[GermanGuy]

Quote:

you don't get something for nothing... higher impedance means more wire, more wraps... which means more magnetic flux... which means more dynamic response given the same mass of movement... means better tighter sound with more headroom.
 
think about it this way, why do sports cars have more powerful engines? it is all about no compromise performance

the main benefit of higher or high impedance coils is
- that you can use a thin but longer or in general a longer wire 
- and so the coil can be designed not very flat but with greater height or more wraps
- so the magnetic field of the coil produced by the flow of current has a better linearity (in the area of the magnet pole)
- result can be a better linearity of the movement of the driver
- a  greater max. moving width of the driver (in german  language called 'Hub', the maximum length or distance the driver diaphragm can move from the zero positon without being in the nonlinear area of the magnetic field
- better linearity for high levels

 

[EthanWiner]

Thanks Saleri for moving this to the public forum. I told you you'd get good replies!

LowZ phones are needed by modern low voltage chip sources.  Studios that use possibly several phones driven at once prefer them to be HiZ to not load down the source if more than one phone is used.  Old tube equipment had higher voltages and low z phones would have not worked well with the higher voltage.

This is the correct answer. Hi-Z phones require a higher voltage but draw less current. Lo-Z phones are the opposite, and so can go louder when driven by battery-powered devices that can output only a limited amount of voltage.

Other than the trade-off between voltage and current, I can't see any reason why the fidelity per se would be different between Hi-Z and Lo-Z headphones or ear buds. But I'm not a headphones expert so I defer to others for a technical explanation.

--Ethan

 

[esldude]

Quote:

 
That is not universally true, there are sound cards that can output 7 Vrms voltage into a high impedance load without clipping. That is normally enough even for 600 Ω headphones. Also, it is common for the output impedance to be in the range 10 to 100 Ω. If it is really high, the maximum power output into a 32 Ω and a 250 Ω load may actually be comparable.

Well nearly nothing is universally true.  I posit that far more often than not, a large majority of sound cards work with limited voltage. And therefore lack enough voltage to drive high impedance phones to their potential.  

 

[mikeaj]

Quote:

Well nearly nothing is universally true.  I posit that far more often than not, a large majority of sound cards work with limited voltage. And therefore lack enough voltage to drive high impedance phones to their potential.  

 
Well, +12 V is abundantly available in desktop computers. The cost of some mechanism for -12 V (yeah, it's available on PCI slots, but I don't think people use it these days especially with PCIe more the norm) shouldn't be too exorbitant on sound cards costing $100+. You see some of these swinging 3 V rms or up to about 7 V rms.
 
If you're talking about onboard sound chips, I think those run off of 5 V (for the analog power input) and ground only, sure. Might be down to 3.3 V on some systems or maybe something else.

 

[Hellbishop]

Fantastic thread with some eye opening info. This probably explains why my low impedance Ultrasone HFI-780 don't sound as good on my Onkyo TX-8555 stereo receiver as they do from the headphone out of my Samsung MediaOne 3a monitor speakers which come with a built in amp. I have noticed the same thing with my Audio Technica Ath AD700 even though it still sounds great on the receiver with the monitor speakers there is an extra layer of clarity like a steamy bathroom that has just had its windows opened letting in fresh air.
 
Looks like from here on out its the stereo receiver for the high impedance cans and amped active speakers with a headphone out etc for the low impedance cans.
 
Thanks for the amazing read.

 

[Chromako]

I'm ignoring output impedance matching and advanced amplifier design and whatnot as that's getting into electrical engineering and physics I don't know. But I do know some stuff about mechanics, so I'll contribute this (and this is for the layperson and my own sanity, so I'm simplifying it. But you can get the picture).
 
If we look at things in most (maybe pretty much all) cases, audio drivers work better if they are lighter (they can respond faster and more accurately- it's about inertia- That's also why we love electrostatics- which I'm ignoring today). Well, maybe not piezoelectrics and plasma drivers, but let's not go there.
 
How do we make them lighter? The major moving parts with respect to mass are the cone and electromagnet attached to it (voice coil).
 
So, as engineers of dynamic headphones, we have made the cone as light as possible by using nano-crystalline membranes grown in distilled baby tears, that are then cryo treated with liquid helium-4 (citation needed). But how do we make the electromagnet lighter? We already are also using nice materials like copper and aluminium and iridium plated dog-hair in the wire, so on we go. There's two major options for lightening this electromagnet:
 
1: Thinner wires.
2: Fewer turns.
 
Thin wires have higher electrical resistance. That's why you want them to be thick if you can to drive them easily (to a limit), and why your small desk lamp can have thinner power cables than a vacuum cleaner's. 
 
So let's go all out, screw the ability to drive them easily- we're making expensive stuff so we can anticipate that the customer will be okay with any drawbacks, in the pursuit of quality and perfection, and good reviews on Head-fi.
 
