AVS › AVS Forum › Audio › Receivers, Amps, and Processors › What limits power in amps and receivers?
New Posts  All Forums:Forum Nav:

# What limits power in amps and receivers?

Power Limit on Amps and Receivers

I have long been interested in what factors limit the power in amps and receivers. This article is an attempt to enumerate these factors. I am not an electrical engineer, and I make no claims all the information contained here is accurate - I have read numerous articles and consulted with electrical engineers to try to understand these concepts better.

I would like to thank the writing talents and knowledge of Gene DellaSala at Audioholics and Rodd Elliott at sound.westhost.com for their many excellent articles on amplifiers and amplifier power. Most of the information here I learned from them.

This will focus on typical class AB amplifiers.

Corrections and comments are always welcome

A few brief concepts

In case you know little to no electronics, here's a brief refresher.

In this article we are mainly interested in voltage, current (measured in amps)and resistance (measured in ohms). If you think of a battery as a storage tank of water, you can think of voltage as water pressure. Current could then be thought of as water flow rate. Resistance is pipe size. A smaller pipe makes it harder to push water through it.

A simple formula, ohm's law relates volts, amps and ohms. The formula is v = i * r, or volts is equal to amps times ohms. i represents current. An example would be a a circuit where you measure 28.3 volts and 3.5 amps. Resistance would be 8.1 ohms.

Power is what we are concerned with here. Electrical power is measured in watts. The formula for power is p = v * i, or watts is equal to volts times amps. Using the above voltage and current, we would have 28.3 * 3.5 or 100 watts.

Some people, like some audio salespeople I have talked to, seem confused by voltage and current. They seem to think they are not related. A salesperson told me that there's voltage watts and current watts, and current watts are what gives you a lot of power or something to that effect. Ohm's law tells us differently.

There are two kinds of electricity, alternating and direct. An audio signal has an alternating current. It's constantly changing. When we talk about the resistance to alternating current, we use the term impedance to differentiate them, because they don't work quite the same way. For example, direct current going through a coil of wire will only encounter resistance from the wire. Alternating current going through the same coil will create a magnetic field which will impede the flow of electricity.

An audio signal in the analog domain consists of a constantly changing voltage. If it's a line level signal, it might vary between -1 and +1 volts. If the signal was a simple 1 khz tone, it would look like a sine wave the repeats 1000 times every second. Audio consists of a bunch of waves added together, but ultimately is still a signal with variable voltage. A line level signal amplified would ideally be nothing but a bigger version of the line level signal. In the real world, it would contain noise.

To drive speakers, you need both voltage AND current. Quite a bit of it. The reason they call it a power amplifier is because that's literally what it does. It amplifiers a low voltage low current signal to a high voltage high current signal.

Speakers need quite a lot of current as they have a low impedance, and ohm's law tells us that a low impedance allows a lot of current to flow. To power an 8 ohm speaker with 100 watts, you need 3.5 amps. If you think 100 watts is not much power, consider how hot a 100 watt light bulb is. It can burn you.

The Outlet

Power starts at the power outlet. Seeing as how I live in the US, I will discuss a typical US 15 amp circuit.

According to what I have read, in the US, code dictates that 15 am fuses and breakers are only designed for continuous 12 operation, or 80% of max capacity. They will likely handle higher loads for short time intervals, but would blow under a continuous load of 15 amps.

You are also limited by other devices plugged into the same circuit.

According to an Audioholics article, there are some real limits to output power based on wall power. Assuming a derating of 15 amp circuit, you can only pull 12 amps from it. Current and voltage may not be in phase with each other, thus also limiting power. A power factor of less than one indicates that the voltage and current are out phase. By that, I mean that peak available current, and peak voltage are not occurring at the same time. Borrowing information from the Audioholics article, power is limited to 12 amp x 120 volts x 0.72 (power factor) = 1036 watts. Approx 1000 watts is the amount of power you can pull from the wall. Nor can you amp/receiver put all this power to the output terminal due to losses which I will cover later.

Interestingly in spite of these limits, you don't hear of a lot of people tripping breakers or blowing fuses while listening to movies or music. Anecdotally, wall power does not seem to be the main limitation of amplifier and receiver power.

