Svart
Well-known member
understood. I am just hoping to stoke some thoughts on this.
What's the distortion adding feature of the new $$L console?
- Thread starter Svart
- Start date
Help Support GroupDIY Audio Forum:
NewYorkDave
Well-known member
Possibly of interest:
http://www.aes.org/e-lib/browse.cfm?elib=12627
http://www.aes.org/e-lib/browse.cfm?elib=12627
recnsci
Well-known member
Tube doesnt "give" second order, circuit topology does. Single ended
BJT stuff (think N word) also gives second. I dont know who started this
false story in the first place (probably some 300B freeks long time ago),
about "toobs givez godly second, tranzisterz givez devils 11th", but that
urban lengend continues to live on. Proper pushpull toob circuit will give
nothin but odd order harmonics (up to the component tolerance and
matching). And if anything, jfets loosely resemble pentodes (mucho
emphasis on loose part).
As far as THD spectrum plots, these can be as misleading as "magick
toobe second" falacy. It wouldn be too dificult to make two circuits,
one single ended thermionic, other clean-plus-FWR solid state, that would
have ,with sinusoidal input of particular amplitude, THD spectrums matching
to 1% boundary. And they would sound nothing alike.
Me personaly wouldnt put any FWR within a mile of "decent" audio path
(note taht me also like indecent things). esesel probably uses lil' deiscrete
transistor amp, where changing of bias of this and that pushes 2nd more
or 3rd more on THD spectrum plot.
cheerz
ypow
BJT stuff (think N word) also gives second. I dont know who started this
false story in the first place (probably some 300B freeks long time ago),
about "toobs givez godly second, tranzisterz givez devils 11th", but that
urban lengend continues to live on. Proper pushpull toob circuit will give
nothin but odd order harmonics (up to the component tolerance and
matching). And if anything, jfets loosely resemble pentodes (mucho
emphasis on loose part).
As far as THD spectrum plots, these can be as misleading as "magick
toobe second" falacy. It wouldn be too dificult to make two circuits,
one single ended thermionic, other clean-plus-FWR solid state, that would
have ,with sinusoidal input of particular amplitude, THD spectrums matching
to 1% boundary. And they would sound nothing alike.
Me personaly wouldnt put any FWR within a mile of "decent" audio path
(note taht me also like indecent things). esesel probably uses lil' deiscrete
transistor amp, where changing of bias of this and that pushes 2nd more
or 3rd more on THD spectrum plot.
cheerz
ypow
recnsci
Well-known member
Widely known Id=Idss(1-Vgs/Vp)^2 is actually aproximation of
one big chunk of equation which is 3/2 based. And that big
chunk is also aproximation of phisics of JFET-alike device with well behaved
geometry. And real life JFET will be further off. Expo law of BJTs is much
better behaved than laws (or should I say "laws") that we think govern
behaviour of fets and tubes.
Every distortion would leave IMD products, and you cant, to the best of my
knowledge, "bandlimit" THD spectrum.
cheerz
ypow
one big chunk of equation which is 3/2 based. And that big
chunk is also aproximation of phisics of JFET-alike device with well behaved
geometry. And real life JFET will be further off. Expo law of BJTs is much
better behaved than laws (or should I say "laws") that we think govern
behaviour of fets and tubes.
Every distortion would leave IMD products, and you cant, to the best of my
knowledge, "bandlimit" THD spectrum.
cheerz
ypow
Svart
Well-known member
I would think they would be OK with a JFET.
My take on this is that they have devised some way of adding way more of this generated signal than simply running the audio through a variable output driver circuit. it sounds much more like the ability to totally saturate the audio.
but I could be wrong.
My take on this is that they have devised some way of adding way more of this generated signal than simply running the audio through a variable output driver circuit. it sounds much more like the ability to totally saturate the audio.
but I could be wrong.
recnsci
Well-known member
Its not that Im ranting bout tubes or transistors. I'm just pissed that lot
of people (especialy in audiophile and guitar amp markets )are purposly telling
a missleading stories about these "misterious" harmonics or whatever.
"band-limit both the input and/or output "-you mean something like
preemphasis-deemphasis ?
of people (especialy in audiophile and guitar amp markets )are purposly telling
a missleading stories about these "misterious" harmonics or whatever.
