What does a speaker prefer to eat?


However, it is the current flowing through the voice coil that provides the motor effect that pushes the cone backwards and forwards.

In order to get a current to pass through the resistance of the voice coil, you must apply a voltage across the coil.

Some speakers are more current hungry than others. Good quality speakers generally respond better with amplifiers with substantial power supplies.

This is a very general overview of a subject that is very complex.

Look up Faraday's Law on Wikipedia.
at a superficial level its not that complex... A current will flow through a resistance if a voltage is applied across the resistance.

No voltage, no current.

A speaker has a winding in it that has a resistance. Apply a voltage to the winding and a current will flow through it.

The greater the voltage, the greater the current that flows.

This implies two things - as long as the voltage is applied from a source that has limitless current capability, an increase in applied voltage will lead to an increase in current flow.

Which gets to the point that Andy makes - a good quality amp starts with a top quality power supply.

Of course not all speakers are the same (some are more equal than others). A low impedance speaker of a particular rating (say, 4 ohm, 91db/w) will require a power supply capable of delivering higher current than an equivalent high impedance speaker (say, 16 ohm, 91db/w). The high impedance speaker will need an amp with greater ability to swing a voltage.

Which is, in part, why tube amps tend to be happier with higher impedance speakers - they are largely voltage amplifiers and not designed to deliver high currents to the speakers, while ss amps are largely current amps and don't apply huge voltages...

I generalise wildly for simplicity - hope it helped.
Last edited:
There are hundreds, if not more, parameters that need to be considered when designing a loudspeaker. However, here is a bit of food for thought.

Some novices think that a BASS woofer flapping is a sign of volume. A good speaker will provide 100dB with no real visual movement of the speaker cone. You will certainly be able to feel it moving.

This is generally achieved by making the suspension of the speaker cone quite stiff.

Immediately we have a conflict of requirement. We want the speaker cone to be able to respond rapidly to small signals from the amplifier AND we want minimum distortion from the cone.

Stiffening up the cone suspension increases the inertia of the cone so that it requires more current to get it moving.

If you prefer "Twing Twang" music then a lighter suspension may sound better to you. If you like "Heavy Rock" or "Orchestra" then tighter suspensions may sound better.

Valve amps are better with more sensitive speakers. However, a good output transformer will be good at transforming the High Voltage/Low Current output of the valve into Low Voltage/High Current for the speaker.

You will notice that a lot of HIGH-END amplifiers are only around the 100W mark. However the output stages contain devices that could easilly cope with 10 times that power. The Power Supplies are also very over-rated.

My Linsley Hood 80W Power Amp can actually deliver nearly 17A peaks, that certainly overcomes any inertia problems in my B&W speakers.
Current drive amplifiers (where they're designed to stick a certain amount of current through, regardless of voltage) don't respond too well to peaky impedances of say, ported cabinets. Where the twin impedance peaks are, the amplifier still tries to deliver the same current as everywhere else. However, because the impedance is very high (can be 10x nominal), the volts are ramped up to try to push the current through. This leads to an "interesting" sound.

Voltage drive amps tend to care little about impedance spikes. It'll apply the volts, supply current when current flows. Of course, a very low impedance will destroy lesser amplifiers as they try to shove too much current through.

That's about as much as I know. If there's errors, let me know and I'll revise things.

Generally it doesn't matter how the watts are derived.
This is the OHM's Law: Ohm's law - Wikipedia, the free encyclopedia


and W=VxA where A (ampere) is actually "I" from the Ohm's Law.

Watt is a unit for power and when we refer to power, we always have in mind amount of work for a given amount of time.

It is the same with horsepower of an internal combustion engine - it is derived by the engine torque and engine speed in revolutions per minute.

1 Watt can be result of endless number of combinations, such as 10V and 0.1A or 10A and 0.1V or even 500A and 0.002V

At 100 mph on level road the air resistance of a vehicle can be overcome with not less than 50 (36.76kw) horse power and the air resistance doesn't care if it is achieved by a petrol engine with 2000 n.m. (1475 lbft) and 500 rpm or with 4000 rpm and 100 n.m. or even by a electric motor driven by 200V and 180A or with 50V and 735A...

If we know the impedance of the speaker and the voltage across it's voice coil, we can find the watts driving it at the moment.
For instance a 8 ohm speaker with 63 volts across the input terminals is being driven by 496 watts.

