Cable length impedance and headphone amplifier

[axero]

I'm wondering when connecting a pair of headphones to a headphone amplifier through a 5 meters (16.5 feet) extension cord if the cord significantly adds to the impedance of the phones and roughly by how much.

So my main question is, how much impedance do you reckon a 3.5mm stereo jack/plug has per length unit if we ballpark here? It seems nigh impossible to find relevant data through google.


I'm using a pair of 250 Ohms headphones paired with a headphone amplifier that can be switched between driving either a pair of <64 Ohms, 64-300 Ohms or 300-600 Ohms headphones.

The 300-600 Ohms mode sounds a lot better than the lower modes in these headphones, even at lower volumes. Isn't it strange that it also sounds better at lower volumes? I thought the differences merely is about level of amplification.

Also, would you say that I would risk breaking the 250 Ohms headphones when driving them with an output rated for 600 Ohms?

[Lightages]

Driving headphones is no different than driving speakers. The output impedance of an amplifier and the wire to the speakers should be less than 1/10 of the lowest impedance of the speaker drivers, as a rule of thumb. The lower the impedance the closer the speakers/headphones will perform to spec. and the lower the better. 1/100 and lower is more typical.

A headphone amplifier that has selections to drive different impedance headphones is just snake oil. Changing the output impedance of the amp can change the sound of the speakers/headphones but not necessarily for the better. If the headphones sound better at lower volume then it could be because the amp is distorting at higher levels and will be more susceptible to this when switched to a higher output impedance. It could also be because the headphones distort, or even your ears!

As far as the length of the cable goes, if the wire is of sufficient cross section that it does not add to the output impedance in a significant way, you can use as long as you want. Most headphone extension cables are made with very low cross section wire and could add some reactance to the system.

You cannot break headphones by running them from any particular impedance output. You can break them by over-powering them or feeding them DC.

[axero]

No, no, no, you are reading my post wrong.

Let us call the pre-amp modes of the amplifier <64 Ohm = Normal Gain (NG), 64-300 Ohm = High Gain (HG) and 300-600 Ohm Extra High Gain (EGH).

When I listen to the headphones at low volume at HG (i.e. the mode that matches the impedance of the head phones), the headphones tend to sound like the speaker of a telephone. When I listen to the headphones still at a low volume but at EHG the headphones now sound ok.

This is what I find strange. It makes me draw the conclusion that there is more to these different gain modes than just level of amplification.



The headphones still sound a bit telephone-ish at higher volume levels at HG whereas they don't at EHG. I would have to crank up the volume to levels that would make my ears bleed before I'd experience any notable distortion. At high volume (at HG) the treble is so pungent that it feels like it could cut my eardrums while it still has a problem driving the bass, although this problem is slightly less significant than at low volumes.

And my worries is that this EHG mode may damage the headphones due to what you call "overpowering" if I accidentally crank upthe volume too high. I don't think this worry is unwarranted. Otherwise it wouldn't make any sense to operate the amplifier in different modes now would it?


I don't think the impedance of the cable is negligible. Otherwise AWG30 wire would more than suffice to drive say 10 50W halogen lights at 12V. When it comes to ordinary speakers then yes, the thickness of speaker cable is indeed relevant, especially when the cables are longer. But of course it gets less relevant the higher the voltage that is being fed through the cable, and sure, the higher the impedance of the speakers, the higher voltage is needed to operate at the same power/volume level.

That's why I was asking about the impedance of the extender cable, which likely is varying with frequency. But I don't know whether the impedance is purely inductive, purely capacitive or purely resistive. We're talking 3.5mm stereo cable here, not the sturdier 6.5mm type...

[linux-works]

reality: output Z on any GOOD amp is very low.  this negates the cable effect.  and the cable effect is negligable at audio freq's.