"band-limit both the input and/or output "-you mean something like
preemphasis-deemphasis ?
clintrubber
Well-known member
[quote author="mediatechnology"]Svart I'm not sure if I understand your first statement but they (SSL ad copy) used the word "or" as in second or third. I glean from that they meant you can switch second or third but not both at the same time.[/quote]
Just some FWIW-remark, who knows this thread gets sent into false directions by info from an ad, which is... not always entirely accurate of course, since an ad.
I'm not in a position to say that the SSL-people can have off-days like for instance Mackie can have when doing their PR, but reading about zero phase shift EQ in their (Mackie) analog board ads made me raise the issue when I was writing them anyway for something.
Got a neat response that they meant to write 'minimal' i.s.o. 'zero' of course. All set.
But how about asking SSL for some little unmistifying of this subject ? Or would that spoil the fun of the riddle ?
Just some FWIW-remark, who knows this thread gets sent into false directions by info from an ad, which is... not always entirely accurate of course, since an ad.
I'm not in a position to say that the SSL-people can have off-days like for instance Mackie can have when doing their PR, but reading about zero phase shift EQ in their (Mackie) analog board ads made me raise the issue when I was writing them anyway for something.
Got a neat response that they meant to write 'minimal' i.s.o. 'zero' of course. All set.
But how about asking SSL for some little unmistifying of this subject ? Or would that spoil the fun of the riddle ?
clintrubber
Well-known member
[quote author="mediatechnology"] but I do know how to make distortion.[/quote]
:thumb:
:thumb:
Kev
Well-known member
[quote author="mediatechnology"]So I'll just shut up now because I'm obviously too stupid to make a contribution here.[/quote]
:shock:
what !! ?
I never said you were stupid
where did that come from ?
... and my father and I disagreed with his math and the relavence to the original premise of the SSL distortion generation.
He did disagree with the identification of the DC term. His math is well beyond mine and I have no hope of explaining what he had to say.
funny how this thread is very near the thread I started at Tech Talk about asymetrical waveforms and the need for polarity reverse at the front end rather than at output ... especialy with single ended topology like the Neve1272 1290 and the typical single ended tube circuits.
back to the math
I think the FFT gives rise to quite a specific point of view and I do agree that generally and anecdotaly the asynmmetric stuff has higher orders of 2nd
( good sounding )
and without getting too far into wind intstruments (which I think is a worthwhile discussion topic and definately worth a thread of it's own ) Mics inside the balloon and mic outside the ballon
Mic's in front of AC air ducts and mechenical offset of the diaphragm ... but does this lead to elecrical offset ?
my basic and specific point was that an AC couple circuit as found in the SSL would probably not have the DC so ... how do they do it ?
:roll:
your suggestion about rectification is interesting and is possible.
create the DC offset with bias so as to get clipping on one part of the waveform ... yes also possible.
Intergration and filtering.
and
still my favourite is Solft Clipping with the above ideas as part of the package.
:shock:
what !! ?
I never said you were stupid
where did that come from ?
... and my father and I disagreed with his math and the relavence to the original premise of the SSL distortion generation.
He did disagree with the identification of the DC term. His math is well beyond mine and I have no hope of explaining what he had to say.
funny how this thread is very near the thread I started at Tech Talk about asymetrical waveforms and the need for polarity reverse at the front end rather than at output ... especialy with single ended topology like the Neve1272 1290 and the typical single ended tube circuits.
back to the math
I think the FFT gives rise to quite a specific point of view and I do agree that generally and anecdotaly the asynmmetric stuff has higher orders of 2nd
( good sounding )
and without getting too far into wind intstruments (which I think is a worthwhile discussion topic and definately worth a thread of it's own ) Mics inside the balloon and mic outside the ballon
Mic's in front of AC air ducts and mechenical offset of the diaphragm ... but does this lead to elecrical offset ?
my basic and specific point was that an AC couple circuit as found in the SSL would probably not have the DC so ... how do they do it ?
:roll:
your suggestion about rectification is interesting and is possible.
create the DC offset with bias so as to get clipping on one part of the waveform ... yes also possible.
