or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › Power available and distortion
New Posts  All Forums:Forum Nav:

Power available and distortion - Page 2

post #16 of 30
Quote:
Originally Posted by bigshot View Post
 
Quote:
Originally Posted by davidsh View Post
 

Solution: Use a speaker amp with headphones

 

Actually, I do that... and I use the headphone jack on my iMac unamped with no problem.

I use my emotiva with HE-500 and 5LE. Made myself a box attenuating the signal ~12dB. Plenty of power and headroom.

post #17 of 30

Got a question about SNR. Suppose we have these specs: 

  • Signal to Noise Ratio: 
    1 watt: > 97 dB
    Full power: > 121 dB

Assuming the worst, it's a-weighted.

 

Let's say we use a headphone that has a sensitivity of 90 db/mW. That'd be 120 dB/W. So the noise floor (according to specs) would be 120dB - 97 dB = 23 dB, right?

Would one be able to hear that with ambient noise at 30dB or so?

 

Generally, how low should the noise be not to be heard?

post #18 of 30

Up to about 10 dB (A-weighted) noise SPL is normally inaudible. It starts to become noticeable at roughly 20 dB (depending on ambient noise and anything else that could mask it); 30 dB is definitely audible hiss, comparable to cassette tape at moderate (<=90 dB SPL peaks) listening level. But loud and low dynamic range music can still mask it.

 

By the way, the specs you listed above look like those of a speaker amplifier, as it is capable of more than 200 W maximum power. If you do not actually drive headphones directly from the speaker outputs, then the actual numbers will be different. Also, the power values (like 1 W for 97 dB SNR) probably assume typical speaker impedances like 4 or 8 ohms, which should be taken into account when the load is a higher impedance headphone instead.


Edited by stv014 - 5/20/14 at 6:00am
post #19 of 30
Thank you, just the kind of answer I sought. Well, I didn't think about the impedance difference, which surely makes a difference as we are likely talking 97 dB at ~2.82 V yielding the noise floor of -97dB for a 8 ohm speaker. That noise floor will of course be some dBs lower with a 32 ohm headphone.
post #20 of 30

Many older stereo receivers would feed the headphone jack with the speaker outputs.  In such gear a resistor pair would be used to both lower the signal level and buffer the headphone from the direct speaker outputs. 

post #21 of 30
Quote:
Originally Posted by esldude View Post
 

Many older stereo receivers would feed the headphone jack with the speaker outputs.  In such gear a resistor pair would be used to both lower the signal level and buffer the headphone from the direct speaker outputs. 

The problem with this, of course, is that you have a very high output impedance due to the resistor pair.

post #22 of 30
Quote:
Originally Posted by cjl View Post
 

The problem with this, of course, is that you have a very high output impedance due to the resistor pair.


Only somewhat.  As the output was quite large voltage wise to the need your resistors usually stepped down the voltage so the resistor to ground could be small enough it wasn't too much of a problem.  Though I have seen one old design in the early 1980's that just ran the signal through a 470 ohm resistor and straight onto the headphone jack.  In any case I agree it isn't an optimum way to do it, just pointing out that sometimes such was done. 

post #23 of 30
Quote:
Originally Posted by esldude View Post
 


Only somewhat.  As the output was quite large voltage wise to the need your resistors usually stepped down the voltage so the resistor to ground could be small enough it wasn't too much of a problem.  Though I have seen one old design in the early 1980's that just ran the signal through a 470 ohm resistor and straight onto the headphone jack.  In any case I agree it isn't an optimum way to do it, just pointing out that sometimes such was done. 

The resistor to ground isn't the problem - any resistance between either leg of the amp and the headphones causes it to deviate from an ideal voltage source, meaning that the voltage supplied by the amp will depend on the impedance of the headphone. This is a significant problem if you have headphones with a large impedance variation at different frequencies. It also hurts the electrical damping of the headphones, which can lead to muddy, uncontrolled bass.  You really don't want resistors between your amp and your headphones.

post #24 of 30
The output impedance can get very low with the right configuration.
post #25 of 30
IMD is a mathematical result of harmonic distortion acting on two simultaneous tones.
post #26 of 30
Quote:
Originally Posted by higbvuyb View Post

IMD is a mathematical result of harmonic distortion system nonlinearities acting on two (or more) simultaneous tones.

here, i fixed it for you :)

 

 

The wikipedia page on intermodulation distortion has a really good introductory section which explains the how intermodulation distortion results in a system with nonlinearities and multiple frequency signals present, and how the resulting sum-and-difference distortion products can appear at non-integer multiples of the signal frequencies (i.e., not harmonic).

 

Cheers

post #27 of 30
Quote:
Originally Posted by ab initio View Post
...and how the resulting sum-and-difference distortion products can appear at non-integer multiples of the signal frequencies (i.e., not harmonic)

which is why it sounds so awful if it's ever present at a high enough level to be audible.

post #28 of 30
Quote:
Originally Posted by ab initio View Post
 

here, i fixed it for you :)

 

 

The wikipedia page on intermodulation distortion has a really good introductory section which explains the how intermodulation distortion results in a system with nonlinearities and multiple frequency signals present, and how the resulting sum-and-difference distortion products can appear at non-integer multiples of the signal frequencies (i.e., not harmonic).

 

Cheers

Technically more correct but less useful as an explanation. The range of possible nonlinear distortions is huge; for example low-level white noise is a type of nonlinear distortion, but you would hardly consider IMD in that context. 

 

On the other hand, harmonic distortion is the most obvious and important result of nonlinearity found in headphones and amps, and hence is most relevant to IMD.


Edited by higbvuyb - 5/22/14 at 4:35am
post #29 of 30
Thread Starter 

IMD is result of signal mixing.  Like law of cosines, two signals mix and creates a baby, which is the IMD(which wasn't intended).  I can't imagine where that would come from for audio signals.  For radio signals mixing is a must so IMD is there.

post #30 of 30
Quote:
Originally Posted by SilverEars View Post
 

I can't imagine where that would come from for audio signals.

 

It can result from any non-linear distortion. Multiplicative "mixing" (amplitude modulation) of a signal with itself is basically pure second order distortion.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Power available and distortion