Head-Fi.org › Forums › Equipment Forums › Sound Science › I have an engineering challenge- need to prove the benefit of headphone amps/etc.
New Posts  All Forums:Forum Nav:

I have an engineering challenge- need to prove the benefit of headphone amps/etc.

post #1 of 45
Thread Starter 

This challenge may seem simple to some audio/electrical engineers, but I being an 8th grader I am having some issues because its beyond my level of math/science. The deal is, my dad being the science nerd that he is (well, physics major, same difference), says he'll buy me a headphone amplifier if I can somehow prove graphically or mathmatically that is can make a difference.

 

Problem is I'm not very good at math or science, and at the least bit I don't have enough experience with these two fields to be able to effectively present that. So this year I participated in my first ever science fair (little off topic I know), so I'm going to attempt to establish my experiment as scientifically methodical as possible! biggrin.gif Personally, I think he just wants me to do this for the learning aspect of it.

 

How am I going to prove that my headphones will graphically/statistically benefit from an amplifier?

 

I. Headphones: Shure 750DJ, the following graphs come from HeadRoom.

 

The isolation graph seemed redundant so I left it out.

 

 

II. Amplifier: FiiO E7 (actually, it is a DAC/Amp combo)

Specs below are from http://content.miccastore.com/fiio-e7

  • Output Power: 150mW (16Ω); 16mW (300Ω)
  • Headphone Impedance Range: 16Ω ~ 300Ω
  • Signal to Noise Ratio (A Weighted): ≥95dB Line In; ≥100dB USB
  • Distortion (10mW): <0.009% Line In; <0.008% USB
  • Frequency Range: 10Hz ~ 1000KHzPower Supply: Internal 1050mAH Rechargeable Li-Ion Battery
  • TI PCM2706 USB receiver, 48kHz 16-bit support for Windows XP/2000/Vista/7 and Mac OS
  • Wolfson WM8740 DAC, ADI AD8692 OP Amp and TI TPA6130A Power AMP

 

Unfortunately I'm not sure where to continue with this and I'm stuck. What can I use/measure to prove a graphical increase in the performance of my headphones??? Another problem is my headphones don't benefit hugely from an amplifier, and the E7 amplifier is a portable amplifier so it doesn't have a lot of power in comparison to a desktop one like the E9. So here I am. For anyone out there who knows what I'm talking about, wow, I could really use your help right about now. If you need more information or if you could tell me what kind of information I need, I will try as hard as I can to find it.

 

Edit: Oh no, it seems the headroom graphs aren't showing up... ugh... well if you can spare the time heres a link to them http://www.headphone.com/headphones/shure-srh750dj.php


Edited by iceshark - 6/1/11 at 5:58pm
post #2 of 45
The Shure 750dj is a pair of headphones with high sensitivity and a flat impedance vs frequency curve, it shouldn't need amplification. Sorry to disappoint.
Now it it were a pair of DT880, it might be different.
post #3 of 45

For the giant changes claimed with "hard to amplify" headphones, there should be some measurable change apparent with "easy to amplify" headphones with measuring equipment with many times the resolution of the human ear.

 

Besides, people regularly claim advantages from amplifying higher-end Grados, which share the impedence and aren't too far off the sensitivity of the Shures.

post #4 of 45
Thread Starter 

Exactly, it seems difficult for me to prove it would make much of a difference. I heard it makes a minor difference, but the real reason I want an amp is so I can use it for possibly a future headphone, but I've decided its better to get the amp first before a headphone that has to be driven by one.

 

Could I still try and measure it somehow? He has a Denon AM/FM Stereo receiver downstairs, and I "felt" as though there was a difference, but I was listening to it with a Pioneer CD rack connected up to it and the volume on it is super loud (he uses it to drive speakers), so I can only keep the volume >3 and when it gets to a comfortable listening level the music only goes into one earcup...


Edited by iceshark - 6/2/11 at 7:04am
post #5 of 45

A headphone amp is just to give you more Mw into a certain ohm load and that will give you more volume, insensitive and high ohm will benifit most from an amp, some headphones may also benefit from the extra current or low output impedance of certain amps, high output impedance of certian amps and devices can mess up the frequence responce and headphones with large 40mm + drivers can benefit from extra current.

post #6 of 45

Well for there to be a difference you need some to compare it against. What is currently making the digital to analogue conversion and what is amplifying the headphones?

*edit-read the mention of the Denon but what is normal for your use?


Edited by JadeEast - 6/2/11 at 7:25am
post #7 of 45

Subscribed.


Edited by nikongod - 6/3/11 at 6:41am
post #8 of 45
Thread Starter 
Quote:
Originally Posted by JadeEast View Post

Well for there to be a difference you need some to compare it against. What is currently making the digital to analogue conversion and what is amplifying the headphones?

*edit-read the mention of the Denon but what is normal for your use?


Oh, I should have mentioned that how stupid of me.

 

Generally, I use my headphones with an iPod Nano 5th Generation 8GB, without any external amplifier. Also, I use my computer, which I think is a Realtek HD Audio sound card. I should probably find specs on both of those, if possible I will return with more info.

 

post #9 of 45

Here's a general approach:

 

Find the datasheet on whatever op-amp you are currently using (i.e. the one in the iPod)

 

Determine (for your headphones) how well that op-amp performs, using the datasheet. This would involve calculating the typical voltage swing and current requirements for listening at normal levels (via the sensitivity and impedance curves), and using that to determine the operating point of the op-amp. Then the datasheet should have the info you need to get a general idea of its performance at that point (i.e., THD).

 

Do the same thing for the op-amp operating into the headphone amplifier. This will be a much higher impedance load, and it should be able to produce better numbers at this point. Add on the distortion of the amplifier itself, operating into the headphones. This will give you the distortion of the amp + headphone setup.