Intergration and filtering.
and
still my favourite is Solft Clipping with the above ideas as part of the package.
pstamler
Well-known member
I'll go along with the simplest idea: somewhere in the channel strip there's a device operating open-loop or with very little feedback. With one level of bias it generates a good deal of 2nd-harmonic distortion; with a different level of bias it generates a good deal of 3rd-harmonic distortion. You can set up just about any active device (or perhaps cascaded pair of them) to do that; for example, put together a couple of triodes, one as a voltage amplifier, the other as a cathode follower, direct coupled. (The preamp I published several years ago in audioXpress, for example.) Mess with the plate, cathode and cathode-follower resistors, and the load, and you can get very clean response, or second harmonic and practically nothing else, or third harmonic and practically nothing else, or a big nasty stream of every harmonic under the sun. Just a matter of bias points.
My guess is that SSL is doing something like that.
Peace,
Paul
My guess is that SSL is doing something like that.
Peace,
Paul
Kev
Well-known member
cool
simple and seems easy to implement
and no soft clip as stupid :wink: Kev suggests
and so, not level dependant
simple and seems easy to implement
and no soft clip as stupid :wink: Kev suggests
and so, not level dependant
clintrubber
Well-known member
[quote author="pstamler"]I'll go along with the simplest idea: somewhere in the channel strip there's a device operating open-loop or with very little feedback. With one level of bias it generates a good deal of 2nd-harmonic distortion; with a different level of bias it generates a good deal of 3rd-harmonic distortion.
Just a matter of bias points.
My guess is that SSL is doing something like that.
Peace,
Paul[/quote]
Interesting take. Wondering how level-sensitive this would be. I mean, how accurately it would for instance remain in '2nd only' mode when levels jump around.
Have to get some textbooks out.
Regards,
Peter
Just a matter of bias points.
My guess is that SSL is doing something like that.
Peace,
Paul[/quote]
Interesting take. Wondering how level-sensitive this would be. I mean, how accurately it would for instance remain in '2nd only' mode when levels jump around.
Have to get some textbooks out.
Regards,
Peter
Kev
Well-known member
:roll:
and there was me thinking it wouldn't be so level dependant
I guess it all depends on where you place things on that non-linear amplitude curve
as Peter said
" ... where are those books ?"
and there was me thinking it wouldn't be so level dependant
I guess it all depends on where you place things on that non-linear amplitude curve
as Peter said
" ... where are those books ?"
clintrubber
Well-known member
[quote author="clintrubber"][quote author="pstamler"]I'll go along with the simplest idea: somewhere in the channel strip there's a device operating open-loop or with very little feedback. With one level of bias it generates a good deal of 2nd-harmonic distortion; with a different level of bias it generates a good deal of 3rd-harmonic distortion.
Just a matter of bias points.
My guess is that SSL is doing something like that.
Peace,
Paul[/quote]
Interesting take. Wondering how level-sensitive this would be. I mean, how accurately it would for instance remain in '2nd only' mode when levels jump around.
Have to get some textbooks out.[/quote]
A bit of a st*pid remark from me of course, of course THD won't remain unchanged when levels vary (all pre-clip).
For illustration let's have a look here. Note these are not tube-stages but BJT's with modest feedback, but OK:
http://www.dself.dsl.pipex.com/ampins/discrete/singleq.htm
While the pics don't break down the THD in spectral components it'll be clear that whatever the 'signature' of the THD is, it is varying with the signal level.
Has the SSL some AGC-like thing before the harmonics are made ?
Regards,
Peter
Just a matter of bias points.
My guess is that SSL is doing something like that.
Peace,
Paul[/quote]
Interesting take. Wondering how level-sensitive this would be. I mean, how accurately it would for instance remain in '2nd only' mode when levels jump around.
Have to get some textbooks out.[/quote]
A bit of a st*pid remark from me of course, of course THD won't remain unchanged when levels vary (all pre-clip).
For illustration let's have a look here. Note these are not tube-stages but BJT's with modest feedback, but OK:
http://www.dself.dsl.pipex.com/ampins/discrete/singleq.htm
While the pics don't break down the THD in spectral components it'll be clear that whatever the 'signature' of the THD is, it is varying with the signal level.
Has the SSL some AGC-like thing before the harmonics are made ?