The Transformer

The transformer(s) likely places the biggest limitation on amplifier power. The transformer takes high voltage AC power, such as the US 120 volt AC power, and lowers the voltage.

Transformers work by passing alternating current through one winding of wire called the primary. Alternating current through a winding creates a magnetic field. This field induces current into a secondary winding. A difference in the number of windings between the primary and secondary changes the voltage.

A receiver transformer likely has taps at various places in the secondary winding. Depending on where the tap is, you can get different voltages. A receiver needs a lot of different voltages to supply various ICs and circuits. An amplifier is simpler and would not need as many different voltages. One of these taps will supply the amplifier.

If a concrete example helps here, I believe a 200 watt amplifier would use a transformer supplied voltage of around 80v (I based this on a cursory examination of Leach's low TIM amplifier.)

Due to the way conventional class AB amplifiers work, we need to convert that 80 volts AC into a positive and negative DC voltage. This will be discussed in the next section.

As mentioned above, the transformer is the primary power limiter. One factor limiting it's output power is it's own resistance to current.

This is a bit of a tricky concept, but as you increase the volume on an amp/receiver, you are decreasing the resistance of the amplifier. This allows more current to flow through to the speaker. This presents a higher load on the power supply.

As the resistance of the amplifier drops, it becomes closer to the transformer's own resistance. Without getting into the math, this causes the voltage from the transformer to drop. If this voltage drops too much, the amplifier will start clipping.

Another factor is heat. The more current flow through the transistor, the hotter the wires in the winding will get. This will increase resistance further limiting performance. I measured the voltage from a wall wart (plug in power transformer,) into a low resistance dummy load, and the voltage on the meter started dropping almost right. The most likely explanation was a heat related increased in transformer resistance resulting in a voltage drop.

Transformers have a definite limit on power handling. One would expect their amp or receiver's protection circuits to kick in before a transformer would be damaged. The transformer could also be equipped with a one shot fuse to protect your receiver.

One more factor to consider is that the transformer is connected to a a rectifier which is connected to filter capacitors. This arrangement results in a stable voltage under normal operation. I don't pretend to understand all the details, but under heavy load the power supply may not be able to maintain a steady voltage and this could result in distortion.

The Power Supply

A standard linear amplifier power supply consists of the transformer, a rectifier and filter capacitors.

The transformer supplies an AC voltage stepped down from the voltage supplied from the wall as described above.

The rectifier separates the negative and positive AC voltage. It provides a positive and negative terminal. At this point, voltage is still AC.

To convert the voltage to DC, filter capacitors are connected to the rectifier. The filter capacitors smooth out the voltage. A steady voltage is important as change in voltage, known as ripple, can cause the amplifier to distort.

To attempt to see power spikes, I hooked a Kill A Watt meter up to my receiver. Sending drum hits to the receiver, I could see power output spike, but not immediately. A possible explanation for the brief delay was that the initial power demand was met by the filter caps, and the spike on the Kill A Watt meter was the demand for power as the caps recharged (further experimentation would be needed to confirm this, but it seems like a likely explanation.)

Amplifiers work by amplifying a small signal into a larger signal. They are limited by how much they can amplify the signal by the voltage supplied from the power supply. This is voltage is sometimes called the rail voltage.

It's not possible for the amplifier to amplify the signal past the rail voltage. If the rail voltage is -60/+60 volts, and the amplifier tries to amplify the signal to 65 volts, it won't happen. The amplifier will clip, and output 60 volts. Clipping causes distortion. It also causes additional power to flow through the speakers as the average voltage has now increased. This power can damage speakers.

Even if you don't clip due to exceeding your rail voltage. clipping can occur due to the voltage being reduced due to high loads (as explained in the section on power transformers.) A multi-channel receiver is likely not going to distort by exceeding normal rail voltage. It's going to distort due to the rail voltage dropping under heavy load.

If you look at a review of an audio video receiver, you will see that the power output starts dropping as more channels are driven. A review of my Yamaha RX-V3900 shows 189, 150, 100 and 88 with one, two, five and seven channels driven ( 1 khz into 8 ohms at clipping.) The limiting factor here is the inability of the power supply to maintain voltage under load. As the load increases by driving more channels, the power supply can't maintain the normal rail voltage which is why the power is decreasing as the channels driven increases.

