For Guitar
So back on the "tube" sound, for guitar at least I run across many sites that provide an explanation. For example: http://www.paia.com/tubesnd.htm
My issue with these explanations are:
1. It's always a simple sine wave. I don't know of any musical instrument, besides a synth, that puts out just a sine wave. It's always more complex.
2. It's always about levels at clipping.
Guitar amps due provide a "master" volume to allow the preamp (class-a) to drive into clipping.
FET amps can be made to produce the same sine wave behavior, but still musicians don't regard them as equal to tube.
So why is a tube better than SS in the area for guitar? Why is it that SS gets accused of being harsh?
I don't know, but maybe some ideas. First it may be hype, Referencing a study done in 1981, it's hard for most people to really tell SS from Tube guitar amps:
http://milbert.com/articles/TvsT/tvtiega.html
I'd like to see new blind tests performed today.
I suspect that people can tell now, and that guitar tube amps are purposely exaggerated to provide a tube sound. I think it has to do with frequency response.
I'd love to see frequency response, and spectrum charts from an amp as it approaches and goes into clipping. This might be the key. I've searched but am unable to find any references. I currently do not own a tube amp to measure.
If you take a read of:
http://www.rane.com/pdf/old/note128.pdf
http://www.st-andrews.ac.uk/~jcgl/Scots_Guide/audio/clipping/page3.html
I conclude from these, if an amp is stable at clipping, that clipping itself doesn't add that much harmonic content. The input signal itself provides more more high frequency content. This HF content is below clipping and can increase in DB up until it clips too. For a good clean amp, you're going to hear more and more of the HF input signal in the output while the bass compresses. This unbalanced HF content sounds "harsh".
I suspect that a tube guitar amp somehow rolls off the HF content at clipping. I don't know how, Maybe it has to do with saturation, under-saturation of the tube and it's ability to pass HF content at these ends. Maybe it has to do with some kind of capacitance. In any case, I think that for tubes, the HF doesn't continue to grow as the LF compresses.
So part of a new blind test might be to let the lows clip for a SS amp, but when it does, start to compress the HF content to keep it in balance with the LF to some reasonable level.
Just some thoughts.
So back on the "tube" sound, for guitar at least I run across many sites that provide an explanation. For example: http://www.paia.com/tubesnd.htm
My issue with these explanations are:
1. It's always a simple sine wave. I don't know of any musical instrument, besides a synth, that puts out just a sine wave. It's always more complex.
2. It's always about levels at clipping.
Guitar amps due provide a "master" volume to allow the preamp (class-a) to drive into clipping.
FET amps can be made to produce the same sine wave behavior, but still musicians don't regard them as equal to tube.
So why is a tube better than SS in the area for guitar? Why is it that SS gets accused of being harsh?
I don't know, but maybe some ideas. First it may be hype, Referencing a study done in 1981, it's hard for most people to really tell SS from Tube guitar amps:
http://milbert.com/articles/TvsT/tvtiega.html
I'd like to see new blind tests performed today.
I suspect that people can tell now, and that guitar tube amps are purposely exaggerated to provide a tube sound. I think it has to do with frequency response.
I'd love to see frequency response, and spectrum charts from an amp as it approaches and goes into clipping. This might be the key. I've searched but am unable to find any references. I currently do not own a tube amp to measure.
If you take a read of:
http://www.rane.com/pdf/old/note128.pdf
http://www.st-andrews.ac.uk/~jcgl/Scots_Guide/audio/clipping/page3.html
I conclude from these, if an amp is stable at clipping, that clipping itself doesn't add that much harmonic content. The input signal itself provides more more high frequency content. This HF content is below clipping and can increase in DB up until it clips too. For a good clean amp, you're going to hear more and more of the HF input signal in the output while the bass compresses. This unbalanced HF content sounds "harsh".
I suspect that a tube guitar amp somehow rolls off the HF content at clipping. I don't know how, Maybe it has to do with saturation, under-saturation of the tube and it's ability to pass HF content at these ends. Maybe it has to do with some kind of capacitance. In any case, I think that for tubes, the HF doesn't continue to grow as the LF compresses.
So part of a new blind test might be to let the lows clip for a SS amp, but when it does, start to compress the HF content to keep it in balance with the LF to some reasonable level.
Just some thoughts.
Re: For Guitar
The way valve amplifiers work mean they are more frequency limited, in particular the output transformer limits both low and high frequencies.
But the classic reason is the type of distortion, valves tend to give even harmonic distortion, and the HF roll-off of the transformer smooths this further. But solidstate amps give odd harmonic distortion which sounds far harsher, and there's no HF roll-off as you don't have a transformer.
