Speaker Impedance - please confirm my interpretation

[NJB]

I have 2 sets of speakers, they are quite different. Both have 86dB sensitivity but one set plays considerably louder than the other (volume set in the same place). One set is 8 ohm and the other is 4 ohm. Given that power equals the square of current times resistance, then the 4 ohm set need more current to generate the same volume. I presume that is why low resistance speakers are called a more difficult load for an amplifier; more current required. However, it is the easier 8 ohm speakers that play quieter.

i am curious to know if my ancient physics is wrong, or whether this is what people would expect? Is it the fact that the speaker impedance, visible as a resistance across the amplifier output, alters the amplifier performance in other ways?

 

[lindsayt]

That's totally what I'd expect. For a given setting on the volume control, the 4 ohm speaker will suck more current from the amplifier.

That's why with speaker design you often get 4 ohm bass drivers mated with 8 ohm tweeters. Tweeters are naturally more efficient than bass drivers. So they compensate by having a lower impedance on the bass drivers to suck more current to boost their loudness up towards the level of the tweeters.

What it also means is that for a given setting on the volume control, especially up near clipping levels, the amp will be delivering more power and therefore having a harder time delivering the power with the 4 ohms speaker than the 8 ohm.

The difference between the 2 speakers should be about 3 dbs for 4 ohms vs 8 ohms at any given volume setting.

Also, manufacturers can be prone to exageration of their efficiency specifications, and the impedance figure is a nominal average. With every speaker impedance varies a lot with frequency, with each speaker design being different.

 

[SteveR750]

Makes sense to my rusty O level physics:

If both speaker have the same power sensitivity ratingh at 86dB for 1W at 1 metre.

I'm assuming that at the same volume position (all other things being equal) there is the same applied voltage to both speakers. The 4 ohm pair are drawing more current, therefore more power from the amp. At the same voltage, the lower impedance speakers are operating at twice the power of the 8 ohm pair.

since V=I*R and W=V*I

Imagine you set 16V output on your amp - this will result in a 2A current in the 8 ohm speaker, and 4A in the 4 ohm.

This means 32W power into the 8ohm and 64W into the 4 ohm load. It's not surprising they sound much louder.

 

[davedotco]

To be honest, the difference in percieved loudness is probably down to other factors. The way the speaker manufacturer manipulates the sensitivity issue mostly, almost everyone quote sensitivity figures that have been measured in a way to make them as high as possible, does not always translate well into the real world.

It is unlikely to be a power issue as steve descibes, speaker impedances vary wildely and the 4 or 8 ohm figures given are pretty nominal, highly unlikely that the 8 ohm speaker will be twice the impedance of the 4 ohm model over more than a small portion of the frequency range. Anyway budget amplifiers are not going to double their current into half the impedance either so the overall increase in the power fed to the 4 ohm speaker is likely to be a lot less than double, certainly nothing like enough to make a really noticeable difference.

Other factors may come into play such as distortion, a modest amount of which will make a speaker sound louder but in reality it is mostly the manufacturers 'fudging' the figures for marketing reasons.

 

[Vladimir]

To my knowledge 32W at 8 ohms and 64W at 4 ohms are of equal loudness at the same voltage, despite curent consumption doubling.

 

[andyjm]

To my knowledge 32W at 8 ohms and 64W at 4 ohms are of equal loudness at the same voltage, despite curent consumption doubling.

I disagree.

Power into a speaker generates useful work by moving the voicecoil producing sound waves and also wasted heat in the voicecoil. If you double the power into the speaker and the sound level remains the same you must be dissipating the extra power in heat in the voicecoil. This makes no sense.

 

[Vladimir]

Ah. The voice coil EMF is determined by the amount of current in it and that determines loudness. Got it.

 

[Covenanter]

Mmm!

Amplifiers tend to have a fixed voltage rail which is the maximum voltage that can be delivered across the load. The power in a circuit is equal to Voltage (V) x current (I) and current is related to voltage by Ohm's Law V=IR so power is equal to V x V / R. So the maximum power output will be twice as high for a 4 ohm speaker as an 8 ohm speaker, BUT it's more complicated than that as (a) amplifiers can't necessary deliver the current to make the maths work and (b) it isn't necessarily the case that all speakers translate electrical energy into sound energy equally.

Chris

 

[Vladimir]

Depends on how much power is lost as heat in all the components where the signal passes, I reckon. Technically amplifiers that double down at half the impedance load are non-existant. If the power loss is small, manufacturers simply declare lower power output at 8 ohms to make it half the 4 ohms output.