I was recently at a 'head-fi' demo or meet (last weekend, in fact) and there was a company selling phones amps that had switchable output Z.  it was funny.  not switchable gains (which does make sense) but output z.  when pressed as to why, he (the sales guy) could not answer it well.  said something about the 10:1 ratio.  I replied that as long as your source Z is low enough, the 10:1 is not a LITERAL figure!  having too low a source is never a bad thing.  think in terms of extremes: picture an amp and preamp connected by an interconnect.  make the source very low Z and see how much noise/hum you pick up at the other end of the cable (the destination side).  now make the source Z much higher and retest.  you'll get noise in the 2nd case.

more extreme example: short the inputs to your amp and vary the volume.  now leave them wide open and vary the volume.

in every single case, lower source Z wins.

other than 'tunings' (sigh), there is no reason to have higher Z on the output of an amp.  some phones may sound different because they expect some series R in the path, but those are broken by design...

[axero]

In my case the headphone amplifier is built into the sound card of a computer so I don't have the ability to short the inputs of the amplifier. The sound card is an ASUS Xonar Essence STX and the three different gain levels can be switched between through the ASUS Audio center software. When switching between these modes I get pretty loud warning messages about damaging my ears and equipment so I just want to make sure I don't do anything wrong here...

When I come to think about it, the impedance/resistance of the extension cable should be negligible since the headphones draw such a small amount of power.

But I can testify that these 250 Ohms headphones (BeyerDynamic DT-990) don't operate properly on this sound card until I set the amplifier into EHG mode which makes me wonder whether there might be something wrong with the card.


Oh, and when talking about headphone vs speaker comparison, I was actually taught that it's better to connect speakers that are stronger than the amplifier than the other way around. That way the sound will clip in the amplifier and not in the speakers risking damaging them. Also I believe it is good practice to not have the volume knobs all the way to max of the final stage amplifier of the speakers. I keep them at 2/3 - 3/4 of max which suits me fine. So maybe switching between these different gain levels is tantamount to switching between three different amplifiers where the EHG amplifier can cause clipping in the headphones and damage them whereas the HG and NG amplifiers cannot.

[linux-works]

your comment about vol control is VERY VERY dependant on how the VC works, amongst other things.

I build (and have designed) attenuators and the relay based attens that I made work perfect all the way up to 0db (ie, max turn to lock, on the pot).  at 0db atten, you are a pass-thru and if the amp can't handle a full input swing (or drive to the gain due to full input swing) its not the volume control's fault

re: asus; its not clear that they are really changing gain.  they might be, but I kind of doubt it.  what they are likely doing is changing 'gain factors' in software either via digital scaling or analog scaling.  its not very likely they are changing the actual gain of the amp (its harder than just changing a feedback resistor, in most cases).

re: spkrs and amps being 'stronger' than the other; that's hogwash.  all that matters is that you don't overdrive things.  don't overdrive the spkrs and don't overdrive the amp.  its just that simple.

[axero]

Well, it is obvious that the sound card has a set of relays and switches in it. For example when switching between speakers and headphones, I can hear a mechanical clicking sound at the back of the computer. Perhaps the switching between the gain modes are operated by semi-conductor based relays because I cannot hear this clicking sound when switching between these gain modes. What switches that really are in place remains to find out.


As for speakers' strength vs the strength of the amplifier. I was shopping a few years ago for a set of studio monitors (that have a max 120W average / 200W peak rating). Then I had to decide upon what amplifier to choose with them since they were passive. I could either go with a 100W (per channel) amplifier or a 150W amplifier. There simply was no 120W amplifier. So I went with the 100W model since it was generally recommended to use this amplifier with these monitors.

That was an interesting take on the attenuators. Then it means that a 100 W amplifier shouldn't be able to overdrive the speakers even if the knobs are at max. But I guess it doesn't hurt to have some margin. Hopefully the sources that are connected to the amplifier are designed in such a way that they won't damage it. I think the amplifier can handle a little bit of overdrive though, it should have some kind of protection. There are clip lights on it so I assume that there is some kind of circuitry to handle it.

I hope you are aware that Atten is a manufacturer of electronics equipment...

