My understanding of amplifiers

[PaulyT]

My father, a EE from the 1950s who designed and built tube amplifers (mostly RF amps) for the military for decades, would always tell me that it isn't an amp's ability to drive high voltage into 100hz (low frequencies) that mattered. He was always asking how well and amp could drive high voltage into 20,000Hz, because that was ultimately a much harder job for the power supply. Why? Because if you calculate the area of a sine wave and compare 1 sec of 100hz to 1 sec of 20,000Hz, the 20,000Hz sine wave is significantly more area - thus more power for given amount of time.

Interesting, never heard of power being equated to area under the curve. Let's see... the area under a half-period of sine wave (f(t) = A*sin(w*t)) for amplitude A and frequency w, where we start at w*t=0 and end at w*t=pi at the half-period, is

:text-link:

-A*cos(pi)/w - (-A*cos(0)/w) = -A*(-1)/w - (-A*1/w) = 2*A/w

A full period of the wave is 2*pi, and (by definition) in 1 second for frequency w there will be w periods. So the total area for 1s of a sine wave of amplitude A and frequency w is

(2*A/w) * 2 * w = 4A

So, the area under a sine wave for a given time period is not dependent on frequency. So there must be something else to your father's assertion...

 

[PaulyT]

Thinking more about power and senstivity...

Let's say two transducers (speakers or headphones, it doesn't matter), have the same sensitivity specified in Watts, but different impedance. That means that when they are receiving the same amount of power (not current, not voltage, but power), they will achieve the same SPL - e.g. have the same apparent loudness.

But remember Ohm's law and Joule's law, which say that P = V^2 / R. Thus if the power (P) is constant, but the impedance (R) is lower, the voltage (V) must also be lower. So, a lower impedance transducer will have lower voltage across its terminals at the same SPL - meaning the amp (or input signal) will need to be turned down, because remember I said the voltage at the amp output is dependent only on input signal and amp gain levels, nothing else (assuming the amp is not clipping).

However, Joule's law is P = I * V. So again if the power (P) is constant, but the voltage (V) is lower, then the current (I) must be higher.

Taken together, this means that if your transducers with the same sensitivity-in-Watts are producing the same SPL, the volume on the input/amp will be turned down on the low impedance transducer compared to the high impedance transducer. But, here's the kicker, despite the fact that the amp is turned down, the current in the low impedance transducer is greater!

This is the counter-intuitive part, that a lower volume setting on the input/amp can actually equate to a higher current. In other words, higher volume setting does NOT mean higher current when you're talking about transducers of different impedance.

So for your high impedance headphones, you may need to turn the amp volume up, yes, but the amp is actually producing less current than it would at the same SPL with low impedance headphones. (Again, assuming same/similar sensitivity-in-Watts.)

Sorry to repeat this over and over, but this is important (to me anyway ) because it explains the perception that "high impedance headphones are harder to drive" - because you have to turn up the amp. Whereas, in another sense, "low impedance headphones are harder to drive" - because they require more current. But because the current is not (simply) a function of the volume control on the amp, you wouldn't necessarily realize this until your amp starts to clip because it can't produce enough current.

 

[heeman]

Earlier you discussed the relationship between Power, Voltage and Current.

Power = E x I

Think of a simple constant load, and a typical 120VAC Source; just for this example.

120W = 120V x 1A

What happens to the current when the voltage becomes less than 120V, it goes up to maintain the load.

120W = 110V X 1.09A 120W = 100V X 1.2A 120W = 80V X 1.5A

One note is to understand that increased current results in increased heat in the load.

The reason for the example above is to show what happens to the current when the voltage continues to drop. Very basic, however necessary to understand.

How does this apply to an amplifier and a speaker. The principle is the same, however much more complex. We have a reactive load, the speaker (that changes it's impedance at different frequencies) and a variable supply of frequencies from the source; Amplifier.

But in simple terms, if the amplifier can not supply the constent voltage, the current increases, however if the amplifier can not supply the required current, we have a source that can not support the load.

Therefore it is necessary to have an amplifier that has a suitable current rating not power rating to support the load during Dynamic Peaks. Steady state does not put the demands on an amplifier like Dynamic Peaks. Dynamic Peaks put instantanous demands on an amplifier that can cause inadequte voltage to the load.

