Hifiman HE-6 Review (Headphone)

[cany89]

When an amp can not deliver enough current the output voltage will clip.

When there is enough W, how do we know if the current is enough to drive the headphones?

(I was looking at more information about headphone amps and just stumbled upon your comment here. I seriously don't get this stuff, so please don't mind if I'm asking a really stupid question.)

 

[solderdude]

There can only be enough Watt when the current is available.

Watt = voltage x current.

So when the Watts are there (in a certain impedance) this means both the voltage and current are there.

 

[cany89]

@solderdude so what do people on youtube or forums mean by saying you need an amp with current, not just watts!
It really doesn't make sense since watt = current x voltage (and voltage = current x resistance )?

 

[Nango]

Amp with current = fat toroidal inside = weight of this headphone amp 5kg and above.

 

[Eddy]

@solderdude so what do people on youtube or forums mean by saying you need an amp with current, not just watts!
It really doesn't make sense since watt = current x voltage (and voltage = current x resistance )?

That's true but not the complete truth: A amp can deliver at a certain Impedance a certain maximum voltage and a certain maximum current, thus a certain maximum watts = (maximum current) x (maximum voltage). See for example diagramms at page 59 in the manual of the RME ADI-2 DAC FS https://www.rme-audio.de/downloads/adi2dac_e.pdf

 

[cany89]

That's true but not the complete truth: A amp can deliver at a certain Impedance a certain maximum voltage and a certain maximum current, thus a certain maximum watts = (maximum current) x (maximum voltage). See for example diagramms at page 59 in the manual of the RME ADI-2 DAC FS https://www.rme-audio.de/downloads/adi2dac_e.pdf

I have to admit I didn't fully understand the current/voltage vs impedance graph as I don't know what I or U (in the labels) are.
Though I found interesting info that I missed - I read almost anything related to XLR out when using it on my hifi but not much on phones out.

It says as long as the volume bar doesn't pass the red sections on the bottom graph, it's distortion free. I'm not sure if this is true when you turn off Auto Ref Level. In any case, I turned off the hi-power and enabled auto ref level, turned the volume knob way above what I listen and it stayed in low gain, let alone reaching the distortion level lol. It seems 1W is really something when connected to LCD-24.

Anyway, I'm still not sure if the headphone gets the current that people talks about. But I'm sure there is more than enough headroom...

 

[solderdude]

@solderdude so what do people on youtube or forums mean by saying you need an amp with current, not just watts!
It really doesn't make sense since watt = current x voltage (and voltage = current x resistance )?

For low impedance headphones to reach a certain Wattage you need a higher current and not so much voltage.
For high impedance headphones to reach a certain Wattage you need a higher voltage and not so much current.

Most planars are low impedance and thus need an amp that can provide enough current. That's where the talk comes from.

All amps are voltage and current limited.
When you want an amp that can drive all headphones to a high wattage you need a design that can reach a high output voltage and current.
High voltage x high current = very high power.

That's where all of this info comes from. Another myth is that some headphones need a huge amount of power.
Most don't. They either require a high voltage (and very little current) or a higher current but very little voltage (portable equipment).
The fact that some drivers can handle a lot of power (and not burn out) doesn't mean they require much power under normal listening conditions.

Most people do not listen to anything higher than a few mW. Rarely do people listen to 100mW (0.1W) in peak levels.
100mW in 16ohm = 1.3V @ 80mA.
100mW in 300ohm = 5.5V @ 18mA
Those are all small voltages and currents.

The tested HE-6 (a notoriously inefficient headphone) puts out 92dB SPL at 1V (20mA) = 20mW which would be a pleasantly loud listening level.
To reach 100dB SPL peak you need 2.5V @ 50mA = 0.125W (which is louder than what most people listen to)
To reach 110dB SPL peak and you need 8V @ 150mA = 1.2W (This is impressively loud and you probably can just listen to 1 song at this level)
To reach 120dB SPL peak (= only endurable for the duration of one song or less) you need 25V @ 0.5A = 12.5W
So for this specific headphone, under the above conditions you really need an amplifier that can provide 25V and 500mA.
0.5A is NOT a lot of current. It is for many headphone amps though.

For this reason (you may want to reach 25V) people recommend to use a speaker amp.
A speaker amp that can reach 25V in 4ohm = 150W rated (25V in 8ohm = 80W).
This makes people believe you need an amp that can provide a LOT of current but the reality is that even in the extreme case a 150W/4 ohm amp is used at full throttle the amp only has to provide 0.5A (12.5W) even though the amp can 12x (and probably even more) higher currents.

 

[Shazb0t]

The tested HE-6 (a notoriously inefficient headphone) puts out 92dB SPL at 1V (20mA) = 20mW which would be a pleasantly loud listening level.
To reach 100dB SPL peak you need 2.5V @ 50mA = 0.125W (which is louder than what most people listen to)
To reach 110dB SPL peak and you need 8V @ 150mA = 1.2W (This is impressively loud and you probably can just listen to 1 song at this level)
To reach 120dB SPL peak (= only endurable for the duration of one song or less) you need 25V @ 0.5A = 12.5W
So for this specific headphone, under the above conditions you really need an amplifier that can provide 25V and 500mA.
0.5A is NOT a lot of current. It is for many headphone amps though.

For this reason (you may want to reach 25V) people recommend to use a speaker amp.
A speaker amp that can reach 25V in 4ohm = 150W rated (25V in 8ohm = 80W).

How much do these numbers change when you also factor in PEQ to the Harman curve?