You can hear objectionable distortion at quite low percentages with transistors, but with valves even distortion well in double figures isn't a problem.
Transistor amps are only harsh if you clip them, and guitarists like it loud (too loud usually!) - once you start a transistor amp clipping it really does sound bad!.
Like yourself, I'd also love to see full graphs from a valve amp!.
raintalk said:
The way valve amplifiers work mean they are more frequency limited, in particular the output transformer limits both low and high frequencies.
But the classic reason is the type of distortion, valves tend to give even harmonic distortion, and the HF roll-off of the transformer smooths this further. But solidstate amps give odd harmonic distortion which sounds far harsher, and there's no HF roll-off as you don't have a transformer.
You can hear objectionable distortion at quite low percentages with transistors, but with valves even distortion well in double figures isn't a problem.
Transistor amps are only harsh if you clip them, and guitarists like it loud (too loud usually!) - once you start a transistor amp clipping it really does sound bad!.
Like yourself, I'd also love to see full graphs from a valve amp!.
I can also find alternative views on the "odd" verses "even" harmonics for tube verses SS:
http://www.pritchardamps.com/pritchardamps/tech_talk/audio_myths.htm#1
I would love to see some spectral analysis done as an amp approaches and goes into clipping.
Even class-d amps get lumped into the SS argument. The output devices are clipped all the time. Yet a filter is added to roll this out.
http://www.pritchardamps.com/pritchardamps/tech_talk/audio_myths.htm#1
I would love to see some spectral analysis done as an amp approaches and goes into clipping.
Even class-d amps get lumped into the SS argument. The output devices are clipped all the time. Yet a filter is added to roll this out.
raintalk said:
So would I, but also with a valve amp heavily over driven, as they are commonly used.
But the audio ISN'T clipped at all, it's similar to an FM radio signal, where the carrier is heavily clipped, but this doesn't affect the audio modulation.
In class-d it's also wrong to say the output devices are ever clipped, clipping assumes the incoming signal is being changed (by clipping the tops and bottoms), but class-d output devices don't do this, they never receive a signal that could be clipped, they are just switches, not amplifiers.
Re: Re: For Guitar
It isn't pretty. At least for the single-ended HiFi amps I've measured, it's a very slowly decreasing string of high level harmonics well past the upper limit of my sound card. BTW, I'ld caution against drawing too many conclusions about pre-clip harmonic content and bandwidth in general from the device used. My current project amp, a KT100 SE hifi amp, is about +5 dB at 80 kHz at full power. By using second harmonic cancellation between stages I've simulated single-ended circuits in which odd harmonics are dominant. My suspicion is topology trumps most everything else.
Nigel Goodwin said:
It isn't pretty. At least for the single-ended HiFi amps I've measured, it's a very slowly decreasing string of high level harmonics well past the upper limit of my sound card. BTW, I'ld caution against drawing too many conclusions about pre-clip harmonic content and bandwidth in general from the device used. My current project amp, a KT100 SE hifi amp, is about +5 dB at 80 kHz at full power. By using second harmonic cancellation between stages I've simulated single-ended circuits in which odd harmonics are dominant. My suspicion is topology trumps most everything else.
Suffice it to say that guitar amps are whole different beast... nothing is overkill in a guitar amp... rather underkill might be a better synopsis.
You WANT:
A lossy power supply that sags.
A output transformer that saturates and adds a whole series of odd harmonics... usually on top of the lower frequencies of the source.
The "soft cliiping" characteristic of tubes... because they are operated out-of-bounds.
You WANT:
A lossy power supply that sags.
A output transformer that saturates and adds a whole series of odd harmonics... usually on top of the lower frequencies of the source.
The "soft cliiping" characteristic of tubes... because they are operated out-of-bounds.
rdf,
The big boys are almost always push-pulls. The little guys are often single ended. Curiously the little single ended guys sometimes have the "best sound"... crunch... grit.
Ping the memeber TROUT. He has some recordings posted somewhere of some little SE Fender clones he has built. He's also a helluva guitar player...
The big boys are almost always push-pulls. The little guys are often single ended. Curiously the little single ended guys sometimes have the "best sound"... crunch... grit.
Ping the memeber TROUT. He has some recordings posted somewhere of some little SE Fender clones he has built. He's also a helluva guitar player...
rdf said:
Most are push-pull, you can't get enough power single ended for anything except a home practice amp - which you can use for studio recording as well.
Easy things to mend though, valve amps are so simple (even crude!), but these days many places can't repair them.