Decreasing impedance has the same affect. Ideally, if you switched out your 8 ohm speakers for 4 ohm speakers, your power output would double. Few audio video receivers can pull this off. It's no different than driving more channels. Receivers commonly share a power supply, and load from lower impedance speakers or load from driving more channels works the same way - if the load is high enough, clipping will occur.

If your load is too high, you will activate the receiver or amps protection circuits - hopefully. Receivers attempt to protect themselves from acts such as shorting the speaker terminals, or overheating.

I have read stories about people driving 4 ohm speakers with a receiver not designed for that, and eventually the receiver fails. Heat is a legitimate problem. You think about how hot a 60 watt bulb gets. Now try to drive your speakers to high volumes, and realize your receiver is maybe 50% efficient, and realize half your power is going into heat.

Power Transistors

The final stage of amplification is performed by power transistors. The power transistors are connected to the power supply voltage rails. The power transistor multiplies the audio signal up to the rail voltage.

These handle very high voltage and current. A power transistor amplifying a signal to 100 watts into an 8 ohm speaker will be working on 30 volts and 3 amps. If this does not sound like a lot, just think of holding a 100 watt lightbulb while it's on. That lightbulb is dissipating 100 watts of electricity as heat ( yes I have said this twice, in case you did not read the whole article, or were not paying attention.)

Power transistors are bolted to large pieces of metal called heat sinks. This helps dissipate heat that's not being turned into power.

Power transistors, like any other electronic component have a safe operating area (SOA.) There's an upper limit to how much voltage and current they can handle without failing.

Something I read from a few sources, is that the gain of a transistor will drop as current increases. I am not sure how much of a factor that is.

Losses

Losses occur everywhere in an amp/receiver. Every device resists the flow of electricity and converts power to useless heat. The transformer has losses, the rectifier and capacitors and wiring have losses.

Class AB amplifiers are inherently inefficient as well. The number differs. A ball park estimate is that %50 of power being pulled from the wall is converted to power at the speaker terminals.

While not part of the amplifier, speakers are probably the least efficient part of the audio chain. A typical speaker used in a home audio system is probably below %5 efficiency. That means 95% or more of the power your amplifier supplies to it is converted to heat, and not sound.

Limiter Circuits

This is an area I don't fully understand. I won't lie. My understanding is that some receiver and amplifiers have limiter circuits. They detect and respond to overload conditions and limit power. Clearly the presence of any limiter circuit could limit power output. Pro amps are designed for abuse, and they often advertise some sort of soft clipping circuit. The idea is simply to reduce the signal so that the amplifier is not clipping - this can protect both the amplifier and the speakers connected to it.

Putting it all together

Going back to wall power, it can be seen that, at best, you are not going to be able to pull much more than 1000 watts from the wall. And with many amps and receivers, less than that. Losses would be around %50, so you are going to be able to put out maybe 500 watts to the speaker terminals. It should not be surprising how much power goes down per channel when trying to drive five or more speakers.

Short term power capability can be higher. The filter capacitors, for example, can store power, which can meet temporary demands, which is a good thing. Peak power needs can be 30 or more times average power needs (15 dB or greater peak to average level.)

This begs the question of why they make amps rated for much higher power than can be supplied continuously. First off, there's a demand for them They can also meet peak needs better than less capable amps. Also, each channel is not always being driven to full power. So it's helpful to have channels each capable of peak output power even though all channels can't be driven to peak power at all times.

So don't take this article as any sort of condemnation of powerful multi-channel 200 watt per channel power amps.

Reference articles -

(Much good information on multi-channel amplifier power limitations)
http://www.audioholics.com/education...er-test-page-3

(Amplifier Power supply design)
http://sound.westhost.com/power-supplies.htm

( fairly technical, but extensive article on transformers)
http://sound.westhost.com/xfmr.htm

### AVS Top Picks

FYI, the theoretical maximum efficiency of a Class AB amplifier is 78.5%
I did read that. But I also read that that's only theoretical. I read an article which gave a range, and 50% seemed like a nice round number.

Do you have another suggestion for a number?
Quote:
Originally Posted by MichaelJHuman

Do you have another suggestion for a number?

