human hearing

[sss]

we all know that most people cant hear above 20khz ,
but can we hear a 20khz square wave ?
i know there is and wount be no speaker that can do this ,but if there is can we hear it?

[li_gangyi]

I dun think it would be "hearable" maybe a bleep of 20k?? continues on...then switches it selfoff...but izzit AC?? I dunno...meaning is the square wave 20K only "on-top"??

[hugobross]

if someone is able to hear a 20khz sine wave he can hear the base tone (1st harmonic) of the 20k square wave too!

But I was told the following: in this case we can't hear the upper harmonics!! so it can be quite dangerous for damaging your ears listening to square waves at high volume!! you don't hear but it might damage your ears!! don't know whether this is right or not, and I'm not going to try it out.

best regards,

HB.

[sss]

i'm just thinking , i fwe can hear somthin like this then we should not limit our amps to 20khz

[li_gangyi]

guess ur tweeter would blow first...

[hugobross]

probably the tweeter will be damaged, exception for piezo stuff

[john curl]

It is more complicated than this. In fact, someone could make a sine-square comparison at 20KHz or even 10KHz, and 'prove' that it is inaudible.
Still, I made a test about 20 years ago with three people. I used a Pioneer ribbon tweeter with a measured response of more than 45KHz, with an Electrocompaniet, Otala based power amp, and a function generator. As I remember, I set the function generator with a 5KHz square wave and deliberately limited the risetime to 3.5us, which is about a 100K response, with a quality film polystyrene cap to ground. The function generator had a buffered 50 ohm output. Then, during the test, I added another polystyrene cap in parallel to ground to change the effective risetime to 10us, or 35KHz. We all could hear the difference. It was fairly easy to, as well. Why? I don't know, but we seem to be sensitive to rate-of-change, more than actual frequency response.

[SY]

Did you do careful measurements of the output of the driving amp and the ribbon to see if you were changing anything else unexpectedly? If I were running a test like that and differences were "easy" to hear, I'd suspect that there was a variable that I hadn't controlled for- like a small level change. Certainly the addition of the cap will do some change to the relative levels of the fundamental and the 3rd harmonic (at 15K, still within the range of normal ears).

[PMA]

Originally posted by john curl

Still, I made a test about 20 years ago with three people. As I remember, I set the function generator with a 5KHz square wave and deliberately limited the risetime to 3.5us, which is about a 100K response, with a quality film polystyrene cap to ground. The function generator had a buffered 50 ohm output. Then, during the test, I added another polystyrene cap in parallel to ground to change the effective risetime to 10us, or 35KHz. We all could hear the difference. It was fairly easy to, as well. Why? I don't know, but we seem to be sensitive to rate-of-change, more than actual frequency response.

Originally posted by SY
Did you do careful measurements of the output of the driving amp and the ribbon to see if you were changing anything else unexpectedly?

Very interesting and I would like to know the answer to SY's question.

[Steve Eddy]

Originally posted by SY
Did you do careful measurements of the output of the driving amp and the ribbon to see if you were changing anything else unexpectedly? If I were running a test like that and differences were "easy" to hear, I'd suspect that there was a variable that I hadn't controlled for- like a small level change. Certainly the addition of the cap will do some change to the relative levels of the fundamental and the 3rd harmonic (at 15K, still within the range of normal ears).

The test also seems to presume that the ribbon tweeter in question was perfectly linear and wasn't producing any intermodulation products down in the audio band.

se