I=V/R so I=63/8, I=7.875A
W=IxV so W=7.875x63, W=496.125 - and that's how a normal amplifier works.

A current drive amplifier will keep the voltage the same and alter the current in accordance with the signal form. - I don't know how that happens without voltage rise when the load is a constant value? But my guess is that a current drive amplifier would supply greater dynamics than the voltage drive ones?

After all Watts do the work and everything other is just it's component.

So the answer would be: The speaker eats Watts.

And a more specific answer will be, that the speaker can not discriminate from volts or watts in theory. In the real world it would be very unpractical to move huge amounts of current through enormous cables or the opposite - to deal with life threatening voltages.
The issue for overcoming the initial resistance in order to be able to drive the voice coil does mot apply to materials which are conductors by nature, this only applies to dielectric substances such as plastic insulators or the human skin. - the human skin if it is dry does not conduct electricity with less than 30 volts no matter of the current.

Qurious Fact: a 100W into 8 ohms amplifier actually delivers around 28.3 volts and 3.5375 amps ;)
Last edited:
Oh if it were as simple as Ohms Law.

Speakers are far from simple resistors.

Even the most basic 2-Way effort from the local Flea Market will be a complex combination of Inductance (The windings of the voice coils), Resistance (The wire itself) and Capacitance (The crossover to the Tweeter).

More complex speakers will also have L,C and R in varying combinations.

This all plays havoc with the impedance and what the amplifier is trying to drive.

My speakers vary from 2 Ohms to over 500 Ohms over the main frequency spectrum. Just do a search on any popular speaker and look at its impedance / frequency graph.
Yes a speaker is definitely a complex load, but it all still works according the OHN's Law.

For instance if fed with 63V your speakers will be driven with no more than 7.938 watts in the 500 ohm impedance region and with not less than 1984 watts in the 2 ohm region of frequencies. (if frequencies with the corresponding amplitude are present in the signal of cource)

As a consequence you will have excessive cone motion in the 2 ohm region of frequencies and a hole in the SPL response in the 500 ohm region. And the amplifier will be unstable.

In practice impedance peaks correspond to something - too big separation in the crossover for instance or a resonance peak.

If it is the first there most probably would be a hole in the frequency response.

if it is the second the frequency response might be either even in that area or it can peak too... Resonance means that the natural retraction to 0 position speed of the driver suspension is equal to the speed it is driven, and due to that the coil presents lesser load to the driving current.
A lesser load in terms of electricity would be always a bigger resistance - less current draw. - The speaker resonates at it's natural frequency and the amp only adds up for the work done other than overcoming speaker's internal resistance.

Impedance dips correspond to voice coil inductance and better membrane to air coupling.

Here is a funny conclusion: a 8 ohm speaker with a resonance peak of 50 ohms at 40 hertz will draw only 8 watts at 40 hz if it is driven with 20V, even though that 20V into 8 ohms is 50 watts.
That means that on every similar, but non resonant frequency 42 watts are wasted only for overcoming of membrane inertia and suspension resistance...
42 out of 50 is 84% ;)
But we know that a speaker rarely has acoustic efficiency of over 3.5%...

We know where 84% of the power goes and where 3.5% of the power goes.

That leaves us to wonder where the other 10.5% of the power goes? - any clues?
note that depending on the impedance peak at resonance frequency and on the actual acoustic efficiency we might have less percents of power to wonder for...
A while ago I was asked to help out a D.J. who kept losing speakers.

After a bit of testing I concluded that the failures weren't down to clipping, nor were they down to be overdrived. The spec said that the speakers should take 500W RMS.

When I fitted thermometers to the speakers, the magnets got up to nearly 160 degrees C at only 300W. The problem in the end was the crossovers. The speakers were being asked to deliver something they just couldn't.


Let's no lose the poor guy that started this thread im detailed conversation. He's asking a simple question and we are now baffling him with science.


After all, out of the two sciences - the one about the loudspeakers (drivers) and acoustics, the one which is a real science is the Acoustics.

The knowledge of loudspeakers is an applied science or empirical science (don't know the exact idiom - sorry) because in real life you can't create a driver by only designing it with given abstract values characteristics and dimensions. - Actually you can, but the result will never be the predicted one.
It's a blend of a number of other sciences physics, electric engineering, wave behavior in solids, mechanics fluids and even acoustics... - most are branches of physics.

