Testing balanced versus unbalanced headphone amplifier drive

[Nathan Raymond]

But, I do not know what a typical headphone's impedance looks like, nor what sort of output impedance most headphones amplifiers exhibit.

It depends. Some headphones have consistent impedance across their frequency response, some have small variations, and some have giant swings up and down. Today it's often the multi-driver IEMs that can have very uneven impedance, but it's also true for dynamic headphones (see my links to the Innerfidelity test results for the Audio Technica ATH-R70x above for instance). Innerfidelity is a good resource for looking at test results from a wide range of headphones and seeing what their actual measured impedance is:

https://www.innerfidelity.com/headphone-measurements

Regarding headphone amps, it depends on the amp in terms of it's output impedance. Some are 5 ohms, 10 ohms, 30 ohms, or 50 ohms. The Objective 2 is one of the lowest ones at 0.54 ohms. Some amps don't readily publish their output impedance and you're left to measure it yourself or try to derive it from other specs.

For those who haven't read it, this is a good resource on the topic of headphone/amp impedance and why it matters:

https://nwavguy.blogspot.com/2011/02/headphone-amp-impedance.html

 

[DonH56]

Thank you!

 

[March Audio]

Just thinking out loud (not thought it through that much) the only issue I can see with a traditional 3 wire headphone connection is that there is a possibility that one channel could modulate the common neg/return wire with respect to the other channel. What real world effect that might have I'm not sure, and if the headphones have two wire connection all the way back to being commoned in the plug, I cant see it as being of an significance.

As has already been said 4 wire connection to the drivers doesnt make the amp inherently balanced or differential.

 

[Deem]

Do they make cheap wiring for it? And at any rate, what do people recommend here?

If you want a cheap balanced cable, check VE clan's website.

https://www.veclan.com/engappliance_sel_one?eng_ApplianceVo.eac_id=16

It gives many options to choose from for most headphones.

Beware of the VE Black Litz cable though, it's known to be microphonic while it looks cool.

Pricing ranges from 20$ to 100$+ depending on the options you choose.

 

[sergeauckland]

Just thinking out loud (not thought it through that much) the only issue I can see with a traditional 3 wire headphone connection is that there is a possibility that one channel could modulate the common neg/return wire with respect to the other channel. What real world effect that might have I'm not sure, and if the headphones have two wire connection all the way back to being commoned in the plug, I cant see it as being of an significance.

As has already been said 4 wire connection to the drivers doesnt make the amp inherently balanced or differential.

I spend an inordinate amount of my time repairing broken headphones at our local radio station. Most headphones only have a three-core cable to the plug, some use screened cable so the common uses the screen, which tends to be of lower resistance than the two signal wires. A very few have four-core cable.

Regardless, Looking at a practical example. if the cable from the headphones to the plug is, say, 2m long, and the common wire has a resistance of, say, 50 mohms, and the headphone driver impedance is 50-75 ohms, any interactions will be over 60dB down. That doesn't take into account the variation in driver impedance, but it's likely to increase at low frequencies as the driver approaches resonance, and rise again at high frequencies due to the inductance of the driver coil, so should still be -60dB or less. As the interaction is crosstalk, is crosstalk of -60dB or less significant?
S.

 

[March Audio]

Is that configuration typical in domestic "hifi" cans? Most I see have seperate cables going to each driver and dont share a common gnd. Maybe thats just me not looking hard enough

 

[sergeauckland]

Is that configuration typical in domestic "hifi" cans? Most I see have seperate cables going to each driver and dont share a common gnd. Maybe thats just me not looking hard enough

Yes, most people use domestic cans, which is why I end up having to repair them so often. The usual problem is those bloody awful 3.5mm moulded plugs that fail. I cut them off and fit decent 6.35mm jacks. Then of course, they complain the jack won't fit their poxy MP3 players.......

S.

 

[mindbomb]

Is that configuration typical in domestic "hifi" cans? Most I see have seperate cables going to each driver and dont share a common gnd. Maybe thats just me not looking hard enough

There are a lot of "high end" headphones that have 3 wire cables. The beyerdynamic dt880/990/770 series. The akg k712. The philips fidelio x2. The beyerdynamics are available in 600 ohm versions, so those should be fine, but the others are low impedance.

 

[amirm]

If you want a cheap balanced cable, check VE clan's website.

https://www.veclan.com/engappliance_sel_one?eng_ApplianceVo.eac_id=16

It gives many options to choose from for most headphones.

Beware of the VE Black Litz cable though, it's known to be microphonic while it looks cool.

Pricing ranges from 20$ to 100$+ depending on the options you choose.

First welcome to the forum! I went to that link but can't figure out how to navigate it.

For a Sennheiser HD-650, what do they make?

 

[SpeedyRodent]

Is that configuration typical in domestic "hifi" cans? Most I see have seperate cables going to each driver and dont share a common gnd. Maybe thats just me not looking hard enough

There are a lot of "high end" headphones that have 3 wire cables. The beyerdynamic dt880/990/770 series. The akg k712. The philips fidelio x2. The beyerdynamics are available in 600 ohm versions, so those should be fine, but the others are low impedance.

