It has been said many times that the ideal/perfect amp is a wire with gain. What would like to know is what would all the measured specs of a wire with gain be? I wonder if we could all agree on what perfection would be or if even this is a topic for endless debate?
Well.... A theoretically perfect amplifier would basically perform the mathematical function, Vout(t) = A*Vin(t).
It would add no noise. It would have no propagation delay. It would be completely linear regardless of the input voltage. In case of a voltage amplifier, it would have infinite input impedance so it doesn't load the source. And it would have zero output impedance so the load doesn't affect the amplifier.
In other words, the "wire with gain" requires materials that are superconductive at room temperature (or that you keep the devices in your "wire with gain" cooled to absolute zero) so you don't have any thermal noise. It also requires the speed of light to be infinite to avoid propagation delay.
Basically, you're asking for something that violates the laws of physics. Never mind that your theoretically perfect amplifier, presumably, would be used to drive a pair of speakers or headphones. Now you're dealing with all sorts of non-idealities. But, as you've already voided the laws of physics, perhaps you can design a speaker that has a perfect spherical radiation pattern at all frequencies, adds no distortion, has no moving mass, no back-EMF, etc.
Good luck and good weekend... 🙂
~Tom
It would add no noise. It would have no propagation delay. It would be completely linear regardless of the input voltage. In case of a voltage amplifier, it would have infinite input impedance so it doesn't load the source. And it would have zero output impedance so the load doesn't affect the amplifier.
In other words, the "wire with gain" requires materials that are superconductive at room temperature (or that you keep the devices in your "wire with gain" cooled to absolute zero) so you don't have any thermal noise. It also requires the speed of light to be infinite to avoid propagation delay.
Basically, you're asking for something that violates the laws of physics. Never mind that your theoretically perfect amplifier, presumably, would be used to drive a pair of speakers or headphones. Now you're dealing with all sorts of non-idealities. But, as you've already voided the laws of physics, perhaps you can design a speaker that has a perfect spherical radiation pattern at all frequencies, adds no distortion, has no moving mass, no back-EMF, etc.
Good luck and good weekend... 🙂
~Tom
When I say perfect, I mean WRT the laws of physics. What are the limits of what is theoretically possible? There are rules that govern what can be done, I just want to know the boundary that those rules dictate are.
I believe such amps do exist today.
It doesn't have to be perfect, it only needs to be transparant and linear to below audibility limits. Although those limits are up for debate, it is not impossible to agree that, for instance, -120dB for all distortion artefacts at all levels should do.
The issue with such amplifiers is that they sound, well, transparent. Clean. Add nothing, subtract nothing from the sound.
If you audition such an amp against, say a tube SE amp, lots of people will prefer the tube unit. Not because is is more 'perfect', on the contrary, but because it just sounds nice and preference is just that, preference.
So, you may well design or buy that straight wire with gain, but will you corner the consumer market with it? Don't think so.
jan
It doesn't have to be perfect, it only needs to be transparant and linear to below audibility limits. Although those limits are up for debate, it is not impossible to agree that, for instance, -120dB for all distortion artefacts at all levels should do.
The issue with such amplifiers is that they sound, well, transparent. Clean. Add nothing, subtract nothing from the sound.
If you audition such an amp against, say a tube SE amp, lots of people will prefer the tube unit. Not because is is more 'perfect', on the contrary, but because it just sounds nice and preference is just that, preference.
So, you may well design or buy that straight wire with gain, but will you corner the consumer market with it? Don't think so.
jan
Agreed. When auditioning for a new amp, I have ended up with Marantz every time so far. Why? It's the most compatible with my preferences. Is it absolutely neutral (other way of saying perfect)? Probably not. The theoretical perfect amp might not be the practical perfect amp.
But for the sake of this argument, I agree with tomchr. Every measurement of the the theoretical perfect amp would result in Vout(t) = A*Vin(t).
Because of my experience with a very badly measuring but wonderfully sounding NOS DAC, I think that in practical situations the perfect amp would effectively be perceived as Vout(t) = A*Vin(t).
But for the sake of this argument, I agree with tomchr. Every measurement of the the theoretical perfect amp would result in Vout(t) = A*Vin(t).
Because of my experience with a very badly measuring but wonderfully sounding NOS DAC, I think that in practical situations the perfect amp would effectively be perceived as Vout(t) = A*Vin(t).
Last edited:
I have OC'ed cpus in the past with liquid nitrogen and "pots" specially designed. For a time I also messed with phase change. If any one here builds an amp I'll gladly but it under liquid nitrogen and see what happens 🙂
"Theoretically possible"? How big is your budget?
The technology has come a long way. I used to design discrete amps, but my latest (non-tube) incarnation uses the LME49811 input and driver stage. According to my measurements it comes pretty darn close to an ideal amplifier when operated within its limits. Still, we're talking 50 V/us, a few hundred kHz bandwidth, THD below 0.002 % up to 100 W, etc. so a far cry from a theoretically ideal amp, but still pretty damn good. You can see my build here: http://www.diyaudio.com/forums/chip-amps/161337-yet-another-lme49811-std03-build.html. The final schematic is in Post #88.
But I think an even more relevant question to ask is, "what is needed". I've performed end-to-end measurements of my stereo system. I.e. measuring the frequency response and distortion vs frequency that is emitted by the speaker. It is very hard to find a speaker that will distort less than 1 % -- even at normal listening levels. At the full rated Xmax, the THD is often as high as 10 %.