A friend of mine was given a fairly modern Fender guitar amp, it had been taken to five 'so called' service departments, all of which said it couldn't be repaired. So my friend was given it free!.
Before he even bought it to me I told him what was probably wrong, the anode loads of the triodes - and true enough, so it was - a simple, common for 50 years fault, and modern 'techs' couldn't repair it!.
This particular kid was competing for the world's #1 Lennon Fan Award. On minimum wage he had custom guitar built identical to one of Lennon's early pieces (it was almost twenty years ago now) and wanted the Vox because apparently it's what the Beatles used in early tours. He would mic it and run it though larger stacks at gigs.
Re: Re: For Guitar
I would think it's the opposite. A musician I know says the digital stuff used in the studios nowadays are so good there's little need for tubes and whatnots. He was in particular referring to limiters and such.
Guitarists are if possible even bigger hams than audiophiles. If Jimi or some other guitar god played it, they will to without asking questions.
I'm not saying tubes aren't better for guitar amps. I wouldn't know. I'm saying it doesn't really matter one way or the other.
raintalk said:
I would think it's the opposite. A musician I know says the digital stuff used in the studios nowadays are so good there's little need for tubes and whatnots. He was in particular referring to limiters and such.
Guitarists are if possible even bigger hams than audiophiles. If Jimi or some other guitar god played it, they will to without asking questions.
I'm not saying tubes aren't better for guitar amps. I wouldn't know. I'm saying it doesn't really matter one way or the other.
Re: For Guitar
A few further comments:
Yes, it is often stated that we listen to music, not sine waves - a logical statement if one looks at the complex waveform representing a full orchestra. But one must remember that any periodic signal, irrespective of its complexity, can be dissolved into sine-waves of 2x, 3x, 4x, etc. the fundamental frequency, each of a specific amplitude. (This is the well-known Fourier Theorem.) In audio it means that if one tests with sine waves up to say 20 KHz, the full capability of an amp is defined - difficult to picture, but it is fact.
A further point is that (regrettably) a large number of ss amplifiers use a quite large amount of feedback, sometimes to make a mediocre design kosher, I am sorry to say. (Larger amounts of stable feedback is possible in ss amps than in tube amps.) When such a device is overloaded (thus destroying linear operation) the result is mostly a real horror show, far more so than with tubes. I have done spectrum analyses on a number of ss amps and can confirm a wide variety of distortion characteristics. You don't want to know of some.
I once helped a friend with a guitar amp design, trying to give him the sound that he wanted. I was grateful to have been successful in the end, but the result was a moderately weird circuit. I guess one could say that to get the sound he desires, a guitarist needs a particular kind of amplifier mostly not hi-fi, but when he wants to replay a guitar recording, then he will use a hi-fi amplifier to give the exact original sound quality.
Regards all.
A few further comments:
raintalk said:
Yes, it is often stated that we listen to music, not sine waves - a logical statement if one looks at the complex waveform representing a full orchestra. But one must remember that any periodic signal, irrespective of its complexity, can be dissolved into sine-waves of 2x, 3x, 4x, etc. the fundamental frequency, each of a specific amplitude. (This is the well-known Fourier Theorem.) In audio it means that if one tests with sine waves up to say 20 KHz, the full capability of an amp is defined - difficult to picture, but it is fact.
A further point is that (regrettably) a large number of ss amplifiers use a quite large amount of feedback, sometimes to make a mediocre design kosher, I am sorry to say. (Larger amounts of stable feedback is possible in ss amps than in tube amps.) When such a device is overloaded (thus destroying linear operation) the result is mostly a real horror show, far more so than with tubes. I have done spectrum analyses on a number of ss amps and can confirm a wide variety of distortion characteristics. You don't want to know of some.
I once helped a friend with a guitar amp design, trying to give him the sound that he wanted. I was grateful to have been successful in the end, but the result was a moderately weird circuit. I guess one could say that to get the sound he desires, a guitarist needs a particular kind of amplifier mostly not hi-fi, but when he wants to replay a guitar recording, then he will use a hi-fi amplifier to give the exact original sound quality.
Regards all.
Re: Re: For Guitar
Music is not very periodic... Fourier analysis can tell us some things but not all....
http://www.t-linespeakers.org/oddsends/mrFourier.html
dave
Johan Potgieter said:
Music is not very periodic... Fourier analysis can tell us some things but not all....
http://www.t-linespeakers.org/oddsends/mrFourier.html
dave
For those of us that are curious we experiment. I connected my guitars directly up to the FFT analyzer in this PC (no amp). There are a few things to learn here. The first is that the harmonic spectra from any guitar is not constant. Every guitar is different and each pickup on the guitar is different, but the common thread is that when you whack a string there is a rich harmonic spectra extending up into the mid Khz region. The higher order harmonics die off quickly and then the whole spectrum decays down to a large fundamental tone with some second and third after a few seconds. The pickup closest to the bridge has the highest harmonic content. Some guitars have a switch that allow operating two pickups at the same time out of phase. This gives a rich harmonic content with a lower fundamental. What does this have to do with the tube vs SS debate. Plenty.
