[Rossliew]

Thanks for the input, Tom. Sounds like it covers a wide range of headphones.

 

[Armaegis]

   
Thus, I suspect the "balanced amplifier" is mostly a feel-good marketing thing. 
 
Balanced inputs, yes. Balanced from input to output ... meh, not so much.

 
Balanced is one of the big buzzwords, right up there with class-A. Most people don't really understand what either of those means anyways

 
That all said, I've mucked around with enough toys that I seem to prefer balanced/bridged/crosscoupled/whatevertheycallit outputs when given the chance. Maybe it's indeed the lower even order harmonics. Maybe it's the increased headroom. Maybe it's because you're using active devices on each side to sink/source current instead of pushing current into ground. Or maybe my ground is dirty and noisy as all hell and taking it out of the equation is better all around.

 

[tomchr]

As I tried to say earlier, I do see technical advantages of differential/balanced signalling. Most importantly, differential signalling moves the ground loop on the input (yes, there is one) out of the signal path. This means the error voltage developed across that ground loop will be rejected by the CMRR of the differential receiver.
In order for the differential receiver to do its job, it needs a differential signal. Such a signal is provided by a differential driver (aka output).
 
If done properly, it is perfectly possible to have a circuit, say a preamp, that has differential receivers on its inputs and differential drivers on its outputs while having the circuit between the receiver and the driver be single ended. That's what my DIFF PRE 8x2 does. Balanced into a THAT1200 -> relay switching -> volume control -> THAT1646 output driver. The chain between the two THAT ICs is single ended.
Similarly, the differential input of the HP-1 is differential up to the differential/single-ended source selector relay. From there it's all single-ended for reasons I mentioned in my post earlier.
 
So basically: Make the connections between boxes via a differential link. Connections and circuitry within boxes can be differential or single ended, but as I tried to explain earlier I don't see any technical advantage of having the circuitry within the box be differential.
 
Hope this makes sense. It's not a simple GOOD vs BAD here.
 
Tom

 

[Armaegis]

Just tossing thoughts out there for discussion Tom. I was curious how you'd pursue the idea *if* it were something you'd consider doing.
 
Another Canadian company, Bryston, also does the majority of their internal circuitry single ended. Their headphone amp derives the balanced outputs by simply inverting/bridging the outputs.

 

[tomchr]

Most "balanced" design are really two single-ended signal paths run in opposite phase. That's not actually differential or balanced. If the two single amp halves match well, you'll get some cancellation of the even order harmonics. If the two amp halves don't match well (which is the more likely case) you'll degrade performance.
 
If I was to approach a fully differential design, I'd likely have to go all discrete. I'd approach it as a power opamp design. I have yet to beat the performance I can get on silicon with a discrete design, so I'm not going to go that route; in particular as the only reason to do so is to get a marketing bullet on the sell sheet.
 
If I wanted a bling amp, I'd pick some exotic vacuum tube and design a headphone amp around it. That'd give me the marketing bling and actually be fun to do...
 
Tom

 

[Armaegis]

  If I wanted a bling amp, I'd pick some exotic vacuum tube and design a headphone amp around it. That'd give me the marketing bling and actually be fun to do...

 
Dunno if this ticks the box for "exotic"...  http://korgnutube.com/en/  

 
 
datasheets: http://www.nutube.us/

 

[bluemonkeyflyer]

   
Dunno if this ticks the box for "exotic"...  http://korgnutube.com/en/  

 
 
datasheets: http://www.nutube.us/

That's Exotic Bling and looks interesting. I wonder if they perform better than old school valves?

 

[bluemonkeyflyer]

  Most "balanced" design are really two single-ended signal paths run in opposite phase. That's not actually differential or balanced. If the two single amp halves match well, you'll get some cancellation of the even order harmonics. If the two amp halves don't match well (which is the more likely case) you'll degrade performance.
 
If I was to approach a fully differential design, I'd likely have to go all discrete. I'd approach it as a power opamp design. I have yet to beat the performance I can get on silicon with a discrete design, so I'm not going to go that route; in particular as the only reason to do so is to get a marketing bullet on the sell sheet.
 
