Voltage source, current source.... huh?

[Coytee]

Can someone give me/us a lesson as to what is what and why this distinction is made as well as the benefits/drawbacks of one over the other?

I thought amps only delivered volts although I realize it takes 1 volt to push 1 amp through 1 ohm of resistance.... I don't even know if that applies []

What is a key difference that makes an amp a voltage source and what is a key difference that would make you call it a current source? Aren't they doing the same thing?

[mark1101]

An audio amp delivers power in watts. Power (watts) = I (currrent) x V (voltage). You can have high voltage and low current such as in a tube amp, or you can have high current and low voltage such as in a solid state amp. Different ways of creating the same result, watts.

[boom3]

Some amplifiers have voltage source stages that "swing" the amplified waveform to the limits (or close to) of the power supply rails, and then have "Current source stages" for sourcing the current required. I am not enough of an expert to say whether this approach has merit or not. I do know that, in general, "electrodynamic" direct radiator drivers require more current than electrostatic or horn-loaded drivers.

The impact on this is that a given amplifier may be able to "swing" the waveform pretty high in voltage, and therefore produce, in some speakers, a higher perceived degree of "loudness", whereas a different amp may not be very loud, but has plenty of current in reserve to pump the woofer(s) at low frequencies. I built an amp like that many years ago, it was limited to a swing of plus or minus 24 volts but it could source 5 amperes to back that up.

There are certain lab power supplies that are "voltage sources" in that they can self-adjust to keep voltage within a close tolerence as load varies, and "current sources" that do likewise with the current.

[Blvdre]

My simple-minded concept is that you ideally want your amp to be a perfect voltage source. In other words, when the load becomes difficult (say, a low impedance), a high amount of current is needed to keep the voltage from sagging. This is just a generalization though.

[Coytee]

So is it fair for me to generalize that if someone else is generalizing about a current amp, they might be referring to a solid state and if they generalize about a voltage amp then they are probably referring to a tube amp?

(I'm presuming there can be exceptions to each topology that would make that presumption wrong)

Do I presume to presume properly?

[billybobg]

I don't know about that generalizing. A voltage source with an output resistor is a current source.

Okay, I'm an idiot trying to answer this.

Ohms Law: V= I x R. Power out is P= VxI. Remember that the current source is a voltage source with a resistor in line. For the whole circuit then Ohms Law looks like V= Ix(R1 +RL) R1 equal to limiting resistor and RL equal to the load such as an 8Ohm speaker. If R1 is large (10 times or more) compared to RL, then mathmatically, RL can be ignored I=V/R. This means that the current delivered is independant of RL (as long as RL stays low).

For a voltage source, its sort of the opposite. In the real world the voltage source is low impedance (Internal.25 ohms or so) so the majority of the voltage drops across the load and therefore the load pretty much sees what voltage is being produced at the source. Think of a car battery and that the voltage is independant from the load.

Put both of these in the power equation P=VxI and sustituting V=IxR, you get P= (IxI)xR (Current Source), or substitute I=V/R and you get P=(VxV)/R (Voltage Source)

What does this mean? My impression is that it indicates how the output power is calculated. A current source can be limited to protect the load with the source resistor. With a voltage source, a diode could limit the voltage that is allowed across a load.

[djk]

Current source amplifiers are generally only used with full-range drivers, they boost the highs and lows (where the impedance peaks).

http://firstwatt.com/index.html

[WMcD]

It would be easier with diagrams, but I'll give it a shot.

An ideal voltage source is a theoretical construct. It is easy to draw one on paper. It means that no matter what the load, the voltage output is what is commanded. It could be a.c. or d.c. or even music. No matter how much current is demanded (because of a low ohmage load) the voltage stays as it should.

Roughly. An AA cell is not a good constant voltage source. As you connect more and more lightbulbs in parallel to it, the voltage sags. OTOH a big auto battery comes closer. This is because you can connect more and more light bulbs in parallel The voltage does not sag. Of course there are limits.

A arc welding machine also has the ability to produce high amounts of current. If an amp, like a Krell, has the ability to produce high current without voltage sag, it is called an arc welder. (Not reference to Sci-Fiction movie with Krell.)

Basically, amplifier or generator or battery outputs are modeled by an ideal voltage source feeding a resistor . The resistor is called the output impedance of the amp.

Ground -- voltage source -- output impedance of X ohms - - - output terminal - - | - - load resistor - - - ground. Please note, that pipe | defines the amp as we know it. Everything inside the amp is to the left. An amp with a low value of X (say 0.01 ohms) has a low output impedance or 0.01 ohms. This means that as the load resistor is made big or small, there is little voltage drop across that internal resistance. Output voltage at the terminal remains close to the value of the voltage source.

But if the output impedance is high, there is a voltage drop. Say it is 8 ohms. And the load is 8 ohms. Then the voltage at the output terminal is half of the voltage source. Voltage is divided across the two. The output voltage is sagging.

You will read that tube amps have a high output impedance (triodes without any feedback) and the output voltage sags as speaker impedance gets low. This is why. Of course these are simplifications.

