• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

6V6 Output?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Exactly, which is supposed to be (to SET freaks) why PP amps supposedly sound poisonous.
I even had from Jack (E) supposed proof that there were things missing from PP outputs, which got cancelled out.... err NO!
'fraid there's lots of hearsay and the rumour mill is going into overdrive!

I don't like distortion, harmonious or no, it's still distortion.
Joined 2002
Paid Member
Then, what do you do to the distortion of the speaker driver?

As we do not live in a perfect world, there may be distortion elsewhere in the chain. But it will sound different. Loudspeaker distortion will sound very different from amp-distortion. And needle/stylus distortion differ again.
This is why some may hear the difference of 0,002 or 0,003% amp-distortion thru 1-3% speaker-distortion
A SE 2A3 amplifier was connected to a well respected 2 way speaker. There was a double throw double pole switch connected from the amp output to the speaker (could reverse the connections, and so reverse the phase).
A 100 Hz pure sine wave was connected to the amp input.

The speaker has 2nd harmonic distortion (like most real woofers).
The 2A3 amp has 2nd harmonic distortion (like most SE 2A3 amps).

The timbre (character of the sound) changed from one switch setting to the other switch setting.
The reason . . . 2nd harmonic is asymmetric, so with one switch setting the 2nd harmonic was partially cancelled; and with the other switch setting the 2nd harmonic was added and increased.

The result showed that a particular kind of amplifier distortion and a particular kind of speaker distortion are alike. Just change the phase and get more or less 2nd harmonic distortion.

This was a blindfold test, one person listened while the other person changed the switch setting back and forth repeatedly; then the persons switched rolls.
If you do not believe the test results, get an SE amp, a 100 Hz signal generator [or CD test tone], and a loudspeaker, then run the test yourself.

I met Eduardo B de Lima at a VSAC conference in Silverdale, Washington about 30 years ago.
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
..or so they say. 0.5V/div, 1k, set, -3dB.


  • os.jpg
    75.6 KB · Views: 227
I mis-spoke.
The 2nd harmonic is a sine wave too. All harmonics are sine waves.
But if you take the fundamental and add the second harmonic, the wave (that contains both the fundamental and 2nd harmonic) is not symmetrical.

This non-symmetrical waveform is in one direction. Take another one that is non-symmetrical in the opposite direction. Add them together, and they will partially cancel the 2nd harmonic (a push pull output stage; or a SE amp and loudspeaker driver that has 2nd harmonic distortion).

Check out a spectrum analyzer.
Check out an FFT.
You will find that no matter how complex the waveform, all the individual parts of the spectrum that is created, will be individual sine waves.

A harmonic is an integer number multiplied times the fundamental (2x, 3x, 4x . . .).
It is not 1.5x, 2.3x, etc.
If you strum two strings at the same time, you will have two individual sets of fundamental and its harmonics.

A square wave is fundamental, and Odd harmonics, 3, 5, 7, 9, .....
An asymmetrical square wave (un-equal 'up versus 'down' time) is Odd harmonics, 3, 5, 7, 9, . . . Plus Even harmonics 2, 4, 6, 8, . . .
Those are all individual sine waves which added together give the shape of the square wave.
Then what the :confused: are you doing listening to 6V6s?

I'm not, but I have a couple of CVx0xx wire end beasties, (quite rare!) a pair of decent SE OPT and would happily add a couple of low power channels to the front 50W stereo amps.

As you may know, rear surround channels require extremely low powers, and are very uncritical as to quality, perfect place for miniature triode wired 6V6, or a simple headphone amp outputting at the same time as to rear channels.
What's not to like? ;)
All harmonics are sine waves.....Not the harmonics coming out of my guitar!!

Quite some time ago a man named Fourier stated that all complex periodic waveforms are made up of a series of individual sine waves and their integer multiples (harmonics) also being sine waves. Nobody ever successfully proved him wrong, and his theory laid the ground work for all the complex DSP stuff that powers everything from guitar effects to cell phones.

When we use a PC and some software to analyze the distortion characteristics of an amplifier, we apply a piece of math called a Fast Fourier Transform (FFT) to take a complex signal apart and display its individual harmonics as a series of sine waves of varying amplitude.