 

[PaulyT]

Right, but it sounds to me like you're talking about input voltage in your equations (e.g. the voltage in the power cord to the wall), where I'm talking about output voltage (voltage at the output terminals of the amp)? Those are very different things... even though governed by the same laws, of course.

And there are two situations where increased current at the amp output is required: 1) if the load's (speaker's) impedance drops with frequency but you want to maintain constant steady-state SPL, or 2) there are transient peaks in the input voltage (the signal, e.g. from the preamp) that need to be translated into transient peaks in the output voltage (to make the speakers reproduce that loud sound). In either case, the amp needs to put more current in the output terminals, and that current has to come from somewhere - the power supply - and has to move through the various transformers etc. in the body of the amp. As the fellow from Parasound in your video said, ability to pass more current from the wall to the speaker requires physically larger transformers and other electrical components.

In other words, as I understand it, the definition of an amp clipping is, as you say, being unable to produce adequate current in order to maintain the output voltage at a constant multiple of the input voltage (input here being the signal from the preamp, not the power supply input from the wall). Mostly what I've been talking about are non-clipping situations where current varies with impedance.

 

[heeman]

Right, but it sounds to me like you're talking about input voltage in your equations (e.g. the voltage in the power cord to the wall), where I'm talking about output voltage (voltage at the output terminals of the amp)? Those are very different things... even though governed by the same laws, of course.

And there are two situations where increased current at the amp output is required: 1) if the load's (speaker's) impedance drops with frequency but you want to maintain constant steady-state SPL, or 2) there are transient peaks in the input voltage (the signal, e.g. from the preamp) that need to be translated into transient peaks in the output voltage (to make the speakers reproduce that loud sound). In either case, the amp needs to put more current in the output terminals, and that current has to come from somewhere - the power supply - and has to move through the various transformers etc. in the body of the amp. As the fellow from Parasound in your video said, ability to pass more current from the wall to the speaker requires physically larger transformers and other electrical components.

In other words, as I understand it, the definition of an amp clipping is, as you say, being unable to produce adequate current in order to maintain the output voltage at a constant multiple of the input voltage (input here being the signal from the preamp, not the power supply input from the wall). Mostly what I've been talking about are non-clipping situations where current varies with impedance.

I used the 120V as an example, however the principle is the same considering the output of the amplifier (source) as the Input to the Speaker (load). Except that the Output of the amplifier, is much more complex and the Load of the Speaker is much more complex, however the principle is the same.

Another anology is the fuel delivery system in your car. If you are just cruising along at 60 mph and the fuel delivery system is adequate, the engine runs fine and smooth, however if you punch the gas pedal to the floor (the audio signal dynamics) and the fuel delivery system can not supply sufficient fuel to the engine, then the engine (the speaker) will begin to stutter and not run properly because the peak demand is to much for the fuel delivery system (the amplifier).

 

[PaulyT]

Good analogy! And a better car (amplifier) will respond more precisely and rapidly to the position of your foot (input signal).

 

[heeman]

Right, but it sounds to me like you're talking about input voltage in your equations (e.g. the voltage in the power cord to the wall), where I'm talking about output voltage (voltage at the output terminals of the amp)? Those are very different things... even though governed by the same laws, of course.

And there are two situations where increased current at the amp output is required: 1) if the load's (speaker's) impedance drops with frequency but you want to maintain constant steady-state SPL, or 2) there are transient peaks in the input voltage (the signal, e.g. from the preamp) that need to be translated into transient peaks in the output voltage (to make the speakers reproduce that loud sound). In either case, the amp needs to put more current in the output terminals, and that current has to come from somewhere - the power supply - and has to move through the various transformers etc. in the body of the amp. As the fellow from Parasound in your video said, ability to pass more current from the wall to the speaker requires physically larger transformers and other electrical components.

In other words, as I understand it, the definition of an amp clipping is, as you say, being unable to produce adequate current in order to maintain the output voltage at a constant multiple of the input voltage (input here being the signal from the preamp, not the power supply input from the wall). Mostly what I've been talking about are non-clipping situations where current varies with impedance.

You can not have it at the output of the amp without it at the power supply section of the amp.