[Lightages]

I just had a look at the manual and IT IS just BS about optimizing for headphone impedance, As has been said this is just marketing crap. The output impedance needs to be AT LEAST AT MOST 1/10 of the load and the lower the better. Gain is not expressed in ohms, it is a ratio or number, ie: 10:1 or just 10. Higher output impedance on an amp will only degrade the capability of the amp to produce proper output. Asus is pulling audiophool BS here.

Edited by backwards wording, sorry.

[true]

On at least Xonar DG, it uses a pretty crap TI line driver and the output impedance is fixed at ~10 ohms. Essence STX may be similar.

I would recommend instead that you mod the non-headphone-amp portion that uses a regular op-amp with a decent op-amp that can drive your headphones or use an external amplifier. I did that with a DG and it is far superior to the TI line driver in "headphone" mode.

[axero]

The only difference I can see between these modes is that in NG mode, the output gain is set to 0dB, in HG mode the gain is set to +12 dB and in EHG mode it is set to +18 dB.

I cannot see in the documentation (which I admittedly didn't read anyway) that the sound card is supposedly adjusting its output impedance or that gain is measured in Ohms. To me it looks like it is providing the ability of driving heavier headphones while not killing lighter headphones by separating these into three different modes. Otherwise these different gains should have been incorporated seamlessly into the volume control.

BS or not, at least it sounds nice in EHG mode and the card seems to be recognized in communities as a good sound card, at least for listening to music or watching movies.

It's just strange that the lower gain modes cannot drive the headphones properly while the highest can. I'd expect the lower gain modes should merely result in a lower volume but there is more to it, and that is kind of a mystery to me.

Well, the Essence STX does have a headphone amplifier that appears to be recognized to have a quality that is said to be comparable to DAC/PreAMP combos that cost more than twice as much. I've seen pictures of such mods in different forums, don't know how good they are though. But I'd rather build myself an O2 or something than make such mods voiding warranty.

[linux-works]

also, volume control position is meaningless.  it can be designed so that full lock on clockwise is 0db (no atten) or it could be that 0db is somewhere less of that and full-lock is in the gain (not atten) region.  even still, a full lock setting of gain may not be exhibiting any distortion, load or no load.  it really depends.  'pot position' really is meaningless, in itself.

[commongrounder]

I just want to make a quick comment on the relationship between amplifier output power and speaker power handling, in case someone finds this thread looking for advice on the subject.  By the way, I service audio equipment for a living.
It is always better to have an amplifier that has more power output than the maximum handling of the speaker it is driving.  It may not sound intuitive, but it is true.  If you drive a speaker with an underpowered amp, and the amplifier clips (distorts), the heavy high frequency harmonics of the clipped waveform will pass through the crossover and allow a tremendous amount of energy into the tweeter(s)/high frequency drivers.  I have seen what happens, which is a thermal failure, i.e. "the magic smoke".  When you use an amp with plenty of power, you will, if you overdrive the speaker, usually hear woofer voice coil bottoming noise first, and can back off the volume.  One can always install an in-line fuse for some protection.  As others have said, ultimately you should trust your ears, if it sounds bad, back it off, or suffer the consequences.

[ciccio]

It is always better to have an amplifier that has more power output than the maximum handling of the speaker it is driving.  It may not sound intuitive, but it is true.  I

I totally agree with the above. I've seen lots of tweeters destroyed by too small amplifiers.
I do not understand well the OP problem with what seem to be different gain settings, but I think that some engineer at ASUS decided that they wanted to keep a constant power level with different headphone's impedance, so they increase gain for higher impedance setting.
This increases the output voltage for the same input voltage.
More voltage on an higher impedance will result in the same power (more V squared / more ohm = constant).
Different headphone efficiency is not taken into account.
This is my humble opinion.
Best regards

[axero]

That was really good knowledge, hopefully falling short of 20W shouldn't be too big of a deal. The comparison to weak amplifier vs speaker explains why high volume at lower gain modes yields stronger treble in the headphones (cans) while still failing to properly drive the bass.