True, the complex signal coming out of your guitar is NOT a sine wave. It is however made up from a collection of pure sine waves. At a given instant in time each of these sine waves will have a unique level and phase relationship with the others.

It is easy to make generalizations about the distortion characteristics of an amplifier and a speaker or guitar and state that with careful manipulation one could be made to cancel the other. Again on paper, and in real life this is true......BUT, that complex series of sine waves flowing from your guitar is NOT constant. There will be much more harmonic content the second the pick leaves the string, than there will be a few milliseconds later. The harmonic content of each individual note on a guitar, or the same note played on a different string has it's own constantly changing harmonic content.

It is possible to use the nonlinear characteristics of say a tube amplifier to PARTIALLY cancel the nonlinear characteristics of a speaker over a narrow range of amplitudes and frequencies. The same technique is applied when the 2H of one triode is used to partially cancel the 2H of another within the amp.

Given the serious processing power available in todays DSP chips it IS possible to design an amplifier or other signal processing system such that it applies "predistortion" to a signal that is exactly opposite of the distortion in another amplifier or system. For this technique to work the DSP must know all of the distortion characteristics of the device it is trying to correct over ALL of its possible operating conditions.

This technique is used today in LTE cell phones and base stations. In a phone where battery life is important a tradeoff is made where the DSP takes out the "low hanging fruit" so that the RF power amplifier can be made more efficient, without the DSP needing more battery power than it saves in the RF amplifier. In some cases intentional clipping or compression is allowed in order to increase the apparent dynamic range of the RF amplifier.

Can these techniques be used today in the audio world? Of course it can. Modern DSP's can do a fair job of analyzing the "tone" of a guitar / amp combination, and generating a "profile" from which the "tone" can be cloned.

For an amp / DSP to be able to completely cancel the speaker distortion, it (or the DSP programmer) would need to know ALL of the distortion characteristics of the speaker over ALL possible operating conditions.... its "profile" to use the words from a popular expensive German guitar amplifier.

As always, some enlightening points from you.

Talking about the transient nature of the guitar string as it is plucked (at startup, and after as certain harmonics die out before other harmonics), reminds me of other musical instruments.

A trumpet, french horn, etc. have a varying harmonic structure from the start of the note until it has time to settle to a constant characteristic.

The air 'Whiff' of a pipe organ or flute at the start of the note is another example.

The second harmonic is probably the easiest distortion to correct enough to get a partial cancellation. Push pull is one way.

The third harmonic is generally not as easily reduced.
Negative feedback is one method.
Shaping the amp's amplitude response to 'expand' the signal more and more as the signal (absolute) amplitude increases is another way. That has to expand both polarity of the waveform. Do that in an analog way, with shaping (a little like using digital numbers representing voltage, and applying a compensating correction table).
3rd harmonic can be either expanding or compressing, depending on the polarity. A class AB amp would have expansion in the beginning class B portion due to the individual tubes 3/2 power law, but then that would change over to compression as the tubes near saturation. Not going to be easy to fix.

There are the beam deflection tubes that can perform an analog multiply operation. Part of the fix. Or can dynamically adjust the tail current to an LTP stage. Some Op Amps use a common mode control loop to control gain. Could insert a HF common mode signal into the tube amp and monitor its common mode amplitude at the output stage. Use the variation of that to control the tail current of an LTP stage to hold the CM gain constant. likely will improve the differential gain flatness too.
Last edited:
If the positive peaks compress, and the negative peaks compress, that is a good example of 3rd harmonic distortion.

A non-fedback class A to AB push pull triode amp generally will do that.
And do not get the wrong idea, I do like many amplifier topologies, and this is one of them.

In the class A region, one plate's rp is increasing, while the other plate's rp is decreasing.
But when one plate turns off (AB), only one output tube is driving a load impedance of 1/2 of what the reflected load impedance is with both plates driving the load.

In this case, just remember that the 2nd harmonic of the Push tube, and the 2nd harmonic of the Pull tube, especially in the AB region is what causes a foreshortening of both the positive and negative alternations. Opposing phases of second harmonic distortion causes 3rd harmonic distortion in this case.
Last edited:
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.