Hi-end headphone amps - alternatives?

I’d like to hear what they did to overcome the distortion problem - it seems to me that the the same basic laws of electricity apply to low cost and high cost amplifiers...

 

This is an interesting thread on balanced headphone amps. I tried to summarize what went on before in my post 112 and wasn't shot down in flames, but it might be worth reading the lot:

Why Balanced Headphone Amps?

 

FYI, in the HPA4 manual, Benchmark states:

"We recommend using the 4-pin jack because the XLR connector provides lower contact resistance than the <1/4"> TRS connector."

... and ...

"The HPA4 included a traditional 1/4" tip-ringsleeve (TRS) headphone jack. This provides a convenient output that is compatible with most headphone cords. Slightly higher performance is available from the XLR4 connector. In our tests we have found that the TRS contacts can occasionally produce more distortion than the entire HPA4 headphone amplifier. We have selected one of the highest quality TRS jacks available, but the performance of the TRS jack is highly dependent upon the quality and condition of the headphone plug. Gold plated plugs are highly recommended. The traditional 1/4" TRS connector uses the sleeve as a common ground for the left and right channels. This connection carries twice the current carried by the other contacts. High contact resistance at the sleeve connection can cause distortion on both channels. With many TRS plugs we have found that the tip connection is somewhat less reliable than the ring connection. This means that there is a higher chance of distortion being added to the left channel. Tip: If the TRS plug is gold plated, clean and in good condition, the TRS jack can match the performance of the XLR4. When conditions are less than ideal, the TRS contact resistance can be the largest source of distortion in the entire HPA4 headphone system."

 

The design of the amp circuit can overcome that distortion problem. That strawman distortion problem is because he's assuming a balanced amp is made by pairing two amps. In a design like that the distortions can double. But some amp circuits are naturally balanced and you can create a balanced amp with just one amp instead of two. As an example, the balanced amps by Schiit are balanced because the amp circuit is naturally balanced, they aren't two amps side by side inside. The PonoPlayer and Ayre Codex are also naturally balanced due to circuit design.

The other issue is economics. Balanced amp designs generally cost more because there are more parts and other issues to deal with. If you're designing an amp to hit a $400 price point it may be better to make the best $400 single ended amp you can rather than trying to make a $400 balanced amp that is two $200 amps. A well designed $400 single ended amp should sound better than two $200 amps. When you get into the higher cost amps those economic compromises aren't compromising the amp design. At $3000 you can have a very good single ended amp or a very good balanced amp. Both with uncompromised design and sound quality.

 

I find the physics on the electronics of such devices far more believable than the non-informative descriptive superlatives such as "sounds amazing" attributed to such devices in this topic. Again I bring up the issue of impedance because I''ve seen what this does to crossover networks that send certain frequencies to one driver over another. Do headphones have caps and coils for separate bass/high drivers? If so, they require the same cap and coil crossover network over the same two wire balanced system as noted in that Benchmark link.

If there is a smidgen of impedance differences happening at the output stage within the headphone drivers, it's going to change the sound no matter what, maybe not good or bad but definitely different. I saw this happen when I mixed a crossover 8ohm capacitor designed for 8ohm tweaters (from Radio Shack) connected to a 4ohm tweater meant for car audio while playing music to test the sound. I connected the cap to the 8ohm then the 4ohm and it changed the crossover frequency of the 4ohm where it played more bass/midrange or it might've been the other way around. It was back in the early '90's so I can't remember which now. But do consider I could hear this in open air in my living room, not the closeness of a headphone environment.

I am not an electronics engineer. I just like configuring my own sound systems on the cheap for my car. So when I relate this "see something, say something" anecdote it is only meant to support my belief of the physics of electronics and that it really does require a person with a degree in electronics engineering to fully explain what these devices are doing to sound.

 

The TRS headphone jack has always been a bottleneck. I wish I could go back in time and get headphones to standardized on a different plug and jack. TRS plugs and jacks introduce additional crosstalk along with sometimes poor contact. The smaller 1/8" mini jacks on portable players are even worse. The TRS jacks and plugs also can cause brief shorts as you plug it in and unplug it. Headphone amps with TRS jacks have to deal with the possibility that the output gets shorted as the headphones are plugged in or unplugged. The 4-pin XLR connector used for balanced headphones fixes all those problems.

Do you hear a difference between the 4-pin headphone jack and the TRS headphone jack on the HPA4 amp? Benchmark never addresses subjective, only objective. And objectively the TRS and 4-pin XLR have the same measurements except for crosstalk. So Benchmark will say they sound the same. But do they? With balanced headphone amps I usually hear better soundstage and imaging using the balanced connector compared to the TRS connector. I can attribute some/most of that to reduced crosstalk for balanced and that the TRS will usually need to go through a summing circuit to become single-ended. But my ears are telling me there is probably more going on.

 

Headphones are like a single driver speaker with no crossover. I think of headphone amps as being more like pre-amps than power amps. But there is some of both going on. Different pre-amps can sound different, even with very similar specs. Expensive pre-amps can (should) sound better than less expensive pre-amps. And you can have different design principles. Like low or zero feedback vs high feedback. I have found that I generally like the sound of good low feedback or zero feedback headphone amp designs compared to high feedback designs. Tubes and hybrids are good too. The actual electrical engineering behind designing an amp is beyond me. So I go by general rules of thumb based on what I've found that I like. If I see an amp has low or zero feedback I'll be more interested in that amp than if I see it has high feedback. And other rules of thumb that may help me narrow choices down and more likely find what I might like.

 

Phhht.

 

I assumed that as well even with my Sony V6's that don't require an amp to sound good but I never assume until I can take it apart and actually see how it's constructed (not going to happen). Did a search on the subject and found not all headphones are built without crossover networks... Need help designing a headphone crossover

I have an antique pair of headphones it appears from around the '60's that have clear plastic enclosures that show off all the internal electronic components of wiring, inductors and transistors. (What's all that for?) Of course it doesn't sound as good as my Sony's listening through my '72 Sansui 2000A amp.

There are passive crossovers I would think would apply to the bass frequencies seeing there isn't much surface material in headphones needed to push that much air to the point it would damage the voice coil and other internals. My Sony's are measured to reproduce bass down to 10Hz which I know from listening can manage around 20-30Hz but not very loud listening through the line out (headphone out) on my Mac Mini. Just a hint of those low frequencies is enough for me.

I don't understand how headphones need amplification unless the headphone plug in on home stereo equipment is already amplified. I wouldn't know nor would I care as long as the headphones played the entire audio spectrum at reasonable volume and that is happening on both my Mac Mini and Sansui 2000A solid state receiver.

 

That's an amazing sound, Doug.