Asymmetrical waveforms !

Subjects that don't have their own home
Post Reply
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#1 Asymmetrical waveforms !

Post by steve s »

Can I bounce this badly put question to you for consideration?

If push pull amplifiers cancel distortion by cancelling asymmetrical signals ie distortion on the push or pull, and the result is a symmetrical output.. which I may have wrong ?

What happens in the stage when you put an asymmetrical signal into it ?
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#2 Re: Asymmetrical waveforms !

Post by steve s »

This seems to confirm my assumption and what my old brain was thinking..

https://digilander.libero.it/paeng/thd_ ... n_push.htm
About 1/3 of the way down

But I put my question very badly
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#3 Re: Asymmetrical waveforms !

Post by steve s »

And I think I've miss read that article.. doooh...
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#4 Re: Asymmetrical waveforms !

Post by steve s »

Copied this from a guitar site
Not a lot different to music waveforms

An AC guitar signal voltage causes electrons to alternate their direction through a circuit. During the positive half of the signal swing the electrons will move one direction and during the negative half they reverse course and move the other direction. With a symmetric AC signal the electrons will move the same distance back and forth through the circuit so there is no real movement of electrons on average, in other words there is no direct current (DC) component to the signal.

But an asymmetric AC signal voltage, where one lobe is larger than the other, will cause the movement of electrons more in one direction than the other which is a DC current. As an extreme example, if you completely clip off the positive lobe of an AC signal it becomes a pulsing DC signal--electrons move in a pulse in one direction. So asymmetric signal voltages have a DC component that will charge capacitors and cause voltage drops over resistors which can cause grid bias voltage excursion.
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#5 Re: Asymmetrical waveforms !

Post by steve s »

Why am I asking this...

Well I read today that most music the waveforms are asymmetrical
Not symmetrical like the test gear we use to test amps
I'm probably going down a rabbit hole though..
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#6 Re: Asymmetrical waveforms !

Post by Nick »

steve s wrote: Sun May 22, 2022 4:25 pm Can I bounce this badly put question to you for consideration?

If push pull amplifiers cancel distortion by cancelling asymmetrical signals ie distortion on the push or pull, and the result is a symmetrical output.. which I may have wrong ?

What happens in the stage when you put an asymmetrical signal into it ?
The answer to your if is no. A push pull amplifier will cancel distortion in the output device, ie make the output stage more linear by canceling even order distortion. However the signal is not distortion and it will not cancel that.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#7 Re: Asymmetrical waveforms !

Post by steve s »

Thanks nick I know after I wrote, that was the answer..
But ive got a worm in my head ..

Bear with me...
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#8 Re: Asymmetrical waveforms !

Post by JamesD »

Nick,

Am I right in thinking that the output device in the PP OPT case is the output transformer and not the active devices i.e. valves themselves.

...And in a OTL or conventional solid state PP device the cancellation takes place in the load - again not in the active device.

...And all of this is complicated by global feedback i.e. the above only applies 100% to non global feedback designs...

James
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#9 Re: Asymmetrical waveforms !

Post by Nick »

Vague wording on my part. I meant output device as opposed to complete amplifier. I guess I was considering the output valves and shared cathode circuit and output transformer primary as a composite device. The transfer function of each valve remains the same (of course) it's only when the two sides lines are overlaid you get the more linear transfer function.

In this case I suppose the driver stage could also be included if it's balanced.

I was thinking more about this and the source of Steve's confusion. The amp only knows about now and now-1 (the previous now) in terms of descret values and the rate of change in the time between n and n-1. So in that sense there is no concept of symmetric or asymmetric. That description made sense to me but may say more about my world view and complete lack of useful descriptive skills.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#10 Re: Asymmetrical waveforms !

Post by Nick »

And that waffle is also wrong in that inductors and capacitors also have a memory on that they will form a summation of current and voltage. So you get rectification effects from the signal.