You can tell whether a certain science is an empirical/applied if there are a number of good professionals with certain even minor achievements in it who actually don't have even a fraction of the knowledge that describes the processes in the devices that they create or modify.

Such science is the internal combustion engine too.

In both cases experience and good observational skills are more valuable than just knowledge, of course logical thinking and intelligence are required too.
Hello all,

Excuse my ignorance, but I have often wonderes ( and haven't found the answer for) what drives a speaker?
I know it is watts...but what is the most substantial part of watts drives the speaker? Is it Voltage? or mainly amperage...(current).


Unless your talking exotic stuff there are two basic answers :

1) Loudspeakers are driven by voltage levels, generally the higher the
power of the amplifier, the higher the voltage it needs to produce and
hence generally internally the higher will be the two DC voltage rails.

2) At the same time the amplifier needs to be able to service the current
required for those voltage levels, and that is determined by the speakers
impedance. The lower the impedance the more current is needed. Some
amplifers will only drive down to 8 ohms, some down to 4 ohms, and
more often in cars some down to 2 or 1 ohms.

VxI = Power, so a 100W into 1R needs far less voltage than 100W into 8R,
100W into 1R also of course needs a lot more current than 100W into 8R.
As speakers are magnetic devices its the current that does all the work.
(1R coils will have ~1/8 of the turns of 8R coils, I x turns = the force.)

Speaker impedance basically determines the ratio of voltage and current
required for a given power level to be fed to a speaker. Car speakers
are low impedance because less voltage (but more current) is needed.

Neverless its the applied voltage levels that vary the output level, i.e.
most speakers are voltage controlled from a low impedance amplifier.

rgds, sreten.
Last edited:
Someone mentioned the peak in impedance would give a hole in frequency response.

I find this unlikely to be true. A peak in electrical impedance suggests the cone is very easily moved at that frequency (next to no current flowing, but lots of movement). Sounds like a resonant frequency to me.

Someone mentioned the peak in impedance would give a hole in frequency response.

I find this unlikely to be true. A peak in electrical impedance suggests the cone is very easily moved at that frequency (next to no current flowing, but lots of movement). Sounds like a resonant frequency to me.


Hi, its not true. With current drive you would get a huge peak, rgds, sreten.
Last edited:


2005-06-17 6:58 am
Drive with more current content is useful in OB speakers

I myself use high output impedance amp for OB speaker (with low Q woofer). The output impedance is a finite 20 Ohm, so I won't call it a current source (or a voltage source). I just need this number to get a proper 'system Q' (working together with the Re, thus Qes... etc.)

Eventually, the amp provides higher total gain in the high impedance region of the woofer. I got over 10dB of boost around fs (or, OTOH, it's also can be seen as attenuating such dB in other low impedance range because of more negative feedback by current content...)

In the end, the FR curve looks the same as driven by voltage amp with EQ. But interestingly, they sound different to me.

I don't have proper skill and equipment in measurement, or capabilty to analyse all these, so I got a lot of questions among them (even with satisfaction of the results with high Zo ones.... ).

Back EMF must be a very important role here. I've seen (and remembered) discussions in one or two sentences only, and sadly I don't fully understand.

I often feel it's a pity that (it seems) very few people use this method (high Zo amp) in their OB speakers, especially for bass. And because of this, I got very few discussions to learn, or very few suggestions or corrections.

I'm happy to see this topic was brought up. I hope to see some more brainstormings by you guys who are more knowledgeable than me :)

Thank you.
You can't have voltage without current or current without voltage, so it is hard to describe a speaker as responding more to one or the other.

If we are talking multiway loudspeakers then I want a low impedance or constant voltage amplifier. Without constant voltage the different sections interact which greatly complicates crossover design. Full range drivers are another matter.

If F=BLi and F=MA we see that constant current gives constant acceleration. Over much of a driver's range constant acceleration equates to flat response. I have a Japanese driver catalog that has response curves and admittance curve (impedance inverted) for dozens of drivers and there is a surprising correlation between the admittance curve and frequency response. Flat current gives flatter response and is the way to go with full range drivers.

David S.


2005-06-17 6:58 am
Well, I use high impedance amp for woofer only, in an active xover'ed system. So in this section, I can avoid the issues of multi-way passive xover.

It's almost a "perfect fit" to OB bass with low Q driver - automatically compensating the early low Q roll-off.

How does it compare to a system with 'normal' low impedance amp with high Q driver? I don't know. And I'd really like to know.