Well, my main point is that a Class AB amplifier CAN NOT be more than 78.5% efficient.

I've measured amps at anywhere between 20-30% efficient and others close to maximum so your figure is as good as any.
Great info MJ,

Something you mentioned that may need repeating, is that an AVR rating with all channels driven, while it may be an indication of a good amp (or amps), isn't needed in a real world model (and is likely how the manufacturer's specs are able to get around it). In a surround system, most of the power will be split between the front three channels (LCR) and little will be used in the side and rear surrounds. So a rating of three (or maybe 4) channels driven, would make a better real world indicator.

Your article made me wonder about those who test amplifiers. If you have all your testing equipment plugged into the same circuit, wouldn't that affect the numbers?
Quote:
Originally Posted by MLKstudios

Your article made me wonder about those who test amplifiers. If you have all your testing equipment plugged into the same circuit, wouldn't that affect the numbers?

I would think it could affect your available power slightlly. but the amp would have to be capable of drawing enough power to trip your breaker before blowing the amp's own internal fuses or putting it into protection.
Quote:
Originally Posted by cavu

FYI, the theoretical maximum efficiency of a Class AB amplifier is 78.5%

This is completely untrue...The Sunfire TGA-7401 can deliver 400 watts x 7 channels (2800 watts) while drawing only 1440 watts maximum off the AC line. It's 200% efficient! This is accomplished using the Sunfire Magical "Tracking Down converter"!

Other inferior amplifiers like Anthem Statement require two AC inputs from different 15 amp circuits just to accomplish 325 watts x 5 channels.

I'm going to put one of those magical down converters on my incoming AC line and cut my electrical bill in half!

A little knowledge like that found in the article the OP posted, helps you avoid being deceived. Good information!

mk
I would deem that the Sunfire is not a typical class AB amp. More like a class H amplifier. Clearly reducing rail voltage just above the needed level should increase efficiency. I remember, years ago, that there were some affordable receivers from Technics (I think,) which used some sort of variable voltage rail design. I am guessing expense, and lack of perceived consumer value are keeping manufacturers from using these designs, but perhaps there are other reasons as well.

My writeup does not discuss amplifiers which support multiple circuits via two power cords. That will of course increase an amp's ability to deliver power without overloading when wall power is the limiting factor.
Keep in mind that with a single chassis multichannel amplifer, a multi-mono design, where each channel has its own (smaller) transformer, will have lower per channel power limits than a similar amp that shares a single (or 2) larger transformer(s), assuming all channels are not being driven equally at the same time. There are very few true multi-mono design multichannels amps out there; the only ones I recall right off are the Sherbourn and the Anthem Statement, but there are probably others.
Quote:
Originally Posted by MichaelJHuman

I would think it could affect your available power slightlly. but the amp would have to be capable of drawing enough power to trip your breaker before blowing the amp's own internal fuses or putting it into protection.

That would have to one ungodly powerful amplifier.
Quote:
Originally Posted by dsmith901

Keep in mind that with a single chassis multichannel amplifer, a multi-mono design, where each channel has its own (smaller) transformer, will have lower per channel power limits than a similar amp that shares a single (or 2) larger transformer(s), assuming all channels are not being driven equally at the same time. There are very few true multi-mono design multichannels amps out there; the only ones I recall right off are the Sherbourn and the Anthem Statement, but there are probably others.

Amps in that category are likely limited by only by your wall power as far as continuous power goes.
On the AC power segment. A 15 Amp circuit breaker will allow an amplifier to play a continuous test tone while drawing 15 Amps for 3 hours before tripping. But amplifiers mostly play music or movie sound tracks, which normally draw much less current much of the time. 15 Amp circuit breaker will not trip on short musical peaks drawing much more than 15 Amps. And the only way a circuit breaker can limit current is by tripping.
Quote:
Originally Posted by Speedskater

On the AC power segment. A 15 Amp circuit breaker will allow an amplifier to play a continuous test tone while drawing 15 Amps for 3 hours before tripping. But amplifiers mostly play music or movie sound tracks, which normally draw much less current much of the time. 15 Amp circuit breaker will not trip on short musical peaks drawing much more than 15 Amps. And the only way a circuit breaker can limit current is by tripping.

