Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homemade speaker Q's: calculate at what freq for impedance?

  1. Nov 24, 2014 #1
    Audible hearing range for humans: Roughly 20-20khz

    That's almost 20,000 values to choose from:eek:



    Other/additional info:

    I don't understand your question BeautifulLight. If your circuit is operating at 2khz, then calculate for 2khz. Simple. Well, this is for a speaker. Human speech can make a pretty complex sinusoid. There will be no one given frequency.

    Have any of you guys/gals handled 30AWG magnet wire? I've used 22AWG in crystal sets, but that's way too thick of wire to be used for voice coils in speakers. My supplier carries 32AWG, but I'm hesitant it'll be too fine to work with.

    I dedicate this last statement to complain about the cost of rare earth magnets.
     
    Last edited: Nov 24, 2014
  2. jcsd
  3. Nov 24, 2014 #2

    Doug Huffman

    User Avatar
    Gold Member

    Considering that the human sense is analogue and not digital, 20-20K is an infinity of frequencies.
     
  4. Nov 24, 2014 #3

    donpacino

    User Avatar
    Gold Member

    If speaker impedance is only evaluated at one frequency, it is typically evaluated at 1k
     
  5. Nov 24, 2014 #4
    I didn't know you could hear a half frequency. It's easy to imagine an infinite set for light (i.e. colors fading into one another), but not so much sound.

    edit: I'm wrong again. Frequency is an element of the objective world. You can't hear frequencies. What humans hear is pitch -a perceptual process. Even so, a half pitch still sounds bizarre to me... Oh, and when I said objective world, I mean a construction of science.



    Because human speech tends to stay around 1khz? Just an educated guess.
     
  6. Nov 24, 2014 #5

    donpacino

    User Avatar
    Gold Member

    The human voice is ~300-3k depending on the person.
    I would guess the reason for 1k is most likely because it is a nice round about in the middle of the human hearing spectrum.
     
  7. Nov 24, 2014 #6

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    Oh you cannot be serious! I mean about half frequencies/pitch. You are telling me that it is impossible to have a sound that is 800.5 hertz? It's fine that we have 800 hertz but if the frequency of the oscillator is changed slightly that it will automatically bump to 801 hertz? THAT is what sounds bizarre to me.
    -
    As for human speech staying around 1000 hertz goes, I suspect that if we were to limit the passband to fairly narrow (relative term) at 1000 hertz as compared with 2000 hertz, more peoples voices would be conveyed in the 1000 hertz passband as compared to the 2000 hertz passband.
     
  8. Nov 24, 2014 #7

    donpacino

    User Avatar
    Gold Member

    Dont forget that speaker are for much more than just human speach
     
  9. Nov 24, 2014 #8

    meBigGuy

    User Avatar
    Gold Member

    Indeed, the impedance of a speaker will vary across the audio range. One article ( http://en.wikipedia.org/wiki/Electrical_characteristics_of_dynamic_loudspeakers ) says the nominal impedance is 1.15 times the minimum impedance across the band of interest.
    Regardless, 1KHz sounds like a fine place to evaluate the impedance. It is high enough that the enclosure doesn't have a huge effect (I think)

    Realize that the impedance will change with load, for example measuring across the band will give different curves ( at the low end, mostly) depending on whether the speaker is in an enclosure. Loading the speaker into a standard test enclosure and measuring impedance is one way to determine speaker characteristics for the design of a properly sized and/or ported enclosure.
     
  10. Nov 24, 2014 #9
    Do you know what the Planck scale is? What Planck did for space, I did for sound. Instead of asking what the smallest unit of space was, I asked myself what the smallest unit of sound was. Okay, let's rephrase that to something like "fundamentally, what do you need in order to produce a sound?". An oscillation! At its simplest level, that would be 1 cycle per second, so 1 hertz. So I labelled 1 hertz as the lowest frequency in order to produce a sound, and anything below that wouldn't make noise ...like the .5 hertz suggested by you!

    Do you see now?

    By the way, I never had a problem with generating non-whole number tones such as 800.5 hertz or something crazy like 21.857694867859 hertz. It was hearing the extra half or eighth frequency that got me. Fundamentally, it wasn't enough to produce "a sound" right? I certainly have my moments.

    Oh, and what is the lowest frequency you can have in order to produce a sound?