One thing that helps an amp have an effective low output impedance is a feedback loop. The loop senses the droop and makes up for it by increasing the voltage at the output of the voltage source.

Constant current sources are also a theoretical construct. They also can be controlled by a loop sensor.

One thing you may be thinking of is something called a buffer. It is actually more of constant voltage source amp with zero gain. This can be an emitter follower or cathode follower. You use this when you have an amp with a high output impedance and which therefore will sag when a significant load is applied to it. The amp feeds a buffer between it an the load. The buffer has the ability to provide lots of current even though it does not increase the voltage. So, essentially, this is a current amplifier to the extent the amp with a high output imedance can not create a lot of current in an accurate way.

- - - - - - -

Another thing mentioned, sort of, is the Stasis type amplifier by Pass. (Another Sci-Fi reference - stasis field.) Here an accurate amplifier with low current capacity (meaning will get screwed up with a load that draws current) is connected to the output terminal. There is actually two current amplifers (maybe of low fidelity) connected too. There is a sensing mechanism which determines whether the load is trying to draw current from the accurate amp. If so, the current amplifiers come on and force the voltage at the terminal high (or whatever), just as high as the accurate amp is putting out. Since the output of the accurate amp is at voltage Y and the current amps have forced the voltage to Y, there is no current drawn from the output of the accurate amp. Its voltage doesn't sag. So we have the accurate amp controlling voltage output and the current amps doing the heavy lifting.

The above is like an automotive power steering unit. The steering wheel is connected to the tie rods, and the tie rods to the road wheel. And there is an engine driven hydraulic pump and hydraulic cylinders connected to the road wheel. But lets say there is a hydraulic valve with a little spring in parallel in the middle of the tie rod. The steering wheel tries to compress the tie rod to move the road wheel. But this actually compresses the spring and the small displacement opens the hydraulic valve. Now the hydraulic fluid and the hydraulic piston move the road wheel so that the tie rod is no longer in compression. The road wheel has rotated on the kingpin to the desired position by almost no force being applied by the steering wheel.

So this means the steering wheel controls the position of the road wheel and the hydraulic system does the work. Of course -- suppose the road wheel hits a bump which causes the tie rod to compress. Then the valve opens and aplies a countering force on the road wheel. The steering wheel never feels the force of the bump. This is the novocaine feel of power assist steering. The above is a simplification of the mechanical set up where the valves are in the steering box.

= = = = = = =

There is also talk of a constant current source being a voltage source with a big resistor. This is true in that a pretty good constant current source can be made this way. It is used in testing the impedance of a woofer to thereby determine T-S parameters. We are looking to graph the impedance of the woofer at different freqs.

Of course we know that the impedance of a device is R = V (across the load) / current through the load. We could do this with a volt meter and a current meter. But is would be nice to do it with only a volt meter. If that current thought the load is a constant at all freqs, then R is proportional to V. So we graph V and it is the same as R or a scaled value of R.

We use a good amp with an oscillator driving it. Then we put in a 1000 ohm resistor between the amp and the speaker.

Ground - - amp - - [1000 ohm resistor] - - - [speaker which varies between 4 ohms and 20 ohms] - - - ground.

We set the amp to put out 1 volt.

But what is the load in this loop of the 1000 ohm resistor and the speaker? The load is the sum of the 1000 ohms and the speaker which is 4 to 20 ohms, which fluctuates depending on freq. So the load is 1004 ohms to 1020 ohms. That is not much of a variation in total resistance. It means that the current through the speaker is always close to 1/1000 amps, or 1 mA. We have near constant current!

That is true even though the resistance of the speaker is varying a lot. (Without the 1000 ohm resistor, the current through the speaker will vary in accordance to its 4 to 20 ohm resistance at various freqs.) Note we've created a near constant current through the speaker with a 10 cent resistor.

So now we can see that R(speaker) = Voltage (across speaker) / 0.001. So, all we have to do is graph the easily measured voltage across the speaker as a function of freq and multiply by 1000. Now we have the magnitude of R versus freq.

That is all for tonight.

Wm McD

[Don Richard]

Here's what Nelson Pass says about the subject:

http://www.passdiy.com/pdf/cs-amps-speakers.pdf

[Guest " "]

This is a whole new universe, like the concept of dark matter.

If you use a current source amplifier, and need to use 2 way , 3 way, 4 way speaker systems, the crossovers in them have to be current sourced crossovers.

Traditional speakers use voltage soruce compatible crossovers.

whats the difference.

In voltage source compatible crossovers, for the most part, each crossover section is in parrallel to the other section. Each driver circut has its own seperate signal path which forms individual complete circuts. Once the crossover seperates the signals to H, M, and L, they do not recombine and pass thru the other drivers circuts. There are multiple seperate + and - paths in the crossover network.

In a current source compatiable crossover network, things change. The seperate crossover circuts are in series. By pass caps and inductors are used above and below the operating range of the driver circut to let the current sourced signal to pass to associated drivers. There are few seperate + and - paths in the crossover network.

see attached

CurrentSourceCrossoverNetworks.pdf

CurrentSourceCrossoverNetworks.pdf