The interesting thing is that when I measure end-to-end with the speakers driven by my tube amp I get results that are nearly identical to when I drive the speakers with the LME49811-based amp mentioned above. The tube amp's THD is on the order of 0.25 % at 1 W, 2 % at 10 W. In fact, I'd probably call it an 8 W amplifier as 10 W is way too far into clipping and grid current for my taste. I still like the sound of the tube amp better. The sound stage is phenomenal and the midrange tone is out of this world. But it's much further from an ideal amp than the LME49811-based amp. What gives?
~Tom
Last edited:
I thinck tomchr nailed it in that there is a profound difference between subjective appreciation and objective measurement.
As with every evaluation one must set conditions, criteria and methdology of testing. The very use of the term "perfect" must be defined as one could easliy use the term of perfection to encompass general (total) consensus over the perceived quality of an amp down to a THD measurements rebus sic stantibus.
The human brain is non linear, it does not allow for variable separation, it does not allow sensory input isolation and is influenced by a miriad of temporary conditions which inherently exclude its use for replicable and objective measurements.
As always it is a question of setting the conditions straight before getting into the "hard testing" part. Set the conditions in terms of THD, methdolody of testing, ambient variables down to room temperature and you have your experiment 🙂
I just think everything boils down to the point where a person is faced with two paths: the crossroad is between pure physics and pure listening experience. There is a time when both are one (a single path) and every improvement in one aspect yields results in the other but up to a point. One might call it the point of diminishing returns but 'd rather believe it is an ideal crossroad. Basically you have a silent testing room with perfect square wave respone on the right and a nice piece of music on the left. I know I will always choose the path on the left 🙂
But hey...a perfect amplifier would make for a killer scientific paper!
As with every evaluation one must set conditions, criteria and methdology of testing. The very use of the term "perfect" must be defined as one could easliy use the term of perfection to encompass general (total) consensus over the perceived quality of an amp down to a THD measurements rebus sic stantibus.
The human brain is non linear, it does not allow for variable separation, it does not allow sensory input isolation and is influenced by a miriad of temporary conditions which inherently exclude its use for replicable and objective measurements.
As always it is a question of setting the conditions straight before getting into the "hard testing" part. Set the conditions in terms of THD, methdolody of testing, ambient variables down to room temperature and you have your experiment 🙂
I just think everything boils down to the point where a person is faced with two paths: the crossroad is between pure physics and pure listening experience. There is a time when both are one (a single path) and every improvement in one aspect yields results in the other but up to a point. One might call it the point of diminishing returns but 'd rather believe it is an ideal crossroad. Basically you have a silent testing room with perfect square wave respone on the right and a nice piece of music on the left. I know I will always choose the path on the left 🙂
But hey...a perfect amplifier would make for a killer scientific paper!
"Mathematically put" auditive perception is based on "attractors". It can happen a piece of music stored on some old LP , badly recorded, distorted, resonances, even monaural,
takes you out of this world....all the fuses of mental control blow off...
basically auditve perception "loves imperfection" . But not each and every imperfection.
More technically however all possible amps fall in one of four categories
input: current output: current
this is a current controlled current source , input impedance zero output impedance "infinite" ( i.e. the current is independent of load impedance)
This is the "innate nature" of BJT. BJTs are current controlled current amplifiers.
This amp does not have any Vas and has no voltage amplification ( same as a current source can produce any voltage drop across a load without amplification of voltage)
This is the ideal amp with BJTs.
input: current , output: voltage
Some BJT amps are designed this way , at some point a conversion of current into voltage has to happen. That works best with a single BJT with bootstrap cap , and rather poor with a push pull Vas. Requires however stabilized supply or a separated
supply for predrivers.
input : voltage output: current
a rare species although conversion of voltage to current is somewhat easier then conversion of current to voltage.
input voltage output voltage
Strictly against the "nature" of BJT, nevertheless 90% of BJT amps are just this.
takes you out of this world....all the fuses of mental control blow off...
basically auditve perception "loves imperfection" . But not each and every imperfection.
More technically however all possible amps fall in one of four categories
input: current output: current
this is a current controlled current source , input impedance zero output impedance "infinite" ( i.e. the current is independent of load impedance)
This is the "innate nature" of BJT. BJTs are current controlled current amplifiers.
This amp does not have any Vas and has no voltage amplification ( same as a current source can produce any voltage drop across a load without amplification of voltage)
This is the ideal amp with BJTs.
input: current , output: voltage
Some BJT amps are designed this way , at some point a conversion of current into voltage has to happen. That works best with a single BJT with bootstrap cap , and rather poor with a push pull Vas. Requires however stabilized supply or a separated
supply for predrivers.
input : voltage output: current
a rare species although conversion of voltage to current is somewhat easier then conversion of current to voltage.
input voltage output voltage
Strictly against the "nature" of BJT, nevertheless 90% of BJT amps are just this.
I do not know enough to comment on the exact specifications of a audibly perfect amplifier. I will say however that all audio hardware must aim to reproduce the original signal as faithfully as possible. What benefits are there to a noisy tube amplifier when similar effects can now be reproduced accurately with modern software?
The best amplifier is the most faithful one, regardless of how flat it sounds. I believe that the same thing goes for speakers, headphones, etc. Modifications that cause your music to deviate from flat can always be made with software and can be changed at will.
The best amplifier is the most faithful one, regardless of how flat it sounds. I believe that the same thing goes for speakers, headphones, etc. Modifications that cause your music to deviate from flat can always be made with software and can be changed at will.
imho perfect amplifier needs perfect components. (superconductor p.c.b. - zero noise - perfect linearity)
but in reality mathematical model is only an approximation often appropriate but sometimes wrong because
it doesn't keep "in mind" parasitic effects.
but in reality mathematical model is only an approximation often appropriate but sometimes wrong because
it doesn't keep "in mind" parasitic effects.
- Status
- Not open for further replies.