Expected roll off with 30 ohm load?

2006-05-02 06:01:51

Hi

Im testing some audio equipment for frequency response...and when I load the headphone amp out with a 30 ohm load I get the expected roll off -- BUT how do I know if its TOO much? or not normal?  Whats a typical 30 ohm roll off on headphone outs look like?

?  anyone ?

thanks.

Expected roll off with 30 ohm load?

Reply #1 – 2006-05-05 16:33:15

Rollofff in the bass region due to load impedance might depend on the output coupling capacitor. For example, an output coupling cap of 100µf loaded by a 32ohm headphone would give a -3db point of ~50hz. (F3=1/2*pi*R*C). Thats theoritical or ideal, or practical only for a purely resistive load. Reactive loads(speakers, headphones) might result even higher rolloff. But the question is what is the -3db point of the headphone you are using. If its higher than the rolloff you are getting from the amp, there is not much to gain.

What are you measuring and how ? Does the rest of the response also get "shelved" down i.e. the response in the entire range drops - say -4db - when the amp is under load ? If so then you should understand that output power or driving capability of the amp is also important. In other words the 'output impedance' of the amp. If what you are measuring is a low power device like a soundcard then output impedance is likely quite high (few hundered or even thousand ohms). Load it with a low impedance load like a speaker or headphone and the frequency response/rolloff and power nose dive. distortion sky rockets.

I had tested the exact same thing sometime ago for my soundcard. I will see if I can post some pitcures later tonight.

Expected roll off with 30 ohm load?

Reply #2 – 2006-05-05 17:51:24

With respect to the reactive nature of electromechanical loads such as speakers and headphones: While speakers exhibit substantial reactive load variations with frequency due to strong mechnical resonances, headphones generally exhibit very little of this.

The reason is simply the degree to which speakers must interface with the air and attempt to deliver power into this most inefficient (and reactive) medium. Headphones are "extreme near field" devices which can ignore those issues of power response.

So, a set of 32-ohm headphones are pretty much 32 ohms across the spectrum with very few inductive or capacitive artifacts. Hence, with respect to the amplifiers point of view the expression V(-3dB)=1/2*pi*R*C will be quite accurate.

For 32 ohm 'phones and a typical 100mfd output cap, you get about 50Hz. This is much better than most real-life loudspeakers do in real rooms.

Sidebar: low frequency perception with headphones is skewed by the lack of "body sensation", and so there is plenty of room to fudge things one way or another. 

Expected roll off with 30 ohm load?

Reply #3 – 2006-05-05 21:32:51

maybe the headphone transducers by itself might not be all that reactive but hooked up with a thin long unbalanced cable into a highly mismatched source-load connection, its a whole new ball game.