Head-Fi.org › Forums › Equipment Forums › Sound Science › Some questions about driving requirements and amplifiers
New Posts  All Forums:Forum Nav:

Some questions about driving requirements and amplifiers

post #1 of 8
Thread Starter 
I'm looking to buy new some headphones soon and trying to figure out which would be best for my setup has left me confused and curious. Here's my considerations and their calculated requirements to deliver a 1kHz 0dBFS sine wave at my maximum listening volume:


White = nominal impedance (@1kHz), grey = maximum point on impedance curve

I couldn't find an impedance graph for the DT880/600Ω, so I just assumed the worst and used the graph from the T1. I also don't know if the 32/600Ω versions of the DT880 have the same sensitivity rating as the 250Ω version, so if anyone knows the exact answer to these it would be great to know.

Right now my amp is the headphone out on my sound card, a TPA6120A2 running on a ±12V supply. The measurements show low distortion and noise at low impedances (32-64Ω) and absolutely minimal for everything above that (600-10kΩ). I doubt the effects would be audible with a 32Ω load, but I think I can safely rule out the DT880/32Ω based on the amp's better performance with higher impedances and the lower price of the DT880/250Ω.

Here are my questions:

Would this make the DT880/600Ω the easiest load for the amp to handle?

Would higher impedance ever be anything but beneficial as long as clipping doesn't occur?

Given the extremely low power requirements in all cases, along with the idea that these are all "hard to drive" headphones, why would an amp be needed in the first place?

Tube amps excluded, what is there to gain by having a full-sized desktop amp over this setup other than more unused power and better performance with low impedance loads? What other factors come into play other than THD/IMD, SNR, crosstalk, and frequency response linearity as long as adequate volume can be attained?
post #2 of 8
the calculation I look at is the the peak I,V required for 120 dB SPL output - read up on Dynamic music peaks before the usual inane "too loud" comments please

a +/-12 V powered TPA6120 is substantially more amp than nearly any portable and lots of sub US$3-400 desktop amps - in both current and V output

I would go for 250 Ohm cans, requiring 11 Vpk, as using more of the Vswing if portable operation isn't a goal - 32 Ohms would be better with lower V battery powered portable amps and the TPA6120 in the soundcard should be OK with it but the noise floor would be higher than it would be with the better matched to the soundcard gain 250 Ohm version

with 600 Ohm requiring 17 Vpk, I would want some more supply V for the 600 Ohm which somewhat restricts even the desktop selection for them if you eventually go that way
post #3 of 8
120 dB peaks are so rare that it's not worth considering IMHO. I would wager that all 'phones are very distorted at that level anyway. Even if the sound pressure doubled, then doubled, then doubled, then doubled, from your 85 dB norm it would still only be 97dB.

Personally I would suggest choosing according to sound quality and then sort the amp later. You may not even need an amp, as your sound card may suite your needs well.

Having said that I have experienced significant / spurious low level background 'noise' on soundcard amps, which is one of the things specifications can not reveal.
post #4 of 8
Quote:
Originally Posted by Leny View Post
120 dB peaks are so rare that it's not worth considering IMHO. I would wager that all 'phones are very distorted at that level anyway. Even if the sound pressure doubled, then doubled, then doubled, then doubled, from your 85 dB norm it would still only be 97dB.

Personally I would suggest choosing according to sound quality and then sort the amp later. You may not even need an amp, as your sound card may suite your needs well.

Having said that I have experienced significant / spurious low level background 'noise' on soundcard amps, which is one of the things specifications can not reveal.
Remember that just because the sound's intensity is 16x, the sound wouldn't be perceived as 16x as loud. For a 10db increase the human ear perceives an approximate doubling in volume. So if you have peaks that are 4x as loud you'd need about 20db headroom over what level you're listening at. So if you were listening at 85db, you'd need headroom of 105db or higher.
post #5 of 8
Quote:
Originally Posted by xxbaker View Post
Remember that just because the sound's intensity is 16x, the sound wouldn't be perceived as 16x as loud...
Yup, I know what you mean.

What I was trying to say is that if the sound increased a very significant ammount it would still be below the 120dB capability that jcx prefers. I just don't think that much recorded / encoded music has peaks that big. I might be wrong of course. (Also as jcx DIYs his own amps he can cater for all eventualities, which is a bonus for him).
post #6 of 8
Quote:
Originally Posted by Leny View Post
Yup, I know what you mean.

What I was trying to say is that if the sound increased a very significant ammount it would still be below the 120dB capability that jcx prefers. I just don't think that much recorded / encoded music has peaks that big. I might be wrong of course. (Also as jcx DIYs his own amps he can cater for all eventualities, which is a bonus for him).
It's true that most music for regular listening volumes doesn't go that high. I'm sure we can all agree though that we're not on head-fi for not sweating the details (meaning we are because we do). That extra headroom could be the difference between 99.999% of the music and 99.9995% of the music I'd say that 110 is a more reasonable number that should be enough to weed out all the weak offerings while not being unnecessarily stringent.
post #7 of 8
xxbaker you might well be right. It's an interesting subject to me, so I'm going to do some more research.

OP, hope you find what you are looking for.
post #8 of 8
due to Fletcher-Munson Loundess effect to get "realistic" frequency balance from a recording of a live event you need to play at the live event SPL

HeadWize - Article: Preventing Hearing Damage When Listening With Headphones (A HeadWize Headphone Guide)

read twice: for protection and then for real live music peak SPL values

"Loudness War" compressed material certainly doesn't require much headroom but dynamic "live" recordings start at >12 dB pk to ave and the most extreme (and rare) go to +20-25 dB over ave level, a big band jazz work might need to be played at near 100 dB SPL average level for full impact - not background music to study by
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Some questions about driving requirements and amplifiers