Do you have links to information on how long it takes fuses and breakers to blow at various current loads?

I was looking for something with that info.
Quote:
Originally Posted by MichaelJHuman

Do you have links to information on how long it takes fuses and breakers to blow at various current loads?

I was looking for something with that info.

I have some printed ones (somewhere) but they are probably not representative of the breakers in your or my house.
Quote:
Originally Posted by MichaelJHuman

There are two kinds of electricity, alternating and direct. An audio signal has an alternating current. It's constantly changing. When we talk about the resistance to alternating current, we use the term impedance to differentiate them, because they don't work quite the same way. For example, direct current going through a coil of wire will only encounter resistance from the wire. Alternating current going through the same coil will create a magnetic field which will impede the flow of electricity.

This is pretty much right, but misses what's really happening. AC and DC current do actually work the same way, but there's a couple differences

Alternating current: has a phase and a magnitude. We model AC current using complex numbers ("phasor" is what it's called, but it's nothing more than the magnitude and argument of a complex number). We'll write, for instance, I = 2âˆ 30Âº, which means that the current has sinusoidal waveform I = 2 sin(376.99 t + 30Âº). 367.99 = 60*2*Pi, since 60 Hz is the operating frequency of AC current in the US, and to convert hertz to radians/sec (as needed here), you multiply by 2*Pi. The +30Âº inside the sine is poor notation, since the input to everything has to be in radians (calculus doesn't work in degrees). To work with that form, you need to convert 30Âº to radians, so really, you'd enter I = 2*sin(376.99t + Pi/6).

Direct current: has phase = 90Âº and has no operating frequency. It's completely constant in that it doesn't vary with time at all. Think of it as AC current with phase = 90Âº running at 0 Hz: I = 2*sin(0t + 90Âº) = 2*sin(90Âº) = 2. Note that lack of time dependence.

Impedance is also a complex number. The real component is the resistance, and the imaginary part is the reactance (comes from inductance and capacitance).

The thing is that Ohm's Law doesn't actually apply to circuits with inductors and capacitors in it. When generalized to impedance, you get V = IZ, where Z is the impedance, so you have the same result, but it's not technically Ohm's Law (although we usually call it that). This result holds for AC circuits, and allows you to do something like this:

Imagine you have a component with impedance Z = 3 + 4j = 5*exp(j*Pi/4) = 5âˆ 45Âº (j is the imaginary unit, j^2 = -1, in EEing, since i is used for AC current). If you have a AC current of 1 A with phase 0Âº, then you write that as I = 1 = 1âˆ 0Âº. You can then use "Ohm's Law" to find the voltage:

V = (5âˆ 45Âº)(1âˆ 0Âº) = (5*1) âˆ  (45+0)Âº = 5âˆ 45Âº = 5*sin(376.99t + Pi/4)

(to multiply complex numbers in phasor notation, multiply their magnitudes and add their phases--phasors are nothing but complex numbers in polar form). If you were to toss an oscilloscope across this circuit, the above waveform is what you'd see for the voltage.

If your current is DC, you cannot use Ohm's Law and instead need to break it down into resistors and inductors/capacitors and figure out what happens there. Suppose you have a very simple RC circuit (a resistor and a capacitor). The current through a capacitor is C*dV/dt, so after using Kirchoff's law, you end up with a linear DE of the form
C*dV/dt + V/R = 0

You can solve this as V = V0*exp(-t/RC), where V0 is the initial voltage in the capacitor. Thus, you see that if you start with a fully discharged capacitor, you end up with 0 current in your wires, since you have no voltage across the circuit. Further, you note that if you subject the circuit to a constant voltage source, you also get no current, since dV/dt = 0 then. The voltage drop across an inductor is V = dI/dt*L. Thus, if the current isn't changing with time, you have no voltage drop across the inductor. In DC current, dI/dt = 0, so you don't get anything from it. This is why inductors only work in AC circuits.

Alternating current doesn't require us to do all this, since the concept of phasors and impedance take care of it all for us. We can instead just use the complex version of Ohm's Law and it all works nicely.