I've seen a fair bit of both. It is really up in the air if manufacturers use 3 or 4 wire cables going to the jack. I have even seen some that have 8 wires all the way to the jack. Headphones like the Beyerdynamic dt880/990/770 series combine the return ground in the left cup connected to a single ground to the jack. To run a differential setup, you need to rewire using a 4 wire cable. I have the DT880 myself and have taken it apart in the past. The AKG headphones with the detachable cable ie. k712 uses a 3 pin mini xlr connector. The connector on the cup would need to be replaced. I have not taken one of these apart myself, but I imagine it would involve a fair bit of modifications. It really makes you wonder if running "balanced" is really worth the effort.

 

[Deem]

First welcome to the forum! I went to that link but can't figure out how to navigate it.

For a Sennheiser HD-650, what do they make?

Make sure that you choose Headphone Cables then click the next icon.
It will give you a list of connector types, choose one and click next. And so on...

https://imgur.com/a/sA3Qs

 

[restorer-john]

Some of his does not fit the normal engineering view, something I find a bit surprising from Benchmark.

I agree, they are talking about bridging, not a balanced system.

Residual noise however is an issue- headphones are phenomenally sensitive and a bridged tied load design makes it that much harder to keep hiss down low enough to not be audible.

But, I do not know what a typical headphone's impedance looks like

They are pretty benign (pretty flat)- I've run impedance plots on several pairs of mine and they certainly don't have the characteristics we are used to seeing in speakers with crossovers, ports and multiple drivers.

 

[DonH56]

^^^ Makes sense, especially the comment on noise. I should have thought another second or two about the application.

Thanks - Don

 

[Vincent Kars]

They are pretty benign (pretty flat)

Bit surprised.
If I look at e.g. the HD800 with a impedance ranges between 340 and 650 ohms I won't call this flat.

https://www.innerfidelity.com/images/SennheiserHD800.pdf

 

[restorer-john]

If I look at e.g. the HD800 with a impedance ranges between 340 and 650 ohms I won't call this flat.

Actually, for a dynamic transducer, it is pretty flat and quite benign. No wild peaks or nasty troughs.

Bear in mind you are only talking about a resonant peak of a little over double the rated nominal impedance (300 ohms) and, according to the sweep, it is quite a gradual rise and fall at both the low end and the top end (probably inductive at the HF end). With a 300-650 ohm impedance, those 'phones would be very easy to drive from any decent class AB amplifiers running decent +/-V rails (the majority of integrated amplifiers out there). I certainly wouldn't be messing around with portable devices with limited available voltage swing for those particular cans.

Here are a some impedance sweeps I ran this morning with my trusty old AudioLab hardware:

AKG-702

AKG-601- cursor at 1KHz

AKG-500

Apple standard white NOS ear buds

Here we have some small bookshelf speakers.

My (new to me) Dali Zensor 1

My favourite small bookshelf speaker, the Jamo E-800

Yamaha NS-5X bookshelf speakers (nominal 6ohms)

As you can see, in general, it is speakers that have the wild fluctuations, not headphones. I have plenty of other speakers with wild impedance curves, particularly the larger bass reflex (ported) designs with compliant woofers.

 

[Graph Feppar]

Will the Amir do the test? I was hoping to find some measurements,I am curious about cable borne crosstalk in typical 2 meter long cables.

 

[amirm]

Yes, I probably need to do that at some point. Meanwhile since this thread was created I have tested a number of amplifiers and most of them produce more power in "balanced" output so there is real value in that regard. In some cases this comes with more distortion, other times not.

 

[Graph Feppar]

I would love if you could test crosstalk in 1/4" (6.35mm) and 1/8" (3.5mm) connectors and also 2.5mm and 4.4mm balanced connectors that high end portable players use.

I am planning to buy a player and I dont need the higher power of balanced output,I have read the Benchmark article but I fear crosstalk,I saw only one measurement of 3.5mm unbalanced connector on the internet and it showed strongly rising crosstalk at high frequencies,I believe it was only -60db at 20 KHz.

 

[sergeauckland]

I would love if you could test crosstalk in 1/4" (6.35mm) and 1/8" (3.5mm) connectors and also 2.5mm and 4.4mm balanced connectors that high end portable players use.

I am planning to buy a player and I dont need the higher power of balanced output,I have read the Benchmark article but I fear crosstalk,I saw only one measurement of 3.5mm unbalanced connector on the internet and it showed strongly rising crosstalk at high frequencies,I believe it was only -60db at 20 KHz.

Why 'only' 60dB? 60dB is some 40dB (100x) more than you need for good stereo, so of no concern whatsoever.

S

 

[Kyle / MrHeeHo]

One thing I've wondered for a long time now is if "balanced" the way it's used headphone amps allows for more power out of small form factors. That's the only thing that would make sense to me for its prolific presence in the headphone market other than audiphiles demanding it so much that everyone is caving to the pressure.