A well designed SS guitar amp has a relatively constant gain VS drive level, up until it clips. In a tube guitar amp (especially an older one with an under sized power transformer and a tube rectifier) this is not constant. When you crank a tube amp all the way up and blast it with a guitar signal the stored energy in the power supply caps (too small by HiFi standards) provides the initial attack, then the power supply sags quickly, the amp gain drops. As the guitar signal decays the demand on the power supply eases and the voltage rises raising the amp gain. This effect causes the amp to exhibit a time (and signal level) dependent compression and distortion that is real hard to duplicate in a SS amp, although modern DSP effects come close.
A skilled guitar player can use this to his advantage to create his particular "sound". It is possible to use some positive acoustical feedback (by standing in front of the speaker cabinet) to add energy to the guitar strings and generate an endless sustain. This is far easier to control if you are using a tube amp.
Notice that I mentioned the tube rectifier, that is how this thread got started. The voltage drop across a tube rectifier is not linear and can not be approximated by a diode and a resistor, at least not in a guitar amp. Some guitar amps have a switch right on the front panel to switch between tube and SS rectification. The Mesa Boogie Dual Rectifier models come to mind. You can flip the switch as you are playing and hear the difference immediately. I have built amps with this switch and there is definitely a difference when the amp is operating outside the linear region. This is true even in a SE amp as the current is no longer constant when the amp in no longer in class A. The tube rectifier selection distorts easier, has lower total power, and has the sustain qualities mentioned above. This is usually used with lead guitars. The SS rectifier has higher total power and an overall cleaner sound. It is also interesting that a choke input filter is almost never used in a guitar amp.
These comments apply only to guitar amps. A guitar amp is not designed to HiFi standards. Often the power supply is undersized. In these applications there is an audible difference.
In the HiFi world it is possible to design a good power supply using either technology. It is often more difficult to design a clean power supply using SS rectifiers since the peak currents are much higher and occur for a shorter period of time. Even if the diodes exhibit no reverse recovery effects, "hash" can be created by as the diodes abruptly turn off by the stray inductances and capacitances in the circuit.
A well designed SS guitar amp has a relatively constant gain VS drive level, up until it clips. In a tube guitar amp (especially an older one with an under sized power transformer and a tube rectifier) this is not constant. When you crank a tube amp all the way up and blast it with a guitar signal the stored energy in the power supply caps (too small by HiFi standards) provides the initial attack, then the power supply sags quickly, the amp gain drops. As the guitar signal decays the demand on the power supply eases and the voltage rises raising the amp gain. This effect causes the amp to exhibit a time (and signal level) dependent compression and distortion that is real hard to duplicate in a SS amp, although modern DSP effects come close.
A skilled guitar player can use this to his advantage to create his particular "sound". It is possible to use some positive acoustical feedback (by standing in front of the speaker cabinet) to add energy to the guitar strings and generate an endless sustain. This is far easier to control if you are using a tube amp.
Notice that I mentioned the tube rectifier, that is how this thread got started. The voltage drop across a tube rectifier is not linear and can not be approximated by a diode and a resistor, at least not in a guitar amp. Some guitar amps have a switch right on the front panel to switch between tube and SS rectification. The Mesa Boogie Dual Rectifier models come to mind. You can flip the switch as you are playing and hear the difference immediately. I have built amps with this switch and there is definitely a difference when the amp is operating outside the linear region. This is true even in a SE amp as the current is no longer constant when the amp in no longer in class A. The tube rectifier selection distorts easier, has lower total power, and has the sustain qualities mentioned above. This is usually used with lead guitars. The SS rectifier has higher total power and an overall cleaner sound. It is also interesting that a choke input filter is almost never used in a guitar amp.
These comments apply only to guitar amps. A guitar amp is not designed to HiFi standards. Often the power supply is undersized. In these applications there is an audible difference.
In the HiFi world it is possible to design a good power supply using either technology. It is often more difficult to design a clean power supply using SS rectifiers since the peak currents are much higher and occur for a shorter period of time. Even if the diodes exhibit no reverse recovery effects, "hash" can be created by as the diodes abruptly turn off by the stray inductances and capacitances in the circuit.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.