 

Hope that the performance of the headphone + amp is better than the headphone by itself.

 

--> Alternative: You could also make an argument that you like to listen REALLY loud. The amplifier would be able to drive higher SPL than the op-amp in the iPod. This might be easier than doing all that work.

post #10 of 45
Quote:
Originally Posted by Iniamyen View Post

--> Alternative: You could also make an argument that you like to listen REALLY loud. The amplifier would be able to drive higher SPL than the op-amp in the iPod. This might be easier than doing all that work.


Instead I'd argue that your iPod or whatever is so limited in volume that you cannot rock out. ;)

 

post #11 of 45
Thread Starter 

Well actually, one reason I want the amp is because these headphones don't sound very detailed at a lower volume (50mm drivers, maybe?), so I wanted to test if it would make a difference. I actually prefer listening low, but I still want my music to sound clear and these headphones don't shine when its under about 50% to me.

post #12 of 45

You can try the analog vs digital attenuation angle. Typical DACs, such as the one in the iPod and many sound cards, are limited to 16 bits of dynamic range. This is the standard for CD quality. The more you turn down the volume digitally, the less of those 16 bits are left to represent all the sonic details in your music. If I am not mistaken, iPods and sound cards can only do volume control digitally. But when you add an amp you get analog volume control, so you can set the music to 100% bit perfect digital volume. There is also the matter of signal to noise ratio. Background noise from a sound card is at a constant level. When digital volume is at maximum, the noise becomes a much smaller percentage of the total audio output.

 

There are of course good reasons for using digital attenuation. And it might be a different story with 24-bit audio. But we don't have to get into that ;)  and the iPod can't do 24-bit anyway. And let's just say channel imbalance is not an issue now, for the sake of argument (and a free amp!).

 


Edited by Yoga Flame - 6/3/11 at 12:10am
post #13 of 45


The problem really is, how do you correlate changes in the graphic or statistic data to audible improvement?  To do this with graphic data, you have to convince your "judge" of the effect of changing performance parameters without knowing if he has a concept of what parameters do to sound.   Statistic data is quite different.  Go for user reviews of the amp, perhaps posted here, or often certain sellers include user reviews.  Some even post a user score (zero to 10) as a quantification of a subjective evaluation.  There's the data you want to use to prove your point.  How many users report an improvement in sound by using the amp vs how many don't report an improvement?  That's your statical proof.  You can ignore measurements or specifications, they won't help you.

 

Coupling specs to audible performance is not  easy to do.  For example, it's been proven using double-blind testing that harmonic distortion is far less audible than we think.  Distortion below .5% seem inaudible in double-blind tests, yet people claim to be able to hear the difference between a device with .001% and .0001% distortion.   Once other influencing factors (like awareness of which device is being listened to, how it looks and feels,etc.) are removed, many "audible" differences are removed also. 

 

That's why I think the statistical proof you want is in customer satisfaction data.  You can't invalidate a person's opinion, but you could invalidate (or at least make very difficult to prove) what test data means to audible performance.

Originally Posted by iceshark

 View PostThe deal is, my dad being the science nerd that he is (well, physics major, same difference), says he'll buy me a headphone amplifier

if I can somehow prove graphically or mathmatically that is can make a difference.

 

Problem is I'm not very good at math or science, and at the least bit I don't have enough experience with these two fields to be able to effectively present that. So this year I participated in my first ever science fair (little off topic I know), so I'm going to attempt to establish my experiment as scientifically methodical as possible! biggrin.gif Personally, I think he just wants me to do this for the learning aspect of it.

 

How am I going to prove that my headphones will graphically/statistically benefit from an amplifier?

 

post #14 of 45
Thread Starter 
Quote:
Originally Posted by jaddie View Post


The problem really is, how do you correlate changes in the graphic or statistic data to audible improvement?  To do this with graphic data, you have to convince your "judge" of the effect of changing performance parameters without knowing if he has a concept of what parameters do to sound.   Statistic data is quite different.  Go for user reviews of the amp, perhaps posted here, or often certain sellers include user reviews.  Some even post a user score (zero to 10) as a quantification of a subjective evaluation.  There's the data you want to use to prove your point.  How many users report an improvement in sound by using the amp vs how many don't report an improvement?  That's your statical proof.  You can ignore measurements or specifications, they won't help you.

 

Coupling specs to audible performance is not  easy to do.  For example, it's been proven using double-blind testing that harmonic distortion is far less audible than we think.  Distortion below .5% seem inaudible in double-blind tests, yet people claim to be able to hear the difference between a device with .001% and .0001% distortion.   Once other influencing factors (like awareness of which device is being listened to, how it looks and feels,etc.) are removed, many "audible" differences are removed also. 

 

That's why I think the statistical proof you want is in customer satisfaction data.  You can't invalidate a person's opinion, but you could invalidate (or at least make very difficult to prove) what test data means to audible performance.


Ah, why thankyou! That was very informative! I will definitely take advantage of that.

 

post #15 of 45

Well; I have been selling tech ideas to customers and management for a long time. The bottom line is know what you are selling and what your audience is looking for. I don't think selling opinion will work. Because the opinion came from a selected base (head-fiers). In order to make it valid the OP will need opinions from a wider cross section of the population. I am not sure you'll get the same opinion from the rest of the world as most people don't even know what a headphone amp is.

The OP should know what his father wants and I think is data, data and more data as most geeks want.

The usual proposal/selling starts with features and benefit analysis and concludes with return on investment analysis. Will $500 for 1% improvement sway his opinion? What is his threshold? What is his hot button.

Maybe a deal making approach works better. Will a payback, additional chores make a difference?

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › I have an engineering challenge- need to prove the benefit of headphone amps/etc.