[Grimbles]

Hi all,

Impedance of the headphones i (think i) understand - it is the resistance provided by the headphone circuit. Higher impedance = harder to drive as more resistance (often thinner wire with more coils = higher resistance, which i think is how beyer do it with their multi impedance options).

I dont quite understand resistance at ouput jack though.

I'm also a little lost with sensitivity and its interrelationship with impedance.

Please can someone school me or point me to some decent reading?

Cheers!

 

[MindsMirror]

Higher impedance = harder to drive as more resistance (often thinner wire with more coils = higher resistance, which i think is how beyer do it with their multi impedance options).

This is not really accurate. It depends on how you define "hard to drive." That term is kind of ambiguous. If you have more turns on the coil, it will be higher impedance, but it will also be more effective, requiring more voltage but less current to effectively get the same result.

Sensitivity tells you how much power you need to deliver for the headphone to reach a certain loudness. If you define difficulty of driving a headphone as how much power it needs, sensitivity (not impedance) is what tells you that.

 

[Grimbles]

Hmm ok, now I'm a bit more confused, and I suspect you have hit the nail on the head which is I have made the assumption "harder to drive" means higher impedance!

So...
1) headphones with high impedance are "hard to drive" as they require a comparatively high voltage, although require less current. Given power=volts*amps, you can conclude they therefore need more power to be driven; and
2) headphones which have low sensitivity need more power to reach a given loudness?

Happy to be told I am wrong, but can we therefore conclude that from a consumer's perspective, "hard to drive" simply means a headphone requires more power (ie energy) to be driven. This can either be because of high impedance or lower sensitivity?

So how does output jack impedance relate? Thanks for your help so far!

 

[Strangelove424]

A good way to describe electricity is with a water hose. Let’s say when you increase the voltage on your amp, it is equivalent to turning the knob on a water outlet. This is the electrical potential for power. Power is a measure of the rate or pressure of the water coming out the end of the hose, or in our case the rate of electrical power used, a watt. The rate of water flowing through the hose is the current, or the amps. It is directly related to the thickness of the hose, the resistance against the current, measured in ohms. The current (amps), along with the initial potential (volts) results in the rate of work done (watts) at the end of the circuit. The output resistance of the amplifier is a ratio of it’s volts over current, or how much power is getting held back by the resistance of the circuit. A lower resistance means more efficient circuit, which is typically ideal, though can present problems for the noise floor of low impedance/high sensitivity IEMs.

 

[castleofargh]

Hmm ok, now I'm a bit more confused, and I suspect you have hit the nail on the head which is I have made the assumption "harder to drive" means higher impedance!

So...
1) headphones with high impedance are "hard to drive" as they require a comparatively high voltage, although require less current. Given power=volts*amps, you can conclude they therefore need more power to be driven; and
2) headphones which have low sensitivity need more power to reach a given loudness?

Happy to be told I am wrong, but can we therefore conclude that from a consumer's perspective, "hard to drive" simply means a headphone requires more power (ie energy) to be driven. This can either be because of high impedance or lower sensitivity?

So how does output jack impedance relate? Thanks for your help so far!

in general a headphone with high impedance will tend to require more voltage than current. and a headphone with low impedance will need more current than voltage. but power could be pretty much the same into 2 very different headphones with very different impedance values as power is current times voltage(at least at max amplitude in AC). that is why headphone specs don't stop at impedance but bother checking sensitivity too. and it's also why amplifiers give different max power depending on the load(headphone's impedance). the electrical formulas link everything together but it's more complicated than "high impedance headphone is hard to drive".
in fact it's wrong to think that way. the hardest headphone to drive is one with both low impedance and low sensitivity. meaning it will let a lot of current flow, and at the same time, it will require high voltage to sound loud enough. as power is current times voltage, that headphone will be the one asking for the most power. not the high impedance one.


as for the question about impedance at the amp output, you would need to learn a little about electricity for it to start making sense. the most basic aspect being that the ratio between the amp's output impedance and the headphone's impedance will determine the power efficiency of the circuit. I can't really tell much without losing you or saying something false depending on circumstances, so you'll have to go read some electrical stuff or look at video crash courses. but the energy used/wasted/dissipated in a circuit is not going to be 100% used by the headphone. we don't have a system that efficient. with an amp's impedance close to 0ohm or a headphone's impedance close to infinity, the headphone will use almost all the power spent(see "impedance bridging"). while if you take something like an amplifier with the same impedance as the headphone, then they will share the energy that the amp will mostly have to dissipate as heat, having no moving part, there are not many other ways to dissipate energy.

you probably didn't need to know that but here it is ^_^. in general most modern headphones and amplifiers expect to work plugged in a fashion that will follow impedance bridging rules. you can often read about the 1/8 or 1/10 damping ratio rule. where we want the amplifier's impedance to be at least 8 times smaller than the headphone's impedance. if you simply try to stick to that, you'll already be very fine for almost all audio combo you'll decide to create in your audiophile life.

 

[Grimbles]

@castleofargh thanks for taking the time mate. I still dont fully understand but you've laid the foundations for some directed reading. My unknown unknowns are unknowns now. Its half the way to wisdom!

Thanks also for your time @Strangelove424 and @MindsMirror - all has been illuminating.

 

[MindsMirror]

1) headphones with high impedance are "hard to drive" as they require a comparatively high voltage, although require less current. Given power=volts*amps, you can conclude they therefore need more power to be driven; and
2) headphones which have low sensitivity need more power to reach a given loudness?

Happy to be told I am wrong, but can we therefore conclude that from a consumer's perspective, "hard to drive" simply means a headphone requires more power (ie energy) to be driven. This can either be because of high impedance or lower sensitivity?

Let's try not to use the "hard to drive" term anymore, as I said, it is ambiguous and can mean several different things related to power, impedance, voltage, current, etc.

There is only one number you need in order to assess how much power the headphone needs. That is the sensitivity. Impedance doesn't even enter the equation. If you have a 16 Ohm headphone and a 600 Ohm headphone who's sensitivity is the same, they require exactly the same power. The lower the sensitivity, the more power is required of the amp.

If you compare these theoretical 16 and 600 ohm headphones with the same sensitivity, the 600 Ohm will require more voltage and less current, but still the same power. Higher voltage means that you will need to set the volume knob higher. This almost certainly leads to the misconception that higher impedance is "harder to drive." However despite having to set the volume knob higher, the amp is not doing more work, since the current is proportionally lower and the power is the same.

In some cases the amp may not be able to supply the required voltage, simply because the amp just doesn't have a high enough gain setting. You might say that this higher impedance headphone is more difficult to drive because it requires more voltage, but this definition of "hard to drive" related to voltage is completely different than the definition related to power. On the flip side, the theoretical 16 ohm headphone with the same sensitivity will require an amp with lower output impedance. This could lead to another completely different definition of "difficult to drive" related to output impedance.

 

[Niouke]

I'm as confused as you, or was until I read about speakers. Most speakers are between 4 and 8 ohms, so over there they only talk about sensitivity as a measure of how loud a speaker is compared to the power in watt you are feeding it, and that value is in practice independent of the impedance of the speaker.

 

[castleofargh]

well getting what the sensitivity value means in the first place is really just a matter of understanding how we get it. and it's the same for most measurements. learning how to do it is often more informative than learning how to read the results.
to measure sensitivity the idea is to play a 1khz tone, set the amplifier to output 1mW of power, and just measure how loud the sound goes at 1khz with a microphone placed in front of the headphone's driver. as straightforward as it gets. so the result is a loudness in dB, and the specs use the nomenclature XXX dB/mW @1khz. which is telling the same thing. the headphone goes XXX dB loud for 1mW of power at 1khz.

you can also go the other way around, set the volume level of the amp while playing your 1khz test tone until you measure 90dB SPL from the mic, and then go measure how much power or voltage you're sending. then it's a simple enough game to estimate how loud we'd get for 1mW of power. and the cool thing with that second method is that we make sure not to blow up our headphone/speaker ^_^.


so you see that you don't have to care about the impedance here. it's interesting to have it for when you go look at amplifier's specs. the amplifier will usually give a max power available without too much distortion when plugged into a specific load. let's say the manufacturer's specs give for an amplifier:
15mW per channel into 30ohm for THD<1%. (and yes they usually measure all that at 1khz too, that way we know where we are going).
86mW per channel into 150ohm for THD<1%.

that's when you need the impedance of your headphone. the "into 30ohm" is saying that if you plug a 30ohm headphone, you can expect to get up to 15mW of power before the signal goes to crap itself. ^_^ and as you know how loud you'll go with the headphone for 1mW of power, thanks to the sensitivity spec, you can get an idea of how loud that specific amp will let you go into that headphone. (you need to go look at the formula and it requires the use of log, but to give an idea, twice the power will get you 3dB louder).
and obviously if you have a high impedance headphone, then you'll want the amp's spec for a load closer to your impedance value instead of the power output into 30ohm, well duh! thanks you captain obvious ^_^.
and that's is why the impedance value of your headphone is not irrelevant at all, even for power considerations. it's just not as simple as saying high impedance is hard to drive, which once again is not necessarily true.


does that make more sense? I should have started here, sorry, and good call from @Niouke to move the discussion that way.

 

[Speedskater]

High impedance loads have greater voltage demands.
While low impedance loads have greater current demands.

But with other things being equal, loudness or power comes down to Voltage times Current.

And Impedance, Voltage, Current and Power are all locked together. You can't change just one.

Current is the dependent variable. That is, you can't just turn-up the Current.

 

[Niouke]

now we have the basics set, lets consider multi BA IEM's whose impedance can vary from 10 to 150ohms depending on the frequency of the signal, I can't wrap my head around that one lol

 

[castleofargh]

now we have the basics set, lets consider multi BA IEM's whose impedance can vary from 10 to 150ohms depending on the frequency of the signal, I can't wrap my head around that one lol

because impedance isn't just resistance, it's the result of all the forces opposing the current, including resistance. the circuit reacts(literally) to the electrical signal sent into it. some headphones have a dead flat impedance response over frequencies because they mostly behave like a stable resistor. other designs like multi BA IEM are electrically more complicated, just take the crossovers. they use electrical components which make the all IEM to be more than just a coil. and then the very technology of a BA driver is a little special and leads to more or less general behaviors. a single BA driver will tend to have high impedance at high frequencies for example. something you don't see much in a dynamic driver. their physical differences end up making them different electrical components with different behaviors. the end result is that some frequency signal will "pass through" easily, while for another frequency the current will not and that's the impedance getting higher for that frequency.

I hope I'm not saying anything stupid because I'm really hitting the oversimplification wall here. people let me know if I'm talking nonsense please.
@Niouke , if you want to really go mad, consider that the measured impedance of your IEM is probably a little different when the IEM is in your ear instead of being unsealed and lying there on the table

 

[MindsMirror]

now we have the basics set, lets consider multi BA IEM's whose impedance can vary from 10 to 150ohms depending on the frequency of the signal, I can't wrap my head around that one lol

In reality, any headphone including a multi BA IEM will have a sensitivity response which varies with frequency due to electrical and physical properties. This is very similar to the frequency response, except that you usually use a constant voltage for get the frequency response, where as for sensitivity you use constant power. Also frequency response is a unit-less measurement, it is only defined relative to other frequencies, where as sensitivity is given with an absolute unit of pressure.

Most consumers don't need or want that much information, so the manufacturers only give it to you at one frequency. Fortunately they all seem to agree upon using the same frequency as each other at 1KHz.

 

[Grimbles]

well getting what the sensitivity value means in the first place is really just a matter of understanding how we get it. and it's the same for most measurements. learning how to do it is often more informative than learning how to read the results.
to measure sensitivity the idea is to play a 1khz tone, set the amplifier to output 1mW of power, and just measure how loud the sound goes at 1khz with a microphone placed in front of the headphone's driver. as straightforward as it gets. so the result is a loudness in dB, and the specs use the nomenclature XXX dB/mW @1khz. which is telling the same thing. the headphone goes XXX dB loud for 1mW of power at 1khz.

you can also go the other way around, set the volume level of the amp while playing your 1khz test tone until you measure 90dB SPL from the mic, and then go measure how much power or voltage you're sending. then it's a simple enough game to estimate how loud we'd get for 1mW of power. and the cool thing with that second method is that we make sure not to blow up our headphone/speaker ^_^.


so you see that you don't have to care about the impedance here. it's interesting to have it for when you go look at amplifier's specs. the amplifier will usually give a max power available without too much distortion when plugged into a specific load. let's say the manufacturer's specs give for an amplifier:
15mW per channel into 30ohm for THD<1%. (and yes they usually measure all that at 1khz too, that way we know where we are going).
86mW per channel into 150ohm for THD<1%.

that's when you need the impedance of your headphone. the "into 30ohm" is saying that if you plug a 30ohm headphone, you can expect to get up to 15mW of power before the signal goes to crap itself. ^_^ and as you know how loud you'll go with the headphone for 1mW of power, thanks to the sensitivity spec, you can get an idea of how loud that specific amp will let you go into that headphone. (you need to go look at the formula and it requires the use of log, but to give an idea, twice the power will get you 3dB louder).
and obviously if you have a high impedance headphone, then you'll want the amp's spec for a load closer to your impedance value instead of the power output into 30ohm, well duh! thanks you captain obvious ^_^.
and that's is why the impedance value of your headphone is not irrelevant at all, even for power considerations. it's just not as simple as saying high impedance is hard to drive, which once again is not necessarily true.

Now this makes sense to me. As the measurement conditions are (practically) constant -1khz signal, 1mW, measure the dB, i understand why this is such a useful number. On this basis, what is the rough scale of low to high sensitivity?

 

[castleofargh]

yes, following standards always helps, it's not fixing everything, after all it's only about 1frequency and considering a fully resistive load which headphones often aren't. and sadly most manufacturers provide too little information for even that to be relevant(like power into only 1 load and often not even specifying which impedance it was), or values without a clear unit are overused marketing tricks to make it look like something is more than it really is.
but all in all those who follow the standards provide specs that are at least indicative of how inadequate a source might be for your specific device.

as for what's high sensitivity, that's not really relevant and people will disagree about it anyway. the right sensitivity is the one that let you get loud enough when you wish to go loud while using a given amplifier. so it's a circular question where the sensi is always right as long as you pick the right amp for it. for IEMs, stuff that can go 120dB loud with 1mW can clearly lead to issues like getting audible hiss from the source, and there is no device that is limited to 1mW anyway so it doesn't serve a purpose and I would personally try to avoid it (everything else being equal, which never happens). my weakest source is a sony DAP that goes at about 10mW per channel into low impedance. and I don't think I've ever owned anything weaker.
for headphone there is a sort of tradition to aim for an amplifier that can push your headphone at around 115dB. it makes for really loud listening sessions, but most of all it allows for some margin of error in more typical listening conditions. so instead of a requirement, it's a more like a very safe reference.