Head-Fi.org › Forums › Equipment Forums › Sound Science › My Theory on Balanced vs Single Ended for High/Low Impedance Headphones
New Posts  All Forums:Forum Nav:

My Theory on Balanced vs Single Ended for High/Low Impedance Headphones - Page 2

post #16 of 92

Beyer's comments are suspect - probably marketing not understanding the tech

 

for the same motor structure - same mag field, same voice coil dimensions - you are "free" to design to "any" impedance you want by using the same weight of copper - trading of turns count against wire size - within the limits of practical wire size, some "2nd order" limitations of fill factor/insulation thickness tradeoffs

 

the effect is like having a "ideal transformer" built in - the electro-mechanical driver properties are identical except for the terminal impedance scaling factor - the the electro-acoustic response should be the same as well

 

for Beyer to claim that the same model # with different impedance has different "speed" means they didn't follow this well known motor design principle and deliberately designed in different responses

 

or the marketing guys just couldn't refrain from making something up

post #17 of 92
Quote:
Originally Posted by goodvibes View Post

 The question could asked. Why not use the thinner wire on the low impedance version? Heat from pulling more current?

 

Anyway, thanks for a 'balanced' view on this.bigsmile_face.gif

 

Yep, the high impedance 'phone uses less current so you can use thinner wire.

The lower impedance 'phone pulls more current so thicker wire is required.

Obviously, in a headphone we are talking about very thin wire either way.

But may as well use the thinnest wire you can, to make the lightest total diagphram mass you can.

 

A balanced view.........................LOL!rolleyes.gif

post #18 of 92
Quote:
Originally Posted by jcx View Post

Beyer's comments are suspect - probably marketing not understanding the tech

 

for the same motor structure - same mag field, same voice coil dimensions - you are "free" to design to "any" impedance you want by using the same weight of copper - trading of turns count against wire size - within the limits of practical wire size, some "2nd order" limitations of fill factor/insulation thickness tradeoffs

 

the effect is like having a "ideal transformer" built in - the electro-mechanical driver properties are identical except for the terminal impedance scaling factor - the the electro-acoustic response should be the same as well

 

for Beyer to claim that the same model # with different impedance has different "speed" means they didn't follow this well known motor design principle and deliberately designed in different responses

 

or the marketing guys just couldn't refrain from making something up

 

I'm just going by what Beyer says.

And I am shocked, SHOCKED! that the marketing guys don't understand the Engineering guys! biggrin.gif

 

Anyway, I'm not a motor or transformer designer.

I guess more windings in the high impedance can don't come into play, which is what I had originally suspected.

Beyer claims a high impedance DT770/880/990 is preferable to a low impedance DT770/880/990.

If what you say is true, I wonder why they would even bother marketing various impedances of the same model since a good desktop amp will drive either a low or high impedance can, so may as well just make a low impedance can and be done with it.  

Beyer's claim is that the high impedance cans are designed for the desktop amp. 

 

But I can see your point: if it's an ideal transformer, 1 mW will always lead to the same SPL, which is what Beyer claims.

Some folks argue that Beyer's efficiency claims are not accurate.


Edited by Chris J - 8/4/12 at 4:41pm
post #19 of 92
Quote:
Originally Posted by TwoEars View Post

 

The really interesting piece is this: "The effect of an inductor in a circuit is to oppose changes in current through it by developing a voltage across it proportional to the rate of change of the current."

 

Basically the the faster the rate of change of current in the wire is the more resistance will be seen.

 

Could balanced drive have any advantages here? That they are less sensitive to resistance imposed by induction in the drive coil since they have a knack for delivering twice the voltage of a SE circuit?

 

More accurately:

The faster the rate of change, or the higher the frequency, the higher the reactive impedance of the inductor, i.e. the more impediment you have to current flow.

 

Anyway, I can't see any advantage to a balanced design here. You can make a balanced amp that can output 5 Volts, or you can make an SE amp which can output 5 volts, which is quite a lot for almost any headphone.

 

If the amp can drive the 'phone with enough voltage and the amp can provide enough current, then Balanced and SE don't matter.

BTW, virtually every headphone and speaker amp is a voltage source, i.e. voltage is applied to the headphone or speaker, and the amp provides current based on the impedance of the amp.

The trick here is to NOT run out of voltage or current.   Running out of either is referred to as clipping. 

 

Edit:  read this:  http://bryston.com/PDF/brochures/BHA1_BROCHURE.pdf

 

Maybe I should use the "multi" button..............redface.gif


Edited by Chris J - 8/5/12 at 7:32am
post #20 of 92
Quote:
Originally Posted by Chris J View Post


BTW, virtually every headphone and speaker amp is a voltage source, i.e. voltage is applied to the headphone or speaker, and the amp provides current based on the impedance of the amp.
Not quite every one. I am listening to an exception right now. The initial two models offered by FirstWatt are transconductance, also known as current source, amplifiers. At least one member here has used the F1 to drive headphones. As long as the outputs are connected directly to the transducers, miracles can happen. Linearity and bass extension are improved re: transducer response. Networks can still be used, but they need to be in parallel instead of series.

