Asymmetrical waveforms !

Subjects that don't have their own home
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#46 Re: Asymmetrical waveforms !

Post by JamesD »

Thats awfully close to QUADs straight wire with gain...

But I guess we all go for closest approach to the original sound :-)
User avatar
Nick
Site Admin
Posts: 15709
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#47 Re: Asymmetrical waveforms !

Post by Nick »

Thats awfully close to QUADs straight wire with gain...
Yep, and equally imaginary.

Maybe I have just given up, but I went full circle and think I now have a amplifier that's good enough for me, and the only problems I am aware of are generally that of omission. The output impedance is maybe too low for some loudspeakers, it would sometimes be nice if it added a bit of 2nd harmonic. That sort of thing. But I have a DAC with a valve output stage and a phono stage that's got valves in it, so there is enough colour. And resistors exist :-)

There is nothing particularity special about the circuit, bipolar nothing that would be out of place in Cordell's book. Somewhat obsessive addition of cascades (once I realized they were just grounded base buffers that allowed you to fix the voltage conditions for the driving device), and the bipolar push pull drive to the industrial mosfet output devices.

I got 25% of the way through building a 2a3 amp for the Quazar copies, but its sat and not been touched for the last year or so. Once I have them finished if the SS amps are not a good match I may look elsewhere, but the elsewhere may well be a MoFo.
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
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#48 Re: Asymmetrical waveforms !

Post by IslandPink »

Nick wrote: Tue May 24, 2022 9:06 pm But I have a World-Class DAC with a valve output stage and .....
Ed
"Once you find out ... the Circumstances ; then you can go out"
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#49 Re: Asymmetrical waveforms !

Post by ed »

don't get me involved! snigger

but seriously, with only half an ear to ground, and half a memory....

there is a thread(can't find it now) about when I was playing with srpp front ends for a 2a3. What I do remember is that when I played with the balanced srpp(balanced with respect to the valve in question), the wave(sine) became much more symmetric. On researching I discovered that the asymmetric was one way of creating distortion in stomp boxes.

just saying because I don't think it has much to do with Steve's OP....
There's nowhere you can be that isn't where you're meant to be
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#50 Re: Asymmetrical waveforms !

Post by IslandPink »

ed wrote: Wed May 25, 2022 8:29 am don't get me involved! snigger
Ha !
"Once you find out ... the Circumstances ; then you can go out"
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#51 Re: Asymmetrical waveforms !

Post by steve s »

Ed, before my initial post I had done as much searching as I could

Most of the references to asymmetrical wave form where as a result to unbalanced components in the push pull and the like, or effects as you describe

As an aside the talk about supplies and modulation and what to many of us has been a discovery, is nothing new, but hardly documented
The fact the that modulation can be clearly heard due to the basic power supplys, that have become a standard
was blamed by many on just being being a valve amp.
The percieved improvement in zero feed back triode amps, was always conteracted by the fact that rock music was not within their best abilities.
And as a result, not truly understood.

There is no doubt in in every component the 'good enough' is the deciding factor, powersupply included

But I'd still ask how many push pull amps are tested to see their reaction to diffences when a unbalanced wave is presented to them.

'Theoretical' is how push pull cancellation is described in the best references
That says something in itself !
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
jack
Thermionic Monk Status
Posts: 5493
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#52 Re: Asymmetrical waveforms !

Post by jack »

Taking the input signal from the time to the frequency domain, ISTR that the result is a DC component which represents any DC offset in bin 0, then the resulting harmonics as pure sine waves with no DC offset. As this is the frequency domain analog of the time domain signal, they are identical. I find it easier conceptually to get my head round the frequency domain. Note that an FFT is taken over a range of input data points, so it's bin 0 value is essentially an average over the number of samples whereas in the true analogue domain, the DC value may be considered as continuously varying.

With varying asymmetric input signals, the DC offset/bin 0 value will vary. This may of course change the operating point of some components in the signal chain, however all the other bins are unaffected.

The implication here is that with global negative feedback, any distortion introduced by any component in the amplifier regardless of the DC condition is removed - this is, of course,. the ideal state. The Quad 405-2 (don't know about other Quads) has a DC servo to try to reduce any average/nett DC condition at the output - pretty effective, ISTR, as mine had a max of about 23mV at the speakers (as measured with a Keithley 2015 true RMS bench meter and an ideal diode with integrator)

Does the above make sense?