Regards,
Peter
Kev
Well-known member
also at Doug's site
http://www.dself.dsl.pipex.com/ampins/discrete/ef.htm
THE SIMPLE EMITTER-FOLLOWER.
from the site
http://www.dself.dsl.pipex.com/ampins/discrete/ef.htm
THE SIMPLE EMITTER-FOLLOWER.
from the site
Mendelt
Well-known member
[quote author="mediatechnology"]Second order "distortion," if I'm not mistaken, introduces a DC component. You see it as waveform asymmetry. Like a saxaphone on a 'scope with its' strong octave resonances. [/quote]
Pure second order distortion will not add a dc component but creating second order distortion with a multiplier circuit will add a dc offset as a "side effect".
When you multiply a signal with itself you're actually amplitude modulating the signal with itself. Amplitude modulation will cause extra sidebands to appear. You can easilly calculate the frequencies of those sidebands. One is sig1 + sig2 and one is sig1 - sig2. When sig1 = sig2 you're creating an extra signal at 2 x sig1 (the second harmonic) and at 0 (the dc offset)
I hope i'm still making sense here
Pure second order distortion will not add a dc component but creating second order distortion with a multiplier circuit will add a dc offset as a "side effect".
When you multiply a signal with itself you're actually amplitude modulating the signal with itself. Amplitude modulation will cause extra sidebands to appear. You can easilly calculate the frequencies of those sidebands. One is sig1 + sig2 and one is sig1 - sig2. When sig1 = sig2 you're creating an extra signal at 2 x sig1 (the second harmonic) and at 0 (the dc offset)
I hope i'm still making sense here
Svart
Well-known member
I don't recall which post I proposed this very same idea some time ago but it was met with almost a dismissive nature. Unfortunately now it's only the clone mentality because $$L has actually done it.
I also mentioned that I think it should be paralleled with the audio chain and give purely the harmonics and/or dist. that we want, not a wet signal but the effect itself.
AnalogPackrat
Well-known member
This is all very interesting. I've spent a fair amount of time learning about harmonic content from the synthesis perspective. I think we've gotten confused here with the "symmetry" and DC offset terminology. DC offset, per se, does not change the harmonic content of a signal. It may be true that passing such a signal through different amplifier topologies produces different results, but that's not what I'm getting at.
If you look at a typical time domain representation of a signal, you can see that there are two axes of symmetry--time and magnitude. I believe there is a simple way to determine if a signal contains even, odd, or both harmonics. A square wave is symmetric in both time and magnitude. You can invert a square wave and shift it half a cycle in time (or invert it in mag and mirror it in time) and it's the same wave. The same is true of a triangle. These contain no even harmonics--only odd. Of course you can also do this with a sine wave (which has no harmonics, of course).
Now take our old friend the sawtooth wave. You have to invert its magnitude and mirror it in time (in half cycle pieces) to get it to match itself. It is not possible to do away with the mirroring. The sawtooth contains only even harmonics.
Look at a pulse waveform (square wave with something other than 50% duty cycle). It has no symmetry. It contains both even and odd harmonics.
My Fourier Transform math is too rusty, but I imagine these time domain symmetries and manipulations have interesting properties in the frequency domain.
Wikipedia has some interesting content along these lines, though there is no pulse train in the examples.
A P
If you look at a typical time domain representation of a signal, you can see that there are two axes of symmetry--time and magnitude. I believe there is a simple way to determine if a signal contains even, odd, or both harmonics. A square wave is symmetric in both time and magnitude. You can invert a square wave and shift it half a cycle in time (or invert it in mag and mirror it in time) and it's the same wave. The same is true of a triangle. These contain no even harmonics--only odd. Of course you can also do this with a sine wave (which has no harmonics, of course).
Now take our old friend the sawtooth wave. You have to invert its magnitude and mirror it in time (in half cycle pieces) to get it to match itself. It is not possible to do away with the mirroring. The sawtooth contains only even harmonics.
Look at a pulse waveform (square wave with something other than 50% duty cycle). It has no symmetry. It contains both even and odd harmonics.
My Fourier Transform math is too rusty, but I imagine these time domain symmetries and manipulations have interesting properties in the frequency domain.
Wikipedia has some interesting content along these lines, though there is no pulse train in the examples.
A P
Svart
Well-known member
reminds me of how they modulate digital onto RF carriers for broadband..