Another useful property is that the circuit ignores "the series impedance elements in the circuit, including the wire, connectors, the inductance of the voice coil, the resistance of the voice coil versus temperature - all that stuff." Quote fron Nelson Pass in the F2 manual. No need for expensive wire or connectors (if there ever was) and power compression is no longer possible.

A lot to be said for this approach. There is an easy way to try this out using resistors, but a great deal of the power of the amp is lost as heat. Not the same as doing it right. I had an F2 hooked up at CanJam in Chicago but only a handful of attendees bothered to listen.
Edited by Clarkmc2 - 8/5/12 at 5:49pm
post #21 of 92

Keep in mind that the F1's output impedance is only 80 ohms and the F2's just 15. That's because they were only intended to drive 8 ohm loads. So if you're looking to try and achieve the same thing with headphones with their typically much higher impedance, you're going to have to up the ante.

 

The headphone jack on my TEAC A-H500 is driven directly off the speaker outputs through 390 ohm series resistors. Sounds quite nice with my 50 ohm LCD-2's. Though you'd likely need to use a network of some kind to tame the bass response when using regular dynamic headphones.

 

se

post #22 of 92
Thank you for the clarifications, Mr. Eddy. I do have to add that the amps have a lot of power for headphone use. My F2JFET can deliver 2.4 amps. While rated at half the wattage of the F1, it has much higher voltage on the rails and can supply a ton of current into a variety of things, or so I am told.

Not to worry, with only a hundred of each model out there, nearly everyone can still enjoy all the advantages of your superbly made, beautiful cables. smily_headphones1.gif
post #23 of 92
Quote:
Originally Posted by Clarkmc2 View Post

Thank you for the clarifications, Mr. Eddy. I do have to add that the amps have a lot of power for headphone use. My F2JFET can deliver 2.4 amps. While rated at half the wattage of the F1, it has much higher voltage on the rails and can supply a ton of current into a variety of things, or so I am told.
 

 

It may be idling at 2.4 amps, but it's not going to be able to deliver that kind of current except into the lowest impedance loads. Note that the F2 is rated at 10 watts into 4 ohms, but only half that into 8 ohms. Which means it's only able to have about 9 volts across its outputs. If that holds true for higher impedance loads, then it's only able to deliver about 800 mW into a 50 ohm load.

 

Quote:
Not to worry, with only a hundred of each model out there, nearly everyone can still enjoy all the advantages of your superbly made, beautiful cables. smily_headphones1.gif

 

biggrin.gif

 

se

post #24 of 92
Quote:
Originally Posted by Clarkmc2 View Post


Linearity and bass extension are improved re: transducer response.

I'm sorry but this works only with speakers that are designed for such type of amp. With dynamic headphones you usually get a peak in the bass and sometimes a bit of an upper treble boost. Also, distortion increases a lot. According to benchmark a 30 ohm output impedance can raise the THD by 50 (!) dB. All of this is pretty bad.

post #25 of 92
Quote:
Originally Posted by xnor View Post

I'm sorry but this works only with speakers that are designed for such type of amp. With dynamic headphones you usually get a peak in the bass and sometimes a bit of an upper treble boost. Also, distortion increases a lot. According to benchmark a 30 ohm output impedance can raise the THD by 50 (!) dB. All of this is pretty bad.
Sort of. The subset of speakers that would benefit are equiped with powerful magnets, light cones and tight gaps. That is more than Fostex, Lowther and other full range speakers. Most decent JBL woofers, for instance, qualify. I can verify good results with the application. My 16 ohm Beyers seem to like it too.

While not in its wheelhouse at all, the F2 was an impressive performer on Spritzer's transformer box. Very versatile amp. I never write off an application until I have tried it, no matter what it is supposed to do on paper. As for distortion, the JFET version of the F2 has 20dB less than the origional. I don't doubt your numbers, just sayin'. Talk about YMMV!.

