Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

How It Works: the Power Amplifier (Part 11)

Otherwise known as the big black box at the bottom of the rack, the power amplifier is the most unglamorous piece of audio equipment known to man. David Mellor seeks out its finer qualities.

Otherwise known as the big black box at the bottom of the rack, the power amplifier is the most unglamorous piece of audio equipment known to man. David Mellor seeks out its finer qualities.

You wouldn't get far without an engine in your car, and there won't be much music made in your studio without a power amplifier to drive the monitors. The power amp provides the motive force that turns the millivolts coming out of the mixer into the megawatts (slight exaggeration!) that get those speaker cones flapping.

If you could walk into a shop and buy an adequate power amp without knowing anything about them, then there wouldn't be much point in writing an article explaining how they work. But getting the best from a studio set-up depends on knowing a little bit about how everything works. The power amp isn't the most exciting link in the audio chain, but get it wrong and it will be the weakest link. And that weak link could mean that the quality of your end product is less good than it ought to be.

Figure 1.


Silly question - or is it? The most basic power amplifier looks a bit like the one in Figure 1. Here, I have shown all controls and connections on the front panel. There is absolutely nothing hidden round the back. As you can see, it looks rather different from a hi-fi amplifier. And so it should, because it is optimised for studio rather than living room use.

The main feature is the lack of controls. There are no treble and bass controls, no input selector switch, no balance knob - and definitely no 'loudness' switch. 'Loudness' on a hi-fi amplifier boosts the treble and bass content of a signal. It's great for listening to (sometimes), but if you make recordings using a hi-fi amp for monitoring with 'loudness' switched on, then it would be better termed a 'limpness' control because you will end up putting less treble and bass on tape. Anyone you send the recording to will hear it that way, not the way you heard it.

The reason for the lack of controls is that all the controls you need are on the mixer. You need to monitor what is going down onto tape. If there were any signal processing functions on the power amp, then they would only affect what you hear, not what you record. A proper power amp should have two audio controls only, a separate Level control for each channel. Nothing more.


The power amp takes the output signal from the mixer, probably around the 1 volt level, and boosts it sufficiently to drive the speakers - up to 28 volts for a typical 100 watt amplifier, with a current capability of around 3½ amps. (Don't get confused between 'amps', short for Amperes of current, and 'amp', short for amplifier). To do this requires some pretty powerful output devices - transistors, valves or MOSFETs.

There used to be a great debate over whether transistor or valve amplifiers sounded better. Valves are supposed to give a 'warmer' sound than transistors, which is largely true. Some commentators have even said that 'valve watts' are louder than 'transistor watts'. Utter rubbish!

Figure 2.

Figure 2 shows the principal difference between valves and transistors. When an amplifier is driven within its capabilities, it makes hardly an iota of difference whether it is valve or transistor. But when those limits are approached, the situation is altered. A transistor amp maintains a clean performance right up to its limit. Go beyond that and a harsh distorted sound is the result. 'Clipping' is the term used to describe this effect. A valve amplifier goes into clipping very much more softly, with rounder edges to the dipped peaks. Not unsurprisingly, this sounds 'nicer' than transistor clipping. That's why many guitarists still prefer valve amps to transistorised ones. But stick within the amplifier's rated capacity and clipping will not be a problem, with transistors or valves.

MOSFETs are a more modern invention. These are a type of transistor but with some of the qualities of a valve. I mention them for the sake of completeness really, because any difference in sound quality between transistors and MOSFETs will be minimal, and a matter of importance for the most exceedingly refined tastes only.


Marketing personnel are well known for their uncanny ability to differentiate people into a range of social classes and to target their products in what they consider to be an appropriate manner. Amplifiers, too, come in a range of classes according to the internal design of the circuitry. There are quite a number of different classes, but the most important to us are Class A and Class B - the upper crust Class A, and the working man's Class B!

I have already said that a 100 watt amplifier is typically capable of putting out 3½ amps of current. That's what comes out, but there are various extra currents necessary for an amplifier's internal workings. Class A amplifiers are very wasteful with current. Class A keeps its output transistors (or valves) switched on and working all the time, which results in a high wastage of current, even when there is no input signal. You can recognise a Class A amp from the temperature of the air around it and from the speed at which the dial in your electricity meter revolves - it's only about 20% efficient!

Class B, on the other hand, can have an efficiency of over 70%. Of the two output transistors (or valves), only one is active at any particular time. The standing current in the amplifier is only a few tens of milliamps. In practice, the two classes are combined together in the design - mainly Class B with a pinch of Class A. Most audio power amplifiers are of this so-called Class AB variety.

In theory, a Class A amp ought to sound better - Class B creates distortion when the transistors switch on and off - but they are harder to make, especially large ones. A 100 watt Class A amp is indeed a big'un. Class AB designs are made up to 1000 watts and more, and most people are satisfied with their performance. There are also Class C and Class D amps (and a few more esoteric classes). Class C isn't usable for audio; Class D at one time promised compact efficient amplification, but it never really caught on.