Non-clipping situations are similar, except may not be as noticable. Dynamic Response (Punch) is related to the ability of the amp to respond quickly and provide the proper voltage during that peak demand and that is accomplished with sufficent current. The result puts more strain on the amp when the system is at a higher output level.


However take a hard hit on a snare drum or a bass drum, the amp must be able to handle the instantanous peak of the signal with respect to the speaker load (at the fequency) without degrading the audio impact (sound from the speaker). The input signal and complexity of the hit on the drum is the same regardless if the volume is low or high, however the amps ability is more difficult if the volume is high.

(This is not easy with typed words)

 

[Dennie]

Right, but it sounds to me like you're talking about input voltage in your equations (e.g. the voltage in the power cord to the wall), where I'm talking about output voltage (voltage at the output terminals of the amp)? Those are very different things... even though governed by the same laws, of course.

And there are two situations where increased current at the amp output is required: 1) if the load's (speaker's) impedance drops with frequency but you want to maintain constant steady-state SPL, or 2) there are transient peaks in the input voltage (the signal, e.g. from the preamp) that need to be translated into transient peaks in the output voltage (to make the speakers reproduce that loud sound). In either case, the amp needs to put more current in the output terminals, and that current has to come from somewhere - the power supply - and has to move through the various transformers etc. in the body of the amp. As the fellow from Parasound in your video said, ability to pass more current from the wall to the speaker requires physically larger transformers and other electrical components.

In other words, as I understand it, the definition of an amp clipping is, as you say, being unable to produce adequate current in order to maintain the output voltage at a constant multiple of the input voltage (input here being the signal from the preamp, not the power supply input from the wall). Mostly what I've been talking about are non-clipping situations where current varies with impedance.

You can not have it at the output of the amp without it at the power supply section of the amp.

Non-clipping situations are similar, except may not be as noticable. Dynamic Response (Punch) is related to the ability of the amp to respond quickly and provide the proper voltage during that peak demand and that is accomplished with sufficent current. The result puts more strain on the amp when the system is at a higher output level.


However take a hard hit on a snare drum or a bass drum, the amp must be able to handle the instantanous peak of the signal with respect to the speaker load (at the fequency) without degrading the audio impact (sound from the speaker). The input signal and complexity of the hit on the drum is the same regardless if the volume is low or high, however the amps ability is more difficult if the volume is high.

(This is not easy with typed words)

I think you're doing a great job, Thanks Heeman!!! :handgestures-thumbup:


Dennie

 

[Zing]

(This is not easy with typed words)

I'd say you're doing a FABULOUS job!

Hat's off.

 

[PaulyT]

Heeman, we agree. Really, I think we do. It's just that you're talking mostly about dynamic response and headroom for transient peaks within a given set of components. I'm talking mostly about how power/current levels are affected by changes in load impedance in different components.

But certainly, the principles are the same. That drum/snare hit is a voltage spike in the input from the preamp, which the amp is supposed to translate into an equivalent but linearly larger voltage spike at the output terminals. Ohm's law (V = I*R) says that when impedance (R) is the same (e.g. you're not swapping speakers), and the voltage goes up, the current must go up. The current has to come from somewhere, and that's where the power supply comes in. And if the power supply cannot provide adequate current to achieve that voltage spike (because either the spike is too large and/or the steady-state level is already near the max), then the amp will clip. The dynamic response is how fast the voltage changes in the output terminals to match what's coming into the input from the preamp.

 

[PaulyT]

What, no love for the originator of the thread? :angry-tappingfoot: Ingrates.

 

[heeman]

The engineering and the theory of why amplifiers and speakers interact the way they do, is a subject that I think Paul is trying show here and is very complex.

I jumped in, to try and explain in terms that may be a little more understandable to some that are not as scientifically savvy.

I feel that I got a little carried away and appoligize if I Highjacked the thread!

Dennie and Zing.......thanks!

It's all yours Paul!!

 

[PaulyT]

No no, don't apologize, I want this to be a discussion. I just wanted to say that I'm not trying to disagree with you or prove that you're wrong.