I should have just mentioned speakers; I did not want to ruffle any feathers or argue with anyone. I admit I am not a fan of dynamic headphones, but I do use a couple almost every day. All I am saying is it works well on my stuff.
Edited by Clarkmc2 - 8/6/12 at 4:59am
post #26 of 92

if talking about properly designed voltage source amps, the doubled output impedance is going to be pretty much meaningless as the outputz from the SE version and balanced version will be rendered tiny compared to (or at least not a big deal compared to) the cable, connector resistance. what I mean by that is, if the damping factor is good enough doubling it is not of any consequence whatsoever

 

Chris, doubling the voltage doesnt multiply the power x 4 in itself, only if the PSU is capable of it


Edited by qusp - 8/6/12 at 5:58am
post #27 of 92
Quote:
Originally Posted by Clarkmc2 View Post

Thank you for the clarifications, Mr. Eddy. I do have to add that the amps have a lot of power for headphone use. My F2JFET can deliver 2.4 amps. While rated at half the wattage of the F1, it has much higher voltage on the rails and can supply a ton of current into a variety of things, or so I am told.
Not to worry, with only a hundred of each model out there, nearly everyone can still enjoy all the advantages of your superbly made, beautiful cables. smily_headphones1.gif


a couple hundred each? unlikely, maybe the commercial versions but if the build threads are anything to go by there is a hell of a lot more builders than that.

 

toying with the idea of an F6 headamp myself, but I shudder to think of the cost of the iron, at least the cost of the iron that I would accept.

on that note, since you are here Steve, any suggestions for high quality but not millio0n dollar tx for the F6? I dont mind spending say 3-400 for the pair


Edited by qusp - 8/6/12 at 5:48am
post #28 of 92
Quote:
Originally Posted by Clarkmc2 View Post

I never write off an application until I have tried it, no matter what it is supposed to do on paper.

That's not some simulation of what could happen. They did real-world measurements and also listening tests: "At a damping factor of 2, the harmonic distortion produced by the headphones seemed to beat against the inharmonic overtones of the piano giving the impression of a poorly tuned instrument. This effect seemed to diminish when the damping factor was increased [...]".

 

Quote:
I should have just mentioned speakers; I did not want to ruffle any feathers or argue with anyone. I admit I am not a fan of dynamic headphones, but I do use a couple almost every day. All I am saying is it works well on my stuff.

I'm not upset at all and also do not want to argue. I'm just saying that for most dynamic headphones you're probably better off with a low output impedance. If not one can use an EQ which is what I do or if that's not possible simply add resistors to raise the output impedance, assuming the amp has enough gain.


Edited by xnor - 8/6/12 at 6:00am
post #29 of 92
Quote:
Originally Posted by Clarkmc2 View Post


Not quite every one. I am listening to an exception right now. The initial two models offered by FirstWatt are transconductance, also known as current source, amplifiers. At least one member here has used the F1 to drive headphones. As long as the outputs are connected directly to the transducers, miracles can happen. Linearity and bass extension are improved re: transducer response. Networks can still be used, but they need to be in parallel instead of series.
Another useful property is that the circuit ignores "the series impedance elements in the circuit, including the wire, connectors, the inductance of the voice coil, the resistance of the voice coil versus temperature - all that stuff." Quote fron Nelson Pass in the F2 manual. No need for expensive wire or connectors (if there ever was) and power compression is no longer possible.
A lot to be said for this approach. There is an easy way to try this out using resistors, but a great deal of the power of the amp is lost as heat. Not the same as doing it right. I had an F2 hooked up at CanJam in Chicago but only a handful of attendees bothered to listen.

 

My post says "virtually every headphone and speaker amp".

I put that in there for the exceptions to the rule.

 

The definition of a transconductance amp is:

 

Amplifier Gain = Current out/Voltage in

post #30 of 92
Quote:
Originally Posted by Steve Eddy View Post

Keep in mind that the F1's output impedance is only 80 ohms and the F2's just 15. That's because they were only intended to drive 8 ohm loads. So if you're looking to try and achieve the same thing with headphones with their typically much higher impedance, you're going to have to up the ante.

 

The headphone jack on my TEAC A-H500 is driven directly off the speaker outputs through 390 ohm series resistors. Sounds quite nice with my 50 ohm LCD-2's. Though you'd likely need to use a network of some kind to tame the bass response when using regular dynamic headphones.

 

se

 

Is this a current source with a really low output (i.e. bad) impedance or a voltage source with a really high (i.e. bad) output impedance?

An ideal current source will have an infinite output impedance,

An ideal voltage source will have zero output impedance.

Quote:
Originally Posted by qusp View Post

 

 

Chris, doubling the voltage doesnt multiply the power x 4 in itself, only if the PSU is capable of it

 

Yes, that was implied in the statement.

 

One of the advantages of balanced output vs. SE output is that you can get 4 times as much power with the same PSU voltage rails using a balanced implementation vs an SE implementation, assuming the PSU and amp can source/sink the required amount of current.

In addition,  to output 4 times as much power to support a balanced implementation the PSU and amp would need to be able to sink/source twice as much current.


Edited by Chris J - 8/6/12 at 6:35am
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › My Theory on Balanced vs Single Ended for High/Low Impedance Headphones