Inductors induce time-varying magnetic fields, which create flux. Faraday's Law says that magnetic flux induces an electromotive force--a voltage--that opposes the direction of the flux. By the right hand rule, this means that there is a voltage drop in the wire due to the magnetic flux. This is how "impedance" impedes the flow of electricity. Resistance always induces a voltage drop, and in AC ciruits, reactance does as well because of Faraday's Law.
I know some of that material from physics and other reading. I was trying to avoid all the gory details.

I was trying to explain in a simple way that impedance takes into account the interaction between the electric current in the wire and the magnetic field generated by the current.

If you want to suggest a wording that does not require people to consider linear differential equations, kirkoff's law, complex numbers and phases I would appreciate it
Quote:
Originally Posted by MichaelJHuman

Do you have links to information on how long it takes fuses and breakers to blow at various current loads?

I was looking for something with that info.

You can find Square D trip curves in their technical library.

In the past, I've posted links or maybe even attached files showing 15A and or 20A Square D QO trip curves.
Quote:
Originally Posted by Raymond Leggs

That would have to one ungodly powerful amplifier.

Not necessarily.

Depending on wire gauge, length of circuit, and current demand, enough voltage drop could occur to negatively impact the amp's output without tripping the circuit breaker.
Quote:
Originally Posted by whoaru99

You can find Square D trip curves in their technical library.

In the past, I've posted links or maybe even attached files showing 15A and or 20A Square D QO trip curves.

Nifty. I will try to find them, thanks.
"What limits power in amps and receivers?"

Only my income...
Quote:
Originally Posted by BNW

"What limits power in amps and receivers?"

Only my income...

Hilarious, but you will forgive me if I don't add that to the writeup
If I have an amp that requires more than 1000 watts, how can I acheive this? Is there a way to "double" power outlets?
Quote:
Originally Posted by sound dropouts

If I have an amp that requires more than 1000 watts, how can I acheive this? Is there a way to "double" power outlets?

Some amps have multiple power cords. If yours only has one power cord, you are limited by the factors discussed in the OP.

Note however, that your peak power can be higher than continuous power which is one advantage to amps who's rated power is more than you can draw from the wall.
You use the appropriate wire gauge, circuit breaker, and plug/receptacle.

However, many of us use very powerful amps on normal 15A or 20A circuits without any problems at all.

Consider that some of the amps I use could draw 25A @ 120V (3,000 watts) at full continuous 8 ohm output. However, that type of output is seldom, if ever, seen outside of a bench test using test tones.

This same amp, playing at "full power" (occasional clipping) using typical music draws only 7.6A (912 watts).
My Yamaha has never come close to hitting it's max power consumption even at excessive volumes. I mentioned this in a thread a long time ago, and a number of people felt that the Kill A Watt meter I used was to slow to measure any peak demand so it did not tell the whole story.

But if you are talking continuous power, I was playing a movie REALLY loud and not drawing more than 200 watts from the wall.

I did not mention 20 Amp circuits in the article. I thought I remember reading that 20 A circuits may not help much because you are limited, with UL certified gear, to 12 amps or so? Power cords I have looked at, were rated for 12 Amps.

I would think your amp would be fused to blow well below 20 Amps. Anyway, if you are not blowing fuses or tripping breakers, wall power is likely not limiting you.
Quote:
Originally Posted by MichaelJHuman

Power Limit on Amps and Receivers

IThe Transformer

The transformer(s) likely places the biggest limitation on amplifier power. The transformer takes high voltage AC power, such as the US 120 volt AC power, and lowers the voltage down. Various regulators convert the AC voltage into DC power for the various circuits in an amp or receiver.

In a typical receiver, one transformer is used for all amplifiers in the receiver. An expensive amplifier could have one transformer per channel. The same transformer which powers the amps might also supply power for other circuits which will reduce, somewhat, the amount of power it can supply to the amplifiers.

The transformer has limitations. One limitation is that it can only draw so much current before being damaged. You are (hopefully) never likely to see this limit, as fuses and other protection circuits should intercede.

The larger the transformer, the more power it can supply. Cheap receivers, in particular, will skimp on the transformers as they are relatively expensive.

The power transformer is the highest cost component within an AVR, in an AVR with an SRP of \$799 or less. The power transformer takes up roughly 15-20% of the bill of materials component costs, alot of copper and steel raw materials in the transformer. Note that UL/CSA requires a internal thermal circuit breaker so if the power transformer heats up and blows it will be an expensive repair job..