 

[heeman]

Heeman, we agree. Really, I think we do. It's just that you're talking mostly about dynamic response and headroom for transient peaks within a given set of components. I'm talking mostly about how power/current levels are affected by changes in load impedance in different components. But certainly, the principles are the same. That drum/snare hit is a voltage spike in the input from the preamp, which the amp is supposed to translate into an equivalent but linearly larger voltage spike at the output terminals. Ohm's law (V = I*R) says that when impedance (R) is the same (e.g. you're not swapping speakers), and the voltage goes up, the current must go up. The current has to come from somewhere, and that's where the power supply comes in. And if the power supply cannot provide adequate current to achieve that voltage spike (because either the spike is too large and/or the steady-state level is already near the max), then the amp will clip. The dynamic response is how fast the voltage changes in the output terminals to match what's coming into the input from the preamp.

The load (speaker) impedance is directly related to the signal from the amp to the speaker. The speaker has no impedance if there is no signal. Yes, the speaker has a nominal Impedance Rating of 8 or 6 or 4 ohms, however that is just a basic rating at a specific frequency.

The signal from the amp to drive the speaker is a complex signal that correlates directly to how the load (speaker) impedence changes.

 

[heeman]

Heeman, we agree. Really, I think we do. It's just that you're talking mostly about dynamic response and headroom for transient peaks within a given set of components. I'm talking mostly about how power/current levels are affected by changes in load impedance in different components.

But certainly, the principles are the same. That drum/snare hit is a voltage spike in the input from the preamp, which the amp is supposed to translate into an equivalent but linearly larger voltage spike at the output terminals. Ohm's law (V = I*R) says that when impedance (R) is the same (e.g. you're not swapping speakers), and the voltage goes up, the current must go up. The current has to come from somewhere, and that's where the power supply comes in. And if the power supply cannot provide adequate current to achieve that voltage spike (because either the spike is too large and/or the steady-state level is already near the max), then the amp will clip. The dynamic response is how fast the voltage changes in the output terminals to match what's coming into the input from the preamp.

Yes, and that directly correlates to whether or not the Amp has enough, Balls, Punch, Current to rise fast enough to make the sound hit you in your face, or to be weak and wimpy sounding. Or like the instrument/band is in your room with you. If the amp does not have the balls, punch or current the sound coming from the speakers will not be realistic sounding.

Yes, we do agree!

 

[heeman]

No no, don't apologize, I want this to be a discussion. I just wanted to say that I'm not trying to disagree with you or prove that you're wrong.

:text-bravo: Likewise!

 

[PaulyT]

The load (speaker) impedance is directly related to the signal from the amp to the speaker. The speaker has no impedance if there is no signal. Yes, the speaker has a nominal Impedance Rating of 8 or 6 or 4 ohms, however that is just a basic rating at a specific frequency.

The signal from the amp to drive the speaker is a complex signal that correlates directly to how the load (speaker) impedence changes.

Yup, the impedance of a speaker is highly dependent on frequency and the spec rating is an approximate average. Some of the links in my posts above talk more about this, why speakers are like this, and what speaker designers do to compensate to some degree.

And of course, even to talk about a single static voltage at the output terminals is a simplification, even if there's only a single frequency; the nature of music and sound and hence electronic audio signals is that they are waves. An audio signal oscillates in a complex way between positive and negative voltage.

FYI that's why volt meters have "DC voltage" and "AC voltage" settings, they are not measured the same way, at least not in a time-average sense.

 

[Dennie]

What, no love for the originator of the thread? :angry-tappingfoot: Ingrates.

Fine...... :angry-tappingfoot:

Thanks for starting a very educational thread Pauly. You are a God!!! :bow-blue:


I forget we need to do this every couple of weeks or so! :doh:

You know "we" gave you a "title" right? For most that would be good enough!!! :angry-banghead:

..... :laughing-rolling:



Dennie (Joking of coarse )

 

[PaulyT]

:laughing-rolling: :laughing-rolling: :laughing-rolling:

That's better, thanks my subject - oops I mean, my friend. :happy-smileygiantred:

 

[Dennie]

:laughing-rolling: :laughing-rolling: :laughing-rolling:

That's better, thanks my subject - oops I mean, my friend. :happy-smileygiantred:

:handgestures-thumbup: ...... :bow-blue:





Dennie :dance: