Balanced Tube Headphone Amplifier?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
A couple caveats:

  • I haven't designed anything balanced before
  • I don't even know if I really believe in all the claimed benefits wrt headphone amps

That said, someone asked me to help them build something and I love to learn by challenging myself. After some thinking I came up with this:



This is going to be used with low Z headphones (32-50 ohms). It is surprisingly difficult to find output transformers with the right ratios for headphones and small tubes that will allow truly balanced output. I ended up landing on the Transcendar 'Trio' replacement. It is a push pull OPT with a 9k primary and secondary taps of 16, 8, and 4 ohms. I think I could get balanced operation by treating the 4 ohm tap as a center tap. Correct, in theory?

I was also exploring the idea of using a similar OPT that includes ultralinear taps. It seems that using the ultralinear taps as a2a might allow the use of higher Z headphones without a sky high load reflected to the tubes. My thinking is that if my turns ratio is about 24:1 on the normal taps, the ultralinear taps at 43% of the turns would be about 10:1 ratio. Is that understanding of UL windings correct?

Now with a normal push pull and a single ended output, the reflected load would be 1/4 a2a impedance (4,500 ohms in this case if 32 ohm load). Because I'm grounding the center tap of the secondary, wouldn't the reflected load actually be 1/2 a2a (9,000 ohms with a 32 ohm load)? This would be Class A operation.

Hypothetical load line with 9k load:



I haven't added any kind of volume control at this point so any comments on that, in addition to corrections of my assumptions/schematic, would be a huge help. This is something I'm trying to design to be as simple as possible, hence the single tube stage and LM317 bias (about 36mA as shown for -6V on the grid). And hey, no caps!

I'm guesstimating the power out at around 100mW with a 2V input. If someone wants to ruin their ears even more, an extra stage or preamp with gain would do it nice and quick.
 
Like that it's for class (A)B.Here the two tubes work together all the time.Gives the result of one in classA doubled.You will get more or less 1W out.Input is (Vgg top) 12volts.
And don't forget with the Raa going up the lowest frequenties goes up too.
Mona
 

Attachments

  • 99.jpg
    99.jpg
    56.8 KB · Views: 531
Like that it's for class (A)B.Here the two tubes work together all the time.Gives the result of one in classA doubled.You will get more or less 1W out.Input is (Vgg top) 12volts.
And don't forget with the Raa going up the lowest frequenties goes up too.
Mona


Hi Ketje. Thanks for the feedback! It makes sense that I don't have to use the flipped loadline for this. That will make calculations a lot easier.
 
I'll check it out. Thank you! I have had a hard time finding balanced output schematics for tubes.

Since a speaker's voice coil is typically floating, I'm not sure why the output needs to be balanced. Why would this have any advantage over a simple secondary winding with one end grounded? Since it is only grounded in one place and the impedance is low, I can't see it making any appreciable difference.

But it certainly won't make things worse, so don't let me stop your creative flow. Just wondering what you are trying to accomplish by balancing the output.
 
With regard to headphone amplifier outputs, "balanced" is nothing more than a gimmick. Nearly every "balanced" headphone amplifier's outputs are what the rest of the audio world calls "bridged" and offers no real benefit unless you're stuck with a low voltage power supply and need to get maximum output levels. Otherwise bridging simply adds more noise, distortion and higher output impedance compared to a non-bridged output.

The sole purpose of balanced interfaces is the rejection of common-mode noise that may be picked up on the interconnect cable. This is seldom a problem in the home where short runs of cable are typical. A whole different case in the studio where you have very low level signals running through dozens of feet of cable.

Common-mode noise simply isn't an issue on the headphone side of things so there's absolutely no benefit to going balanced or bridged.

So don't fall for the "balanced" hype.

se
 
@SpreadSpectrum

Someone with balanced headphones is looking for an amplifier and I figured it would be interesting to learn about differential and balanced design. I'm not sure I buy all the claimed benefits, but there's no denying that balanced headphones have become kind of an audiophile *it* thing for can lovers. Besides, it might come in handy down the road if designing a mic pre or piece of studio gear down the road.

Speaking of non-traditional differential applications, I loved your write up on the unity coupled design!

@Steve

I mentioned my own opinion on balanced design to the person that asked me (pretty much the same as yours), but one must do something with one's extra brain power. Might as well be learning about new topologies.
 
Last edited:
Someone with balanced headphones is looking for an amplifier and I figured it would be interesting to learn about differential and balanced design. I'm not sure I buy all the claimed benefits, but there's no denying that balanced headphones have become kind of an audiophile *it* thing for can lovers. Besides, it might come in handy down the road if designing a mic pre or piece of studio gear down the road.

Yes. It has indeed become a thing.

Speaking of non-traditional differential applications, I loved your write up on the unity coupled design!

This is directed at me? Are you talking about the Journal entry on my website about the headphone amp? If so, thank you for the kind words.

se
 
Hi Steve,

The unity coupled comment was directed at SpreadSpectrum, you got caught in the crossfire. I edited the post afterwards so as to be less confusing.

But, I do love the aesthetics of that JFET amp on your website. That's really gorgeous. Reminds me of Universal Audio mixed with Tivoli.
 
@SpreadSpectrum

Someone with balanced headphones is looking for an amplifier and I figured it would be interesting to learn about differential and balanced design. I'm not sure I buy all the claimed benefits, but there's no denying that balanced headphones have become kind of an audiophile *it* thing for can lovers. Besides, it might come in handy down the road if designing a mic pre or piece of studio gear down the road.

Speaking of non-traditional differential applications, I loved your write up on the unity coupled design!
I'm not sure what a balanced headphone is. That's a new thing to me, but I'm not really a headphone guy.

And thanks, I don't know what it is, but I really like that unity-coupled circuit.
 
Hi Steve,

The unity coupled comment was directed at SpreadSpectrum, you got caught in the crossfire. I edited the post afterwards so as to be less confusing.

Ah, gotcha, thanks.

But, I do love the aesthetics of that JFET amp on your website. That's really gorgeous. Reminds me of Universal Audio mixed with Tivoli.

(Insert "Thank you for the kind words" from my previous post here)

:D

se
 
I'm not sure what a balanced headphone is. That's a new thing to me, but I'm not really a headphone guy.

A "balanced" headphone is simply a headphone that has had its cable's TRS plug (which has a common ground for the left and right channels) re-terminated with a four-pin XLR, a dual three-pin XLR, or any of the other dopey connectors the headphone community has adopted for "balanced" headphone amps.

Really this is only for bridged amps since you can't common the inverting outputs of the left and right channels. However with transformer coupled outputs, you can common the bottoms of the output secondaries and drive the headphones from a balanced source using a regular TRS jack.

se
 
Yep, that's one of the really nice things about an output transformer, you have a lot of flexibility on that secondary.

Yeah. Though personally I prefer dispensing with output transformers and just using them on the inputs. But if you're wanting to drive low impedance loads using tubes, it's either a transformer or cap coupling with a bunch of parallel devices.

Thanks for the explanation.

No problem.

se
 
To answer your question, you are correct. This should give equal and opposite voltages wrt ground.

Except headphones don't have center taps.

He would essentially have a floating balanced output by hooking it up as per the schematic.

As a side note, headphone and loudspeaker drivers are differential inputs with very high common-mode input impedances and like transformers will offer good common-mode rejection even when they're fed from wholly unbalanced sources.

se
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.