So, we'll use thin wires. That increases electrical resistance. The Sennheiser HD800 voice coil wires are something  on the order of 80 microns in diameter, give or take. That makes for an awful lot of resistance. We could have used thicker wires to reduce resistance and make the headphones easier to drive, but that would be compromise (so we could run them off of an iPod or something). Never compromise when your customer has money!
 
We can also reduce the length of wire by making fewer loops in the voice coil. We will, however, have to drive more electricity through to get the same magnetic flux as we would with a normal set of headphones- that is, the same amount of magnetism and therefore mechanical power available to move the cone. Fewer turns: we now have less sensitivity, which will demand more voltage out of our amplifier. Again, no compromises!
 
That's why our desktop amplifiers have to drive something on the order of 15 volts. Cheaper headphones are designed to work with iPods and such, which put out something like 1.2 volts. Those, however, require thicker wires and more turns to drive an equivalent cone. That's a compromise, and this is Head-fi! 

 

[stv014]

Quote:

Originally Posted by Chromako /img/forum/go_quote.gif
 
1: Thinner wires.
2: Fewer turns.

 
Both of these trade mass vs. efficiency (the amount of sound pressure from a given input power). 1. increases the impedance, while 2. decreases it, but both make the efficiency worse (and also make the driver benefit less from electrical damping). If you use both thinner wire and fewer turns, then the impedance of a light voice coil can remain the same, but the efficiency will be poor (the AKG K70x might be an example of this). To prove that option 1 (high impedance voice coil with thin wire) is better, it would need to be shown that it can achieve the same efficiency with lower mass, and preferably the amount of improvement should also be quantified.
Other factors like better design or a stronger magnet may make a greater difference, however.

 

[jcx]

didn't read the 1st page? http://www.head-fi.org/t/670303/why-do-we-have-amps-and-headphones-with-large-resistance-levels#post_9572602
 
changing turns and wire diameter to keep the same total conductor mass, voice coil geometry, cross section in the (same, of course) magnet gap is the same as adding an ideal transformer == no efficiency change
 
sensitivity in dB/mW will be unchanged, but dB/1V does change with the Z
 
so a manufacturer offering different Z in the same model number at the same sensitivity in db/mW has followed this principle - this is a of basic motor design
 
Thiele-Small parameters of the drivers look like an ideal transformer were included in the electircal side - and the bass resonance, the Q,s, BL product, efficiency are all the same
 
 
changing voice coil mass does change everything - constiutes a new motor design - not just a change of the Z

 

[Chromako]

Quote:

 
Both of these trade mass vs. efficiency (the amount of sound pressure from a given input power). 1. increases the impedance, while 2. decreases it, but both make the efficiency worse (and also make the driver benefit less from electrical damping). If you use both thinner wire and fewer turns, then the impedance of a light voice coil can remain the same, but the efficiency will be poor (the AKG K70x might be an example of this). To prove that option 1 (high impedance voice coil with thin wire) is better, it would need to be shown that it can achieve the same efficiency with lower mass, and preferably the amount of improvement should also be quantified.
Other factors like better design or a stronger magnet may make a greater difference, however.

 
Indeed. But if we expect  the consumer to use powerful amplification, we don't have to worry about balancing efficiency so much. It simply wouldn't be a problem (if one spends 1k USD on a headphone, we can probably anticipate that they will have a nice amplifier anyway).
 
Then again, if we aren't careful and pull this screw-efficiency trick with less expensive phones, then we end up risking angering customers in that price range and making them angry. Then the angry customers go out and murder kittens. Kinda like how a few AKG people got their K70x's and found that they sounded like garbage with their un-amped setup (and then in desperation they came to Head-fi). 
 
Of course, you're very correct about other innovations. Like how Beyerdynamic did with their Tesla ubermagnet and Sennheiser did with their ring driver voodoo. But why not use many strategies? HD800's are also rather hard to drive, and the T1's don't make it easy. 
 
Speculation: Beyer's DTx80's have pretty decent sensitivity but very high resistance (600 ohms in one of the variations), so they may have gone with the thin wire strategy, while Denon Dx000's have low resistance but low sensitivity (so may have gone with fewer turns). Audio-Technica's A900's sound fantastic without dedicated amplifiers, but they definitely sacrificed some potential (I'm not saying that it's a bad thing).

 

[stv014]

There is also a question of possibly diminishing returns when trying to reduce the mass of the voice coil too much, especially since the other moving parts still remain the same. Even if efficiency is ignored, it will have disadvantages like reduced electrical damping, and reduced power handling (i.e. a thinner but otherwise identical voice coil will not only be less efficient, but also overheat from less power).
 
The DTxx0 drivers are in fact not particularly efficient. They may seem to be more efficient than expected when someone first tries them after reading all the exaggerated amplification requirements (that is, reading other forums on Head-Fi, one would never expect that a 250 Ω headphone can actually sound good and loud enough from a laptop or similar device), but 96 dB/mW - which can in reality be only 93-94 dB/mW depending also on random driver variation - is mediocre compared to the 100-105 dB/mW of modern headphones that are designed to be powered by portable sources.
Interestingly, all versions of the DT880 measured at InnerFidelity seem to have about the same efficiency, with only minor variation.

 

[jcx]

"thin wire" is a meaningless design criteria as I have been apparently failing to explain
 
BL product can be changed with more/less conductor cross section in the same magnetic field/gap/voice coil geometry - but how you divide it up in thin wire/more turns or thick wire/few turns is a "free" design choice - doesn't impact "speed", efficiency - lets the designer choose Z independently

 

[stv014]

Quote:

"thin wire" is a meaningless design criteria as I have been apparently failing to explain
 
BL product can be changed with more/less conductor cross section in the same magnetic field/gap/voice coil geometry - but how you divide it up in thin wire/more turns or thick wire/few turns is a "free" design choice - doesn't impact "speed", efficiency - lets the designer choose Z independently

 
Where did I disagree with that ? I only mentioned an efficiency vs. mass trade-off (e.g. thinner wire but not more turns).

 

[jcx]

using "thinner wire" as a proxy for reduced cross section/mass of voice coil conductor is probelmatic - confuses the principle of Z being a "free variable" with your changing the BL product with less total conductor cross section, reducing voice coil mass
 
"thinner wire" is ambiguous - lots of head-fier aren't educated in this field but are very bright - will try to push "semantic analysis" of words beyond the engineering aplicability
 
I think this is a especially needed distinction in light of headphone manufacturers offering the same model with different Z, same sensitivity - and the resulting speculation over their purported different "sound"

 

[Chromako]

Quote:

using "thinner wire" as a proxy for reduced cross section/mass of voice coil conductor is probelmatic - confuses the principle of Z being a "free variable" with your changing the BL product with less total conductor cross section, reducing voice coil mass
 
"thinner wire" is ambiguous - lots of head-fier aren't educated in this field but are very bright - will try to push "semantic analysis" of words beyond the engineering aplicability

I do realise that my wording wasn't the most precise... I did the best I could.
 
I was only intending to offer a contribution to answering the question... (did I say that, I think?)  certainly not the entire one. And I said that I have no clue how this applies to amplifiers... 
 
In any case, I'm not going to be so arrogant to say that I understand all the implications... my knowledge is naturally incomplete. 
 
The issue is that I haven't been educated (or learned) in anything more than the most basic aspects of electrical engineering and physics, and I'm having a very difficult time making sense of the jargon you're using, but I really want to understand the basics. And my god, Wikipedia is writing their articles with the assumption that the reader already understands the concepts, which in my opinion is completely missing the point. Is there a way you can explain some of it to us luddites?  
 
The extent of my understanding of amplifier Z is that my DT880's, T1's and HD800's sounded awful on my High Z amp output but are fantastic on the low Z one (The Corda Opera has both options), and that High Z outputs is good for K701's for some reason I don't understand except that it was engineered that way by more educated people. And someone told me that output resistance might have something to do with damping factors.  But I really don't know much of anything about what this Z output thingy is, except that Z has something to do with output resistance and is measured in ohms. And that interconnects jacks have really high Z. No clue why that matters. 
 
For example, I really don't understand your first paragraph's use of "Free variable" and "BL product," or how your concerns of "semantic analysis" apply in this particular case (I know what it means, just not how it applies here)?  

 Or how wires with smaller cross sections don't really contribute anything to the original question, as I though that everything else being equal, wires with smaller cross sections (what I meant with "thinner" but didn't have the term on hand) does increase resistance? Or do they? My head is spinning. 
 
After all, I'm an audiophile who is a psychology and biomaterials researcher... This amplifier design stuff certainly is far from my usual stomping grounds.  

 

[stv014]

Quote:

using "thinner wire" as a proxy for reduced cross section/mass of voice coil conductor is probelmatic - confuses the principle of Z being a "free variable" with your changing the BL product with less total conductor cross section, reducing voice coil mass
 
"thinner wire" is ambiguous - lots of head-fier aren't educated in this field but are very bright - will try to push "semantic analysis" of words beyond the engineering aplicability

 
Admittedly, English is not my native language, but I do not see how "thinner wire" by itself means anything other than "lower cross section". It is entirely your assumption that for others it also implies increased length (so that the mass remains the same), which is a straw man (i.e. "stupid people argue high impedance is better, let's prove them wrong by debunking claims they did not make").