete amps. The Szekeres is a good example, as is the Borbely headphone amp.

 

[TWIFOSP]

D'oh. This has gotten over my head already

Off to do more research apparently. . .

 

[JaZZ]

It's not all that clear. If we think we can make a solid-state amp sound like a tube amp, we're supposing that it sounds neutral from the beginning. But fact is that solid-state amps don't all sound the same, but show equal sonic differences as tube amps among them. So we could argue just as well that tube sound is synonym for the absense of solid-state colorations.

Even though I don't think that this is an absolutely adequate scenario, I guess there's more in tube sound than just distortion. Some tube amps have very low distortion (and not even dominating 2nd order harmonics) and nevertheless don't sound like solid-state amps -- or still sound like tube amps in their positive characterization.

 

[JohnFerrier]

Quote:

Originally Posted by JaZZ Even though I don't think that this is an absolutely adequate scenario, I guess there's more in tube sound than just distortion.

Yep, higher noise. Tubes also have the disadvantage of not coming in complementary devices that I suppose contributes to what you refer to as solid state "coloration".


JF

 

[JaZZ]

Quote:

Originally Posted by JohnFerrier Yep, higher noise. Tubes also have the disadvantage of not coming in complementary devices that I suppose contributes to what you refer to as solid state "coloration".

I don't get what you are trying to say. Noise has never been an issue with tube amps for me. And solid-state amps can indeed sound quite different. I've not heard two headphone amps sound identical. Far from that. And tube amps don't necessarily alter the input signal more than solid-state amps -- that's my experience gained with the direct-path method.

 

[JohnFerrier]

Quote:

Originally Posted by JaZZ I don't get what you are trying to say.

Well, JaZZ. You say that unmeasureable differences have an affect on sound. Why wouldn't measureable differences have an affect on sound?

Tube amplifiers measure appreciably higher noise across the audible band than solid-state amplifiers.


JF

 

[antonik]

There are some other interesting differences as well. Tubes do not suffer from some serious problems of solid state devices - such as strong temperature dependency of most parameters, unlinear change of internal capacitances with voltage, and changes in the frequency parameters from the operating point. In essence, a transistor has many more unlinear variables than a valve. Usual way of counteracting these is with NFB, introducing a loop and producing much more complex patterns of distortions, thought usually at much lower level. Many unlinearities of a transistor amplifier are not really measurable on a (usually used for it) simple waveform continuous signal. Music usually is a very complex signal, with huge changes in the signal level. Valves are reacting on such a signal in a much more natural and predictable way, and the result could be more natural for our ears.

This does not mean it is impossible to create a very naturally sounding transistor amplifier, but it is in many ways much more difficult than with valves, especially without NFB.

Valves have other problems - limited current capabilities, necessity of transformer coupling in most cases, microphony, limited life and slow change of the parameters with time, etc. Transistors and IC's are much simpler in mass production, however if one wants to create a really good sounding amplifier, all options are open, and valves can give you an easier way to achieve this goal.

I dedicated some of my time during last year and half to a possibility of creation of a solid-state circuit which in many ways can "copy" the behavior of a valve triode, producing low distortion amplification without any NFB. It is a very interesting subject and one day I'll write an article about it.

Alex

 

[JohnFerrier]

Quote:

Originally Posted by antonik I dedicated some of my time during last year and half to a possibility of creation of a solid-state circuit which in many ways can "copy" the behavior of a valve triode, producing low distortion amplification without any NFB. Alex

In your definition of NFB, you probably exclude follower circuitry (degenerative FB, i.e. emitter and source followers), right? This is another subject of debate. To avoid that controversy, I describe follower circuits as short feedback or single stage feedback (rather than long global feedback).


JF

 

[antonik]

Quote:

Originally Posted by JohnFerrier In your definition of NFB, you probably exclude follower circuitry (degenerative FB, i.e. emitter and source followers), right? This is another subject of debate. To avoid that controversy, I describe follower circuits as short feedback or single stage feedback (rather than long global feedback). JF

No, I don't. A follower is a feedback circuit. If we need a low output impedance without a transformer coupling we have to use a follower. There are very few possibilities to build an amplifier without NFB altogether. However fewer loops (even local) is in my opinion better.

Alex

 

[DrewWinters]

Someone has already done it! Or maybe not, I don't know, but they claim to have made a solid state amp sound like a tube amp.... The description of Sunfire amps (these are loudspeaker amps, not headphone amps) says this:

"Two output options: current source for a warmer, more open tube sound or voltage source for tight response"

Here is a more detailed description of this :http://www.sunfire.com/pdf/sunwhite05.pdf

So then, how does this hold up to the scrutiny of any of our audio engineers? I've wondered about this question for as long as I have known about tubes. Has anyone here ever heard a Sunfire to verify whether or not this little trick actually works?