    It's not 20hz. Forget the audible range. I'm just talking about longitudinal pressure waves in general. I assume it goes to infinity and approaches (but never reaches) 0 hertz?
     
  11. Nov 24, 2014 #10
    I knew I came across this dilemma on another forum. Here,

    "...you're assuming 'dead centre' is the algebraic mean of the extremes. But why not the harmonic mean? For sound I'd say dead centre was 632Hz being the square root of the product of the two extremes, because sound pitch is perceived in ratiometric terms.
    "
     
  12. Nov 24, 2014 #11

    meBigGuy

    User Avatar
    Gold Member

    No "oscillation" is required to produce a "sound". Just a pressure change in a medium. It could be a single impulse, or anything. Even then, the concept of "sound" might be construed to be what you can hear, which is significantly different from the absolute pressure changes that might have stimulated your hearing.

    I'm not sure what are are getting on about now, BTW. There is no "optimum" or "official" frequency to evaluate a speaker at. You pick what is useful to you for what you are trying to determine.
     
  13. Nov 24, 2014 #12
    I created this thread because I wanted to make a voice coil with a specific ohm rating so I wouldn't damage my amplifier. Every speaker has one. You often see it listed on or near the terminals of the speaker or the manufacturer has it stamped on the bottom of the magnet. Could be 2, 4, 8, or even 16 ohm. I knew impedance rose with frequency, so I was unsure which frequency to calculate for.

    Well I guess this is all wrong. Those 2, 4, 8, etc. figures I've been looking at having nothing to do with impedance. They are simply the DC resistance of the voice coil. Now I'm utterly confused. So when people say they are matching their 4 ohm (says it right on the terminals, right?) impedance loudspeaker to their amplifier, they are really saying they are matching their 4 ohm DC resistance loudspeaker to their amplifier?

    So now when I go to wind my voice coil, all I need to worry about is a DC resistance of 4 ohms?
     
  14. Nov 24, 2014 #13

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

  15. Nov 25, 2014 #14

    donpacino

    User Avatar
    Gold Member

    Like I said earlier when they give 1 value it is usually at 1kohm
     
  16. Nov 25, 2014 #15

    donpacino

    User Avatar
    Gold Member

    Sound is simply a mechanical wave. It is infinite.
    and a 0.5 hz sound is a thing
     
  17. Nov 25, 2014 #16

    meBigGuy

    User Avatar
    Gold Member

    The stated impedance of a speaker is NOT (NEVER, EVER) the dc resistance. It seems like you pay little attention to what people post. Please read the wikipedia article I referred to earlier. http://en.wikipedia.org/wiki/Electr...ic_loudspeakers#Load_impedance_and_amplifiers We post reference articles for a reason.

    For the purposes of not having amplifier problems, you can use the DC resistance, since it is easiest to measure. The impedance will be "a bit" higher, but if you want you can measure it at 1KHz and everything will be accurate enough.
     
  18. Jan 4, 2015 #17
    meBigGuy, if you were to record your favorite song and play it back, would you agree that the microphone would spit out the exact words you sang into it?

    I think so. Speaking of the source of the signal alone, it's almost as if your voice and the microphone are interchangeable, right?

    This is where I went wrong.

    We don't need an oscillation to produce a sound, correct? Okay, so let's take a body and move it from point A to point B and then set up a microphone and a scope at sme distance, x.

    Move the body from point A to point B and tell me what the scope registered. If I understand how sound travels, then the diaphragm should have moved in one direction and then the other creating a sine wave.


    I think you probably see where this is going. If I were to play that sound back again with the microphone, would the diaphragm only move in one direction now? Remember, the source of original signal was just a body that moved in one direction. from point A to point B, and stopped. You'd think that if the source (now our microphone) moved differently (now oscillatory -not just like a body moving from point A to point B and stopping), that the sound would be different. But it's not. Its still spitting out your favorite song.

    I'm not sure if you're following me but it seems strange to me.
     
    Last edited: Jan 4, 2015
  19. Jan 4, 2015 #18

    Averagesupernova

    User Avatar
    Science Advisor
    Gold Member

    You are grasping at straws here. An object moving from point A to point B is an event that occurred. The sound generated by said event is almost never a sine wave. It is a complex group of waves. These waves will be oscillatory in nature but the cause was not an oscillation. I think an oscillator needs to be defined as more than picking up a glass and moving it to a different part of the table. When set down the impact will surely causing ringing in the glass