I had very little sleep last night & am going back to my coffee now.
Last edited by jack on Wed May 25, 2022 1:44 pm, edited 1 time in total.
Vivitur ingenio, caetera mortis erunt
User avatar
Nick
Site Admin
Posts: 15709
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#53 Re: Asymmetrical waveforms !

Post by Nick »

Does the above make sense
Yes, and its close to my understanding of the definition of what a asymmetric signal is, one that has a net DC offset over the window that the sample is taking place.

But I don't know for certain if that is what is being meant by a asymmetric signal. I could for example construct a signal with a sine form above the 0 line and a triangle below the zero. I could also construct it such that the area under both half was equal, so over one period the net DC offset was 0. Now that may meet what Steve means by such a waveform, or it may not. But as James points out, if you are interested in the RMS power value of the wave instead of the voltage, then it will be different, because of the squaring function meaning the two area were no longer equal. But given that RMS is a average function, not sure if looking at such a small duration is valid.

But I think the main question from Steve is
But I'd still ask how many push pull amps are tested to see their reaction to diffences when a unbalanced wave is presented to them.
I would answer that every such amp that has been tested with a intermodulation signal. For example:

Image

And DIM would also probably fit:

Image

It may be those tests dont meet the requirements for a unbalanced wave If so then a better definition of such a wave is needed.
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
jack
Thermionic Monk Status
Posts: 5493
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#54 Re: Asymmetrical waveforms !

Post by jack »

Nick wrote: Wed May 25, 2022 12:32 pm
Does the above make sense
Yes, and its close to my understanding of the definition of what a asymmetric signal is, one that has a net DC offset over the window that the sample is taking place.
...
That's a relief - it's a while since I wrote any FFT/DFT stuff. As a slight aside, FFTs assume a continuous and repetitive waveform - as the practical realisation of FFT algorithms don't allow this, windowing is used to reduce leakage due to the initial and terminal step functions. i.e. the FFT is an approximation, albeit a good one, to an ideal frequency domain analog.

As the sampling window size get asymptotically closer to zero, the bin 0 value (DC) will approach the instantaneous amplitude of the source. A sensible approach might be to monotonically increase the exponent of the window size until the bin 0 value stabilises. Who knows? If the effective DC component varies with time, it just becomes another AC component, albeit at a very low frequency (*). And then there's Nyquist.

If you wait long enough (or short enough) any physically measured nominally DC signal will become AC, either with an incredibly short or long cycle - think Johnson or Shot/Poisson noise through to DC drift. Reducto ad absurdum.

Oh well...

(*) I once wrote an FFT suite to analyse the kernel mode CPU usage of a huge Sun mainframe system - I found repetitive peaks in the noise with a cycle time of 25 secs...
Vivitur ingenio, caetera mortis erunt
User avatar
jack
Thermionic Monk Status
Posts: 5493
Joined: Wed Dec 29, 2010 8:58 pm
Location: ɐılɐɹʇsnɐ oʇ ƃuıʌoɯ ƃuıɹǝpısuoɔ
Contact:

#55 Re: Asymmetrical waveforms !

Post by jack »

jack wrote: Wed May 25, 2022 2:36 pm ... found repetitive peaks in the noise with a cycle time of 25 secs...
Senior moment. It was about 400 seconds.
Vivitur ingenio, caetera mortis erunt
steve s
Shed dweller
Posts: 2829
Joined: Wed May 30, 2007 6:19 pm
Location: east yorks

#56 Re: Asymmetrical waveforms !

Post by steve s »

Taken the below from a pro music site regarding reason that an asymmetrical system exists in music


.'This asymmetry is due mainly to two things, the first being the relative phase relationships between the fundamental and different harmonic components in a harmonically complex signal. In combining different frequency signals with differing phase relationships, the result is often a distinctly asymmetrical waveform, and that waveform asymmetry often changes and evolves over time, too. That's just what happens when complex related signals are superimposed.