If I wanted a bling amp, I'd pick some exotic vacuum tube and design a headphone amp around it. That'd give me the marketing bling and actually be fun to do...
 
Tom

 
Tom,
 
Please let me know if you do this.
 
I'm loving Neurochrome HA-1!
 
~ BMF

 

[Armaegis]

  That's Exotic Bling and looks interesting. I wonder if they perform better than old school valves?

 
I think someone brought it up a while ago on some bereft alternative forum, but found that the specs were lacking to make any real informed decision. Still, it'd be kinda neat.
 
 

   
Tom,
 
Please let me know if you do this.
 
I'm loving Neurochrome HA-1!
 
~ BMF

 
 
p.s. the HA-1 is a rather malign offering from Oppo

 

[tomchr]

  That's Exotic Bling and looks interesting. I wonder if they perform better than old school valves?

 
I did come across a data sheet for them at some point. They're surprisingly linear. They also run at incredibly low currents (tens of uA), so one needs high impedances to make them work well. Maybe a CCS load or something... It could be a feature (low current -> low power -> small size) or a bug (low currents -> high impedances -> higher noise).
 

  Please let me know if you do this.
 

 
The best way to get notified is by subscribing to my newsletter (www.neurochrome.com/newsletter). I hope to get up to maybe 2-3 newsletters per year, so you won't drown in spam. Not from me anyway. The main idea is to be able to give those interested a heads up about future products and maybe the opportunity to provide inputs on what future products they'd like to see.
 
Tom

 

[bluemonkeyflyer]

   
I think someone brought it up a while ago on some bereft alternative forum, but found that the specs were lacking to make any real informed decision. Still, it'd be kinda neat.
 
 
 
 
p.s. the HA-1 is a rather malign offering from Oppo

My bad. I blame chronic insomnia.  
 
Correction: I'm loving Neurochrome HP-1...sorry, Tom!

 

[bluemonkeyflyer]

   
I did come across a data sheet for them at some point. They're surprisingly linear. They also run at incredibly low currents (tens of uA), so one needs high impedances to make them work well. Maybe a CCS load or something... It could be a feature (low current -> low power -> small size) or a bug (low currents -> high impedances -> higher noise).
 
 
The best way to get notified is by subscribing to my newsletter (www.neurochrome.com/newsletter). I hope to get up to maybe 2-3 newsletters per year, so you won't drown in spam. Not from me anyway. The main idea is to be able to give those interested a heads up about future products and maybe the opportunity to provide inputs on what future products they'd like to see.
 
Tom

Tom, 
 
Thanks. I'll sign up for your newsletter. I would actually welcome more frequent newsletters or a blog about your designs, general technical information, and "how to's." 

 

[tomchr]

Hoping I'm not overstepping too many boundaries by posting this.
 
As many of you know I'm currently studying psychology and am taking a class in consumer psychology. As part of the coursework we are to conduct a market analysis and design a marketing campaign. For the analysis, we have devised a brief survey. I would greatly appreciate it if you'd take the survey: 

https://ucalgarypsych.co1.qualtrics.com/SE/?SID=SV_4Ysedh1YPDx35eR​

The survey is completely anonymous. It will be a big help to me, both for my psych degree and quite possibly for my business as well.
 
Thanks,
 
Tom

 

[Armaegis]

  Hoping I'm not overstepping too many boundaries by posting this.
 
As many of you know I'm currently studying psychology and am taking a class in consumer psychology. As part of the coursework we are to conduct a market analysis and design a marketing campaign. For the analysis, we have devised a brief survey. I would greatly appreciate it if you'd take the survey: 

https://ucalgarypsych.co1.qualtrics.com/SE/?SID=SV_4Ysedh1YPDx35eR​

The survey is completely anonymous. It will be a big help to me, both for my psych degree and quite possibly for my business as well.
 
Thanks,
 
Tom

 
Done!

 

[bluemonkeyflyer]

   
Done!

Also Done.