 

[Robbo99999]

How much do these numbers change when you also factor in PEQ to the Harman curve?

Judging by the frequency response graph and boosting the bass by around +5dB then I think the figures would just be shifted by 5dB:

So it would be 95 / 105 / 115dB rather than 100 / 110 / 120dB that solderdude quoted. I think that's right, I'll let solderdude correct me if I'm wrong in that. (And subtract another 2dB from that if your Harman EQ brought down 1kHz to meet the target accurately)

 

[Shazb0t]

Judging by the frequency response graph and boosting the bass by around +5dB then I think the figures would just be shifted by 5dB:
View attachment 144594
So it would be 95 / 105 / 115dB rather than 100 / 110 / 120dB that solderdude quoted. I think that's right, I'll let solderdude correct me if I'm wrong in that. (And subtract another 2dB from that if your Harman EQ brought down 1kHz to meet the target accurately)

And that bumps the power requirements from what @solderdude calculated at the same levels to what?

Power, voltage, current.

 

[solderdude]

The Harman boost is 'desired' around 70-80dB average levels.
It won't be at 110dB SPL and higher because of equal loudness contours.

 

[cany89]

The Harman boost is 'desired' around 70-80dB average levels.
It won't be at 110dB SPL and higher because of equal loudness contours.

But what about the transients? They are 20-30 dB - sometimes even more - above your listening level. In order to get them you need a good headroom. No?

(I do believe in my case, Adi-2 with LCD-2, there shouldn’t be problem.)

 

[solderdude]

Can you show me the existence of transients of 20-30 dB above average level in any normal music recording ?
If that is the the case the average levels of such recording would need to be -30dBFS.
This can be had in recordings with lots of silence in it but this is more of a 'measurement error' than actual DR.

I do believe in my case, Adi-2 with LCD-2, there shouldn’t be problem

LCD2 = 106dB/V. It can handle 15W peak power (music) = 32V = 0.46A = 135dB SPL peak = pain !

ADI 2 DAC = 10V (at 100ohm so about similar in 70ohm) so peaks of 126dB SPL can be reached.

120dB SPL = 5V @ 70mA = 350mW
ADi-2 DAC can supply 460mA.

 

[Jimbob54]

Can you show me the existence of transients of 20-30 dB above average level in any music recording ?
If that is the the case the average levels of such recording would need to be -30dBFS.
This can be had in recordings with lots of silence in it but this is more of a 'measurement error' than actual DR.

I don't think I'd like 30db drum hit above the rest of the music when I'm already listening at 90db or whatever

 

[solderdude]

The mixing/mastering engineer would have compressed/brick-walled it if it happened during the recording sessions.
You can't sell recordings with average levels of -30dBFS.

 

[cany89]

Can you show me the existence of transients of 20-30 dB above average level in any normal music recording ?
If that is the the case the average levels of such recording would need to be -30dBFS.
This can be had in recordings with lots of silence in it but this is more of a 'measurement error' than actual DR.

I read it here: https://www.audeze.com/blogs/technology-and-innovation/sensitivity-impedance-and-amplifier-power

Saying: Here’s why that’s important for reproducing transient sounds: transients can have loudness peaks of 30, 40, or even 50dB greater than the average SPL. Knowing what we do about logarithmic scales, this translates into current requirements which are FAR greater than what’s needed to simply reproduce an average SPL.

I can check or ask someone about it. Thought I’m not sure in which genre / music there is enough dynamic range. Probably on live recorded orchestras.

 

[solderdude]

Well yes... one can find a dynamic range (between softest and loudest signal) like that in music recordings.
That, however, is not the dynamic range between average levels and peak levels.
This would be anywhere between 3 and 8dB for pop recordings and 10-15 dB for well made recordings and maybe 20dB or so for classical recordings.

We are not talking about the same ' dynamic range'.

 

[cany89]

This would be anywhere between 3 and 8dB for pop recordings and 10-15 dB for well made recordings

It’s less than half of what the article says but ok I’ll take it as 10-15 dB than.

 

[solderdude]

This is a 'normal' range for well recorded music.
It means at 'active' listening levels (which are already far higher than 1 hour music enjoyment) you could have peaks of 90 to 95dB SPL.
Add in some bass EQ and you might have occasional peaks of 100dB.

This requires less than 1V (14mA) peaks for your LCD-2 (14mW) and the HE-6 = 2.5V @ 50mA = 125mW peaks.

This is far from needing a power house. The LCD-2 can do this from a phone.
You will need more power when you want to turn things up.
To double the SPL (+10dB) you need 10x the power. This will get you to a loud level that you can listen to for the duration of a song.
When you want to reach impressive levels you need another +10dB so basically 100x the power levels calculated above.

 

[cany89]

This is a 'normal' range for well recorded music.
It means at 'active' listening levels (which are already far higher than 1 hour music enjoyment) you could have peaks of 90 to 95dB SPL.
Add in some bass EQ and you might have occasional peaks of 100dB.

This requires less than 1V (14mA) peaks for your LCD-2 (14mW) and the HE-6 = 2.5V @ 50mA = 125mW peaks.

This is far from needing a power house. The LCD-2 can do this from a phone.
You will need more power when you want to turn things up.
To double the SPL (+10dB) you need 10x the power. This will get you to a loud level that you can listen to for the duration of a song.
When you want to reach impressive levels you need another +10dB so basically 100x the power levels calculated above.

Thanks for the explanation! It really cleared some confusion around amps and power.