Head-Fi.org › Forums › Equipment Forums › Sound Science › Homemade headphone frequency response
New Posts  All Forums:Forum Nav:

Homemade headphone frequency response

post #1 of 9
6/22/11 at 6:54am
Thread Starter 

Hi all,

 

I have just finished building my own homemade super-thin ear hook headphones (I need them to be as thin as possible to fit under my bike helmet and not totally block my ear canals from outside sound). The drivers I used are from China, metal, 23mm diameter, 3.9mm thick, with the magnet side glued to a round 23mm disc of plastic that is connected to an earhook  (Yes, I know there's no sound chamber, but audio-quality is not of high importance in this application). The front of the metal driver is simply covered with a piece of stretched out foam taken from a set of earbuds. The drivers came soldered to a run-of-the-mill 1.3m cable. The headphones actually work very well for their intended use and I do not plan on changing them. 

 

The specs included with the drivers is as follows: 

 

  • Sensitivity: 116 dB/V ±5dB @ 1 kHz
  • Impedance: 32 Ω
  • Frequency Range: 20-20kHz

Frequency Response:

  Frequency Response.jpg

 

Here are my questions (please bear with me because I am a newbie at speaker science):

 

-Based on the specs above, please comment if you can about the sound quality of these drivers. Poor/no high-end? Is the curve too smooth to be real?

 

-Why didn't the manufacturer include a graph that shows response over the entire 20-20kHz Range?

 

-The sensitivity is listed as +/- 5db. Does this mean that the real frequency range should be listed as 200-2.3kHz (~2.3kHz is where the response intersects the lower sensitivity range of 111dB)?

 

-Being that the drivers are simply glued to a plastic disc with no enclosure, can I assume this frequency response of the driver itself would be somewhat similar to the completed headphone as described? If not, why, and do you recommend any labs that can do a test for me (I'm very curious)?

 

I appreciate any information you guys have for me. Thanks. 

Reply
post #2 of 9
6/22/11 at 7:27am

My unscientific guess is that the graph is there to show the drivers presence in the region of the human voice. It's probably made for cellphone use and not High-fi and they didn't measure the rest as it wasn't important to their engineering goals. 

Reply
post #3 of 9
6/22/11 at 7:32am

Now that is quite an impressive roll-off of the high frequencies. However, as JadeEast said, this is clearly not a driver with even the slightest pretension to high fidelity. But hey, they sound good to you, so keep 'em.

Reply
post #4 of 9
6/22/11 at 7:54am

I would ASSUME that these drivers were originally intended to be used as speakers in wireless home telephones - hardly hi-fi at all. The limited frequency response range and poor sample to sample consistency supports that. Id look elsewhere for parts for a project. 
 

Quote:
Originally Posted by Jebbhd View Post

-Based on the specs above, please comment if you can about the sound quality of these drivers. Poor/no high-end? Is the curve too smooth to be real?

 

-Why didn't the manufacturer include a graph that shows response over the entire 20-20kHz Range?

 

-The sensitivity is listed as +/- 5db. Does this mean that the real frequency range should be listed as 200-2.3kHz (~2.3kHz is where the response intersects the lower sensitivity range of 111dB)?

 

-Being that the drivers are simply glued to a plastic disc with no enclosure, can I assume this frequency response of the driver itself would be somewhat similar to the completed headphone as described? If not, why, and do you recommend any labs that can do a test for me (I'm very curious)?

 

I appreciate any information you guys have for me. Thanks. 



In order:

will probably sound like poo. Imagine listening to your buddy's hi-fi through your telephone. 

 

To paraphrase an old professor on the subject of reading specifications:

if they didnt say it, its probably not very good.

Why give a graph 20-20K when its going to be 36db down at 20khz(extending the graph)?

 

Due to the sensitivity spec, you can safely assume that they are inconsistent from sample to sample. Many units will need to be purchased and hand-matched for a stereo pair.

 

You can get an adequate measurement of a DIY headphone with a measurement mic on a DIY setup. As long as you compare it to your reference on the same setup your all good. 


Edited by nikongod - 6/22/11 at 7:55am
Reply
post #5 of 9
6/24/11 at 9:49am

Headphones should roll off higher in the treble, after 10khz. That hump at 1khz could be very noticeable as the ear can discern 1-5 khz very well. At least according to a few speaker-related articles, a mild depression at those frequencies will be regarded as "mellower" and more euphonic.

Reply
post #6 of 9
6/24/11 at 10:03am

Sure, a soft roll off is fine, but we see a 20 dB drop from 1 kHz to 4 kHz here. (!)

 

Also it looks like this driver has a very high resonant frequency, so like others have said I also wouldn't use it for anything but speech.


Edited by xnor - 6/24/11 at 10:05am
Reply
post #7 of 9
6/27/11 at 9:53am
Thread Starter 

Thanks for all the replies! I'm still gonna use them but it's nice to know why they sound the way they do. 

Reply
post #8 of 9
6/28/11 at 6:31am
Quote:
Originally Posted by anetode View Post

Headphones should roll off higher in the treble, after 10khz. That hump at 1khz could be very noticeable as the ear can discern 1-5 khz very well. At least according to a few speaker-related articles, a mild depression at those frequencies will be regarded as "mellower" and more euphonic.


considering it's 12db biggrin.gif ... veeeery noticeable biggrin.gif
Reply
post #9 of 9
7/3/11 at 1:05pm

This is an example of how the full graph should look like from 20Hz to 20kHz

I doubt anyone needs translation, but if you do, just holla

 

86125.jpg

Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Homemade headphone frequency response