The other element involved in this is that many acoustic sources inherently have a 'positive air pressure bias' because of the way the sound is generated. To talk or sing, we have to breathe out, and to play a trumpet, we have to blow air through the tubing. So, in these examples, there is inherently more energy available for the compression side of the sound wave than there is for the rarefaction side, and that can also contribute to an asymmetrical waveform.'

It also appears that there is little chance of a dc offset with this type of signal

The following describes the possible effects.. again pro music

''However, a DC offset can be very easily corrected by passing the audio through a high-pass filter tuned to a low frequency (typically 10Hz or lower). It is important to correct DC offsets when they do occur, because editing between an audio clip with a DC offset and one without results in a loud thump or plop at the edit point, which is not good!

In contrast, natural waveform asymmetry cannot be 'corrected' with a high-pass filter, and a rather more complicated solution is required called a 'phase rotator'. Generally, there is no need to 'correct' a naturally asymmetrical signal, but occasionally the asymmetry can restrict how much the signal can be amplified because the stronger half of the waveform will reach the clip level before the weaker side. By using a phase rotator process to alter the harmonic phase relationships, a more balanced symmetry can be established, allowing slightly more gain to be applied before both sides reach the clipping level at the same amplitude. Asymmetrical waveforms can also sometimes confuse the side-chain level-detection circuitry (or algorithms) of some compressors, resulting in less effective compression than might be expected.''  
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

#57 Re: Asymmetrical waveforms !

Post by steve s »

So .. again thinking out loud..

If we look as a push pull amp as two separate amplifiers (this is correct for class b, less so for a or ab but bear with me) and taking into account my last post about a naturally accuring asymmetrical signal

The distortion in both sides could be different, as its related to the work each amplifier has to do, and that distortion is related the signal being amplified ?

Now if that makes sense, when the opposing distortions are combined in the speaker or op tx, for the theoretical cancellation, does the theoretical cancellation occur?
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: 15709
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#58 Re: Asymmetrical waveforms !

Post by Nick »

The distortion in both sides could be different, as its related to the work each amplifier has to do, and that distortion is related the signal being amplified ?
But at any one time the signal will not be both positive and negative at the same time, it will be whatever value it is.

Any amplifier will have a transfer function, that is a graph of the input signal against the output signal. For a push pull amplifier either one or both or the other device will be active at any one time. Its the lack of linearity in the transfer function that generated distortion in the amplifier.
Now if that makes sense, when the opposing distortions are combined in the speaker or op tx, for the theoretical cancellation, does the theoretical cancellation occur?
Again, I think there is confusion here, I think you are wanting to think of the amplifier as producing signal and distortion and they are in some way separate and then summed at the output and can add or subtract. Its the same sort of thinking you used with the regulator where you wanted to consider the B+ and the signal as being separate.

As I said thinking of a push pull amp as two separate amps is wrong, both in the case of a valve amp and a transistor amp. In the valve amp case, there will be a current from 0 to some value in each leg of the primary. The flux in the primary will be generated by the combined value of the two currents.

The asymmetric nature or not of the signal is only a thing if you look to the way the signal varies for a period of time. The amp only ever knows about now and the last now. if you pick a short enough period of time then all signals are asymmetric, and in practice all real life signals will be asymmetric but other than maybe creating a small DC shift in the operating points of the valves if there are capacitors in the unshared cathode (which will also happen with a single ended amp) I still don't see the problem.

As the article says, it may clip on one side before the other, but as long as (if its a push pull amp) everything is close enough to being balanced and equal, then its all good. Ok if you made a push pull amp with a el34 on one side and a el84 on the other it would be bad, but that's because the amp is bad. It would be equally bad amplifying a perfectly symmetric sine wave.
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

#59 Re: Asymmetrical waveforms !

Post by steve s »

Thanks nick, looking back I think my confusion is in the way push pull distortion cancellation occurs..
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
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#60 Re: Asymmetrical waveforms !

Post by izzy wizzy »

My understanding in a simplistic way of distortion is a single ended device doesn't swing as linearly pos as neg compared to a pp amp.

It's not so much that a pp amp cancels signal/distortion but it doesn't generate even order distortion to the same level as the push and pull is generated by devices going in the same direction. And in class A, they are offset by the other device going in the opposite direction. This is only looking at the output stage. What happens before that is another story.

Cancellation only occurs for common mode signals such as power supply noise.
Post Reply