 

[Thompsonuxb]

The amps output is fixed. It will not deliver anymore or any less than its specified to at a given level.

It either has enough continuous juice to drive the speakers or it won't.

(These are not intelligent devices)

The speaker is were things change across the crossover - the P/V/C - has already noted.

Lower the resistance more power feed to the drivers from the incoming signal.

 

[Vladimir]

The voltage is fixed but current varies with speaker impedance, which varies with frequency. You can generate 15kV just by walking across your carpet, but that voltage has very very low current and will not power anything.

 

[Covenanter]

Depends on how much power is lost as heat in all the components where the signal passes, I reckon. Technically amplifiers that double down at half the impedance load are non-existant. If the power loss is small, manufacturers simply declare lower power output at 8 ohms to make it half the 4 ohms output.

Yep, I was simplifying and hence my comment after the BUT about amps not necessarily being able to suppy the current.

I looked up my amp and it can generate 70W into 8ohms and 100W into 4 ohms so you can see how much it falls away from the linear relationship.

Chris

 

[lindsayt]

To be honest, the difference in percieved loudness is probably down to other factors. The way the speaker manufacturer manipulates the sensitivity issue mostly, almost everyone quote sensitivity figures that have been measured in a way to make them as high as possible, does not always translate well into the real world.

It is unlikely to be a power issue as steve descibes, speaker impedances vary wildely and the 4 or 8 ohm figures given are pretty nominal, highly unlikely that the 8 ohm speaker will be twice the impedance of the 4 ohm model over more than a small portion of the frequency range. Anyway budget amplifiers are not going to double their current into half the impedance either so the overall increase in the power fed to the 4 ohm speaker is likely to be a lot less than double, certainly nothing like enough to make a really noticeable difference.

Other factors may come into play such as distortion, a modest amount of which will make a speaker sound louder but in reality it is mostly the manufacturers 'fudging' the figures for marketing reasons.

Oh yes they will!

As long as the budget amplifier is not clipping into the 4 ohm load it will deliver twice the current into it, for a given setting on the volume knob.

The vast majority of people with budget amps reading this, will spend the vast majority of their time listening at sub clipping levels, even with 4 ohm speakers.

The impedance vs sound levels thing is simple to prove. Get two identical speaker drivers. Play a test tone through one of them and measure the output with a sound pressure meter. Keep the volume knob at the same position. Now place the 2nd driver next to the first and wire them in parallel (switching off the amplifier whilst you're changing connections). The test tone will be about 3 dbs louder. Now wire them in series and the test tone will be about 3 dbs quieter than with the single dirver.

 

[andyjm]

To be honest, the difference in percieved loudness is probably down to other factors. The way the speaker manufacturer manipulates the sensitivity issue mostly, almost everyone quote sensitivity figures that have been measured in a way to make them as high as possible, does not always translate well into the real world.

It is unlikely to be a power issue as steve descibes, speaker impedances vary wildely and the 4 or 8 ohm figures given are pretty nominal, highly unlikely that the 8 ohm speaker will be twice the impedance of the 4 ohm model over more than a small portion of the frequency range. Anyway budget amplifiers are not going to double their current into half the impedance either so the overall increase in the power fed to the 4 ohm speaker is likely to be a lot less than double, certainly nothing like enough to make a really noticeable difference.

Other factors may come into play such as distortion, a modest amount of which will make a speaker sound louder but in reality it is mostly the manufacturers 'fudging' the figures for marketing reasons.

Oh yes they will!

As long as the budget amplifier is not clipping into the 4 ohm load it will deliver twice the current into it, for a given setting on the volume knob.

The vast majority of people with budget amps reading this, will spend the vast majority of their time listening at sub clipping levels, even with 4 ohm speakers.

The impedance vs sound levels thing is simple to prove. Get two identical speaker drivers. Play a test tone through one of them and measure the output with a sound pressure meter. Keep the volume knob at the same position. Now place the 2nd driver next to the first and wire them in parallel (switching off the amplifier whilst you're changing connections). The test tone will be about 3 dbs louder. Now wire them in series and the test tone will be about 3 dbs quieter than with the single dirver.

Indeed. Hifi amps are voltage amplifiers with very low output impedance. As long as the amp has sufficient headroom, halving the load impedance will double the current.

 

[davedotco]

Whilst this is correct in theory, power supply limitations in budget amplifiers will mean that this is not the case in a real world system other than at very modest output levels.