To explain the circuitry of the sound card, the heart of it is a TI PCM1792A DAC chip. The outputs of this chip are amplified by a triad of OP-amps consisting of two NJM2114D (connected as a differential amp) and an LM4562NA. In headphone mode the output signal of these amplifiers is fed into a Texas Instruments TPA6120A2 circuit which is the headphone amplifier. When looking at the datasheet of this amplifier:

http://www.ti.com/lit/ds/symlink/tpa6120a2.pdf

I cannot see that there are separate gain modes that can be switched between so it makes me suspect that this is purely a software thing, unless there is a switch somewhere else that yields different V_cc voltages of this amplifier circuit.

But with the speaker/amplifier reasoning above I conclude that there are no dangers whatsoever with driving a pair of 250 Ohm cans with a signal designed for 600 Ohms cans, especially not with my setup. Heck, why bother with a headphone amp when I could instead try and get a small ordinary speaker amplifier of good quality rated for 10W or more. But then otoh if I want to get say 4 W out of a pair of 250 Ohms cans, a rather large amplifier that delivers 250W@4 Ohms would be required in order to compensate for the higher impedance.

Edit: I suppose I could supply the PCB of it...

http://ko.com.ua/files/u4/Xonar_STX_big.jpg
http://www.ixbt.com/multimedia/asus/stx/oub.jpg
http://www.xbitlabs.com/images/multimedia/asus-xonar-essence-stx/p2.jpg

[Lightages]

Amplifiers are rated at some level of distortion. An example; 10 watts at 0.1% distortion at 1kHz. This does not mean the amplifier cannot put out 100 watts at 100% distortion which is mainly made up of a square wave. The tweeters can't take the high level of power at high frequency and burn out.

So the lesson is to never start clipping the output regardless of whatever power the amplifier is rated for. There are various ways of testing for distorted output with using special instruments but I don't have time at the moment to look up the references and pass them on. I will do so later.

[ConKbot]

I just had a look at the manual and IT IS just BS about optimizing for headphone impedance, As has been said this is just marketing crap. The output impedance needs to be AT LEAST 1/10 of the load and the lower the better. Gain is not expressed in ohms, it is a ratio or number, ie: 10:1 or just 10. Higher output impedance on an amp will only degrade the capability of the amp to produce proper output. Asus is pulling audiophool BS here.

Not exactly right Any speaker is going to have a non-flat frequency vs impedance plot, with dips and peaks, and what happens at those dips and peaks changes how the speaker or headphones sound.   If the impedance of the driver dips at 1khz and you have a perfect voltage source (0 ohm output impedance) then as the impedance dips, power goes up in that frequency band.  The other extreme is infinite output impedance(current mode output), in which the power goes down as the impedance dips.  Not too hard to imagine that those would sound different as your power vs frequency distribution changes

Not only that, but headphones are sometimes meant to operate with a series impedance, hence the ability to change the output impedance. 
See http://sound.westhost.com/project56.htm for more info on it. 

Also, to the TS, no if you can keep the headphones on your head, and listen comfortably, youre probably not going to break your headphones. 5M of cord would only add a few ohms, which as others said should be negligible (unless you have a low impedance set of phones, 8-16 ohms, and depending on how thick the wires are, maybe 32 ohm cans) 

[olsenn]

Driving headphones is no different than driving speakers. The output impedance of an amplifier and the wire to the speakers should be less than 1/10 of the lowest impedance of the speaker drivers, as a rule of thumb. The lower the impedance the closer the speakers/headphones will perform to spec. and the lower the better. 1/100 and lower is more typical.

Headphones use much less power than loadspeakers. Delivering that much power into a 250-ohm load requires a decent amount of voltage, but not a lot of current... you don't need to worry about the resistance of the cable. That being said, make sure you get a decent amplifier (look at the O2 by NwAvGuy) that has low THD and a high voltage gain. Refer to the sensitivity specification of your headphones if you need to know how much power is needed to produce a desirable volume.

[Lightages]

The output impedance needs to be AT LEAST 1/10 of the load and the lower the better.

Sorry, I was trying to say "at most" but I was think of the size of the denominator in the fraction as being "at least". SO the impedance needs to be 1/10 or much less. The lower the output impedance of an amplifier the better up to the point of diminishing return.