The first thing you need to consider when buying a power amplifier is the power rating. The object is to have the monitoring as loud as you want it to be, without any possibility of the amp running into clipping - and preferably have a safety margin. When choosing an amp, professional studio designers will make an intricate calculation, taking into account the desired SPL (sound pressure level), the contribution the reverberation in the particular room makes to the overall level, the efficiency of the loudspeakers, and the current relationship between the Moon and Pisces!

The money spent on the studio designer's fees might be better spent on buying a nice big amp in the first place - particularly in a home studio. With averagely efficient speakers, 50 watts per channel is an adequate power rating for living room use. 100 or 150 watts won't give that much more audible level, but the added headroom (safety margin) before the onset of clipping will be an advantage. 300 watts would be more than enough for any single-box passive crossover loudspeaker system. Anything above that is really into the big-boy league. They can afford a pro studio designer.

One important point to remember is that these wattages are RMS values - for example, 50 watts RMS. It's not necessary to know that RMS stands for Root-Mean-Square (that's too mathematical!), just remember that the RMS wattage is a real indication of the 'oomph' it has. There are other ways of quoting output power, such as 'music power' or 'peak power'. If it doesn't say 50 watts RMS per channel, then it's down to luck what you get.


Something which crops up in many an amplifier specification sheet is the damping factor. Once the cone of a speaker is set into oscillation, then it stands to reason (and the laws of conservation of momentum) that it isn't going to stop until something makes it stop. One of the 'somethings' that bring the cone to a halt is the damping effect of the amplifier. Some amplifiers are better than others.

This is the technical bit: The cone of the speaker moves in response to the current supplied by the amplifier. As a result of this motion it produces its own current, which goes back down the speaker cable and into the output stage of the amplifier. If this current is somehow dissipated, this will tend to counteract any spurious motion of the speaker - that which is not initiated by the amplifier. If the output impedance ('impedance' is similar to electrical resistance) of the amplifier is low, this will be the case. The lower the better.

That's it in rough terms. The only problem is that the amplifier's output impedance can be as low as it likes but it is still connected in series with the resistance of the speaker cable, and indeed the resistance of the loudspeaker voice coil itself. If the output impedance of the amplifier is in the region of a tenth of an ohm or less, then all is well. If the output impedance of the amp was claimed to be a millionth of an ohm, it wouldn't be a lot better in terms of end result. Damping factor is sometimes expressed as a ratio - the nominal impedance of the loudspeaker, probably 8 ohms, to the output impedance of the amplifier. If the output impedance is a tenth of an ohm, the damping factor is 80. End of technical bit.

Distortion and signal-to-noise ratios of power amplifiers can be of similar values to other pieces of equipment. Look for distortion measurements of less than 0.1 % and signal-to-noise ratios of better than 10OdB. A ruler flat frequency response is the norm these days.


Unfortunately not. I can't think why there aren't more good, low-cost, power amps available. We ought to be able to get away from using those less-than-suitable hi-fi amplifiers in our increasingly professional home studios.

Credit where it's due. I use a 15 year old Quad 303 power amp, which boasts 45 watts per channel, in my home set-up. After fitting it with some XLR connectors I couldn't hope for anything more suitable. It isn't manufactured anymore, but Quad still make proper power amps at sensible prices that are worth investigating.

There are plenty of high-end manufacturers, such as Amcron, BGW, Turner and others, who make whopping great amps at whopping great prices, so the secondhand market may be something to consider. Look for a no-frills construction (you don't really need VU meters) and if you're after a 300 watter, make sure you use the correct lifting posture! Bear in mind that if an amp is fan-cooled, the noise of the fan may be annoying in the studio.

If you are forced to go for a hi-fi amp out of necessity, then there is no harm done, because you can still use it as though it were a dedicated power amp, by making sure the tone controls are set to flat and the balance is dead central. Since the hi-fi amp will contain a pre-amplifier as well as the power amp part, it will probably be best to use it on its least sensitive input, to avoid any headroom problems.

Personally, I think it's time some of those nice manufacturers started to address the low-cost power amp market. I'll bet there are quite a few people waiting.


Many power amplifiers run their output transistors right up to the limit to get the quantity of watts required, if the maximum current is exceeded, then transistor failure is the probable result - the inevitable result if the speaker cable develops a short circuit.

Any decent power amp will have some sort of protection mechanism built in, so that even if you connect the output terminals together with heavy gauge copper wire the amp will come to no harm.

The simplest method of protection is the fuse. Unfortunately, it reduces the damping factor of the amp (see main text). There is also the snag that transistors can be very quick to blow, when they want to. So the transistor blows to protect the fuse!

More sophisticated amplifiers use electronic protection, reducing the drive to the output transistors when excess current is being drawn. This can, unfortunately, increase the distortion level of the amp. The better designed - and more expensive - the amp, the less this is likely to happen.


The amp forces voltage into the loudspeaker. The speaker draws current from the amp. Remember this - and that the lower the impedance of the speaker, the more current it draws - and you will have no problem with impedance. Remember also that the more current that is drawn, the greater the resulting sound level - within the power limits of the speaker.