 

[TrevC]

Whilst this is correct in theory, power supply limitations in budget amplifiers will mean that this is not the case in a real world system other than at very modest output levels.

It will still be more than the maximum wattage output available at 8 ohms though, so perhaps not really all that modest at all.

 

[Vladimir]

Can someone name an amplifier that exactly doubles its output at half the impedance - measured, not manufacturer declared?

 

[TrevC]

Can someone name an amplifier that exactly doubles its output at half the impedance - measured, not manufacturer declared?

I name it George.

 

[Laurens_B]

Devialet 120 and Devialet 200 from what I've seen

 

[Vladimir]

I name it George.

As you should.

Devialet 120 and Devialet 200 from what I've seen

Do you remember the magazine/blog/lab that did the measurements?

 

[unsleepable]

I disagree.

Power into a speaker generates useful work by moving the voicecoil producing sound waves and also wasted heat in the voicecoil. If you double the power into the speaker and the sound level remains the same you must be dissipating the extra power in heat in the voicecoil. This makes no sense.

Different speakers must perform different amounts of work to generate the same loudness—that makes lots of sense. For example, in order to improve bass in a small speaker, sensitivity must be reduced, while larger speakers need to work less to generate the same bass at the same loudness.

A simple way of looking at this is that the amplifier is telling the speakers with voltage what to do, while the speakers tell the amplifier with impedance variation how much current they need to do what they are told.

Sensitivity is usually given in reference to volts or to watts. Two speakers with the same sensitivity should sound equally loud when fed an input signal with same volts or watts—depending on how the sensitivity is specified, and as long as the amplifier can provide the required current. When sensitivity is specified in volts, then the speakers should sound equally loud at the same position of the volume pot.

All this is of course in theory. In practice, sensitivity values given by vendors are rarely precise, and there is also perceived loudness and other factors to take into account.

 

[andyjm]

Vlad,

Not sure your point.

As long as the amp has headroom, it can be modelled as a perfect voltage source in series with an output impedance. As most amps use negative feedback, the effective output impedance is extremely low, in the order of 0.1 ohms or less. In this case, halving the load impedance will double the current.

There will come a point where the output drivers are saturated, the power supply is flat out and the amp just can't deliver more current. At this point, halving the load impedance won't double the current.

For an amp comfortably within its limits, the half / double relationship holds true.

 

[Laurens_B]

I name it George.

As you should.

Devialet 120 and Devialet 200 from what I've seen

Do you remember the magazine/blog/lab that did the measurements?

I was wrong, I saw the Devialet 240 measurement, not th 120/200, but I think it is safe to suppose they are quite similar in performance. Link below. It is a Polish website, and I don't know any Polish myself, but using google translate you can actually understand what is said. Scroll all the way down for the measurement results.

http://audio.com.pl/testy/devialet-240:2

 

[SteveR750]

Mmm!

Amplifiers tend to have a fixed voltage rail which is the maximum voltage that can be delivered across the load. The power in a circuit is equal to Voltage (V) x current (I) and current is related to voltage by Ohm's Law V=IR so power is equal to V x V / R. So the maximum power output will be twice as high for a 4 ohm speaker as an 8 ohm speaker, BUT it's more complicated than that as (a) amplifiers can't necessary deliver the current to make the maths work and (b) it isn't necessarily the case that all speakers translate electrical energy into sound energy equally.

Chris

And few if any speakers have a flat impedance curve, which explains why active speakers sound so much better, and why trying to obtain a ruler flat frequency response in an amp is at best a compromise. I'm guessing this is how the devialet SAM process works by assuming a standard impedance curve for a given speaker model and tailoring the output accordingly? Maybe not, as I haven't read up on it.

 

[Vladimir]

Vlad,

Not sure your point.

As long as the amp has headroom, it can be modelled as a perfect voltage source in series with an output impedance. As most amps use negative feedback, the effective output impedance is extremely low, in the order of 0.1 ohms or less. In this case, halving the load impedance will double the current.

There will come a point where the output drivers are saturated, the power supply is flat out and the amp just can't deliver more current. At this point, halving the load impedance won't double the current.

For an amp comfortably within its limits, the half / double relationship holds true.

No problem with that. I'm just curious in extreme conditions what amps out there come to the theoretical line of doubling their output at half impedance load. Check Laurens' link of that Devialet. Impressive measured performance, would you agree? (lab info at the bottom)