Quote:

The Power Supply
A simple amplifier power supply consists of the transformer, a rectifier and filter capacitors. The transformer supplies an AC voltage stepped down from the voltage supplied from the wall. The rectifier has two output terminals one of which contains a negative voltage, and the other positive. The filter capacitors convert the rectifier voltage to a steady DC voltage and can also store power in reserve. The amplifier is pulling power from the capacitors, and the transformer/rectifier are refilling it. ( You can see this with a Kill A Watt power meter and a piece of music with mainly kick drums on it - you will see the power consumption spike after each drum hit.)

These work together to provide the voltage supplied to the amplifiers. This is often called the rail voltage. The rail voltage will differ from receiver to receiver. A 30 volt supply would be sufficient to produce 100 watts into 8 ohms. Technically, the supply would consist of a -30 volt and +30 volt line. This is needed due to how amplifiers work. When the input signal is negative, a transistor or transistors supplied by the negative voltage amplify the signal. When the input signal is positive, a transistor or transistors supplied by the positive voltage amplify the signal. Technically, there's a switchover between the two, so both sides of the amplifier will be active part of the time.

Key component costs are the power supply capacitors, rectifiers and regulators besides the power transformer. Also depends upon its design..
Can be a high current..
Can be high voltage..
Amount of regulation..

The majority of AVRs today are designed with higher voltage rails which start to collpase when more channels and current are demanded..
A high voltage can deliver higher peak output power..
Again loudspeaker sensitivity, impedance, room size and target volume level all enter the equation as well.

Quote:

Power Transistors

The final stage of amplification is performed by power transistors. The power transistors are connected to the power supply voltage rails. The power transistor multiplies the audio signal up to the rail voltage.

These handle very high voltage and current. A power transistor amplifying a signal to 100 watts into an 8 ohm speaker will be working on 30 volts and 3 amps. If this does not sound like a lot, just think of holding a 100 watt lightbulb while it's on. That lightbulb is dissipating 100 watts of electricity as heat ( yes I have said this twice, in case you did not read the whole article, or were not paying attention.)

Power transistors are bolted to large pieces of metal called heat sinks. This helps dissipate heat that's not being turned into power.

The power transistors limit output power because they can only handle so much voltage and current before failing. This limit should not be reaching under normal operation though.

Think about a power device like a fuse or lite bulb, it is designed to handle a certain level of voltage and current..
If exceeded they will short out and blow a channel or even a loudspeaker in extreme cases. Also here thermal efficiency of the heat sink & fans and its ability to clear out the hot air is crucial...
Typically an output device's current/voltage specs will decrease significantly once its case temperature exceeds 70 degrees C. Sometimes multiple output devices are used, pairs in series for higher voltage handling or in parallel for high current capability.
Also chassis ventilation is critical and air space around the chassis will determine its reliability, AVRs like other electronic products do not like high heat temperatures..

Quote:

[Limiter Circuits

This is an area I don't fully understand. I won't lie. My understanding is that some receiver and amplifiers have limiter circuits. They detect and respond to overload conditions and limit power. Clearly the presence of any limiter circuit could limit power output. I want to say more on this, and I am trying to find out how common this is, and whether this is a common cause of the power drop of you see on benchmarks as more channels are driven at the same time.

The limiters protect the amplifier from destroying itself either to higher temperatures or destructive conditions such as a short ciruit or even low impedances. Thats why Yamaha and others limit their power supply output as to protect the output stage.. By the impedance selector, which basically lowers the power supply output (different power transformer windings) as the loudspeaker impedance decreases..

Just my \$0.025..
I did not see any actual corrections to what I said, M-Code. It looks like you were just clarifying some points. Did I miss something?

You mentioned an internal thermal device of some sort. You mean that transformers have a built in thermal shutdown device?

I was confused about your comments on power transistors. Seems you said what I said. So if you are trying to clarify something, I am missing it. I think I could have worded it better, and I will probably do so.

I am curious about your comments on limiters. Were you talking about a fixed impedance switch which lower rail voltage by using a different tap, or some sort of automatic reduction in rail voltage?