Most speakers these days are nominally 8 ohm impedance. 'Nominally' means that at some frequencies the impedance is more, at others it is less. An amplifier may say that it is suitable for speakers with impedance from 4 to 15 ohms. It really doesn't matter how high the impedance gets, the level will just reduce. But if the impedance is too low, the amp may fail or its protection mechanism may operate, causing distortion. Many so-called 8 ohm speakers do in fact drop down as low as 4 ohms at some frequencies. There needs to be a safety margin in the impedance capability of the amplifier. Don't expect to run a 4 ohm speaker from a 4 to 15 ohm amplifier successfully.


In the main text it is described how transistor amplifiers and valve amplifiers clip in different ways. But even when they are working within their limits they produce different types of distortion.

If a transistor amplifier is fed with a 1000Hz tone, it will also produce low levels of 3000Hz, 5000Hz, 7000Hz - in fact, all the odd multiples. Even numbered harmonics are very nearly absent. The effect on a sine wave input is this (exaggerated):

A valve amplifier produces even multiples of the input frequency: 2000Hz, 4000Hz, 6000Hz etc. The effect on a sine wave is this:

Many people would say that valve distortion is more pleasant than transistor distortion. It's certainly different. MOSFETs produce the same type of distortion as valves.


Or how to make a 100 watts per channel stereo amplifier into a 200 watt mono amplifier. This sort of thing goes on more in the PA trade than in the studio, but it's a useful technique to be aware of. The trick is to feed the same signal to both channels, like this:

But before it reaches the amp, put it through a transformer box which turns one input upside down, like this:

If the stereo output of the amp has two red terminals (hot) and two black terminals (ground), the output to the speaker is taken from the two red terminals. The ground terminals are ignored.

In the normal (unbridged) situation, the input signal is referenced to ground (earth). In the bridged configuration it is referenced to an inverted copy of itself. The amplified signal is effectively doubled.

In theory, this should quadruple the output power of the amplifier, but in practice it only doubles it, as only the voltage of the output is increased. The amp cannot supply any more current. Even so, this is still useful as it gives extra versatility. Some power amps have a switch which will perform the bridging operation instantly.


Multiple loudspeakers can be driven from the same amplifier, as long as the impedance stays within the right range. If you link speakers of equal impedance like this...

...the impedance of the whole lot is the impedance of one speaker divided by the number of speakers. Thus, four 8 ohm speakers connected in this way would have an impedance of 2 ohms. If you link them like this...

...the impedances add. In this case, they make 32 ohms. The greater the impedance, the less level results, even with more speakers, so the best answer is to combine these two methods of connection to get a resulting impedance of 8 ohms...


A great source of pub conversation is the question of whether a speaker should have a greater power rating than an amplifier or vice versa. Let's examine the pros and cons:

Amp rating 100 watts, speaker rating 200 watts.
It looks as though the speaker is unblowable with this amp. The 100 watts safety margin ought to be enough. But if the amp is driven into clipping it will produce a lot of high frequency distortion. Even if the bass unit of the speaker is rated at 200 watts, the high frequency unit will not be. The excessive high frequencies will blow the tweeter. But rest assured that you are unlikely to damage the woofer, unless you try really hard!

Amp rating 200 watts, speaker rating 100 watts.
Yes, here there is a strong possibility that the speaker will be damaged - but only if the amp is driven above the speaker's rating. At lower levels the speaker will be perfectly safe, the tweeter included. Hopefully, the engineer will be sensible enough to hold back the levels to sensible values.

If you want to be in control, the amplifier should be at least as powerful as the speaker. You may blow it, but it will not happen spontaneously. Having a lower powered amplifier means that the speaker is safer, but the high frequency unit is still at risk. For myself, I like to be in charge. I have used 300 watt amplifiers with speakers rated as low as 40 watts without mishap. The key is to listen to when the speaker starts to complain, then back off the level.


Read the next part in this series:
How It Works - Hard Disk Recorders (Part 12)

Previous Article in this issue

Writing Music for a Documentary

Next article in this issue

The New Standard?

Sound On Sound - Copyright: SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.


Sound On Sound - Nov 1988

Donated & scanned by: Mike Gorman


Sound Fundamentals


How It Works

Part 1 | Part 2 | Part 3 | Part 4 | Part 5 | Part 6 | Part 7 | Part 8 | Part 9 | Part 10 | Part 11 (Viewing) | Part 12

Feature by David Mellor

Previous article in this issue:

> Writing Music for a Document...

Next article in this issue:

> The New Standard?

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

Donations for December 2019
Issues donated this month: 1

Issues that have been donated this month.

Funds donated this month: £0.00

All donations and support are gratefully appreciated - thank you.

Please Contribute to mu:zines by supplying magazines, scanning or donating funds. Thanks!

Monetary donations go towards site running costs, and the occasional coffee for me if there's anything left over!

Small Print

Terms of usePrivacy