Kev
Well-known member
[quote author="AnalogPackrat"]If you look at a typical time domain representation of a signal, you can see that there are two axes of symmetry--time and magnitude
Now take our old friend the sawtooth wave. You have to invert its magnitude and mirror it in time (in half cycle pieces) to get it to match itself. It is not possible to do away with the mirroring. The sawtooth contains only even harmonics.[/quote]
thanks AP
I too wanted to look at the waveform and look at possible techniques from the synthesiser world
but I was trying to move away from continuous repetition of waveform (cycles) and look closer to the rate of change of the amplitude
I think the non-linearity we might want is continuos and without instant change
Wayne is right that this would be level dependant and initially I was thinking that would probably be the implementation of the SSL
which was the point of the thread ... but I'm happy to widen it and let it be something new and not a clone.
... even though I think it is very hard to do anything with a simple circuit and handful of the basic components that hasn't already been done before. The stomp box world hasn't had a new idea in my lifetime ... at least I don't think so. I'm sure you could patent a resistor and pass it off at a new circuit idea ... sorry ... pet peeve
AP
"the saw tooth has only evens"
I thought both saw and reverse saw contained all integer harmonics ??
they sound the same anyway
but can help hightlight a circuit that is asymmetrically clips ... sory I'm back on clip and soft clip again
perhaps I'm not getting the significance of the mirror in half cycle pieces.
...
wouldn't that make it more like a triangle wave ... lots of odd ?
Whatever the process might turn out to be it probably is working directly and in sync with the waveform itself so Wayne's cool circuit leaves me thinking envelope because of the RMS detector rather than waveform.
just for fun
the above doesn't show a crossover distortion ... or step thing
but then as Wayne pointed out above (earlier page)
we probably don't want that
not having seen or heard the SSL product
but a couple of questions before I get into more trouble
how many controls does it have ?
odd - even - on - off ?
mix - wet/dry ?
is it level dependant ?
etc
or simply straight out of the recorder as pos or neg polarity ?
this was very much the content of that thread I started ... long ago in the wayback machine
it was also the subject of one of the Group DIY updates.
Now take our old friend the sawtooth wave. You have to invert its magnitude and mirror it in time (in half cycle pieces) to get it to match itself. It is not possible to do away with the mirroring. The sawtooth contains only even harmonics.[/quote]
thanks AP
I too wanted to look at the waveform and look at possible techniques from the synthesiser world
but I was trying to move away from continuous repetition of waveform (cycles) and look closer to the rate of change of the amplitude
I think the non-linearity we might want is continuos and without instant change
Wayne is right that this would be level dependant and initially I was thinking that would probably be the implementation of the SSL
which was the point of the thread ... but I'm happy to widen it and let it be something new and not a clone.
... even though I think it is very hard to do anything with a simple circuit and handful of the basic components that hasn't already been done before. The stomp box world hasn't had a new idea in my lifetime ... at least I don't think so. I'm sure you could patent a resistor and pass it off at a new circuit idea ... sorry ... pet peeve
AP
"the saw tooth has only evens"
I thought both saw and reverse saw contained all integer harmonics ??
they sound the same anyway
but can help hightlight a circuit that is asymmetrically clips ... sory I'm back on clip and soft clip again
perhaps I'm not getting the significance of the mirror in half cycle pieces.
...
wouldn't that make it more like a triangle wave ... lots of odd ?
Whatever the process might turn out to be it probably is working directly and in sync with the waveform itself so Wayne's cool circuit leaves me thinking envelope because of the RMS detector rather than waveform.
just for fun
the above doesn't show a crossover distortion ... or step thing
but then as Wayne pointed out above (earlier page)
we probably don't want that
not having seen or heard the SSL product
but a couple of questions before I get into more trouble
how many controls does it have ?
odd - even - on - off ?
mix - wet/dry ?
is it level dependant ?
etc
once they have passed through a circuit that introduced distortion ?
or simply straight out of the recorder as pos or neg polarity ?
this was very much the content of that thread I started ... long ago in the wayback machine
it was also the subject of one of the Group DIY updates.
Similar threads
Latest posts
-
-
-
-
-
-
-
-
For Sale Group buy for Grayhill 71 Series 3 deck, 1 pole switches
- Latest: Tungstengruvsten