Bring back analogue computers :-).
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#11 Re: Asymmetrical waveforms !

Post by JamesD »

Indeed!

Completely agree that the asymmetric, symmetric signal argument is irrelevant - music signals are often asymmetric (more often than not) and by the arguments logic we would not hear them in PP amps when they are...

Its often overlooked that all components, moment by moment, react to what comes next from their present state and not from a resting state and that reaction might well be different from the resting state one... in time domain as well as in frequency domain (and both the time response and the frequency response reflect into each others domain i.e. they are not truly orthogonal domains...)

argh..... back to Steve C's simple and beautiful EL34 SE before I do an Ouzelum bird impression...
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#12 Re: Asymmetrical waveforms !

Post by steve s »

Thanks gents great info..
Just to clarify (I know !!!)
And I'm bring very hypothetical

If the signal is asymmetrical
And as a result, that results in a partly dc waveform.

And that dc can charge caps ?

Would that have any effect on the original waveform ?
The tube manual is quite like a telephone book. The number of it perfect. It is useful to make it possible to speak with a girl. But we can't see her beautiful face from the telephone number
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#13 Re: Asymmetrical waveforms !

Post by Nick »

If the signal is asymmetrical
And as a result, that results in a partly dc waveform.
Well, it depends on what you class as DC. Over time its likely that the signal will average out as symmetrical, so there will not be a DC component, however if you look at a smaller window of time, then its likely that the average signal value will have a value that is either +ve or -ve. But strictly speaking, as you are looking at a small window of time (small as in less than forever), its not actually DC, its a low frequency AC signal. As you look at smaller and smaller windows then the frequency of the lowest AC signal will increase and the likelihood that the average signal value will be more +ve or more -ve. However if you then look at a sequence of these small windows, the average of each windows value will tend towards 0v again.

The above is getting close to sampling theory and I should stop now.
Would that have any effect on the original waveform
Would what have any effect on the original waveform? The original waveform will have the same value. What you are actually describing is the low frequency response of the amplifier and how its gain will vary with frequencies that get more DC like.

And again you are back to James's distinction or lack of between time and frequency domains.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#14 Re: Asymmetrical waveforms !

Post by JamesD »

It is a fallacy that an asymmetric waveform always produces a partial dc offset - what ever the partial means in that context.

A symmetrical waveform or an asymmetric waveform can produce a dc offset it all depends on how the waveform sits relative to the long term average excursion value for the waveform and that value can be defined in several different ways to i.e voltage excursion, current excursion, power excursion - and its possible for each definition to produce a different value for the average dc waveform Note long term average means over several cycles of the waveform repetition frequency - if it doesn't repeat then all bets are off - maybe.... it depends...

OK its more likely that an asymmetric waveform will be defined in such a way as to show a dc offset but in practice for a capacitively coupled signal the waveform if repeating at a audio frequency or higher will balance itself around the circuits dc value in the same way that a symmetrical waveform will...

Of course some circuits are designed to react differently to symmetric and asymmetric signals and they are called guitar amplifiers :-)
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#15 Re: Asymmetrical waveforms !

Post by Nick »

If you intentionally add A DC component to the signal, in effect redefine the signal's 0v, then that may have an effect on the amplifier depending if its DC coupled or not. If its DC coupled (or at least the part of the amplifier that is DC coupled), then it would be expected to amplify the DC part as part of the signal. So a 1v RMS sine wave with a +2v DC offset will vary from +0.6v (2-1.414) to 3.4v (2+1.414). If that went through a DC coupled amp with a gain of 2 the output would be a signal from +1.2v to +6.8v, with an average DC value of +4v. The same signal though a AC coupled amplifier with a gain of 2 would produce an output that varied from -2.8v to +2.8v with an average value of 0.0v.

Depending on the use you want to make of the amp, you may consider the different output as either distortion or just fine. For audio it may be fine, for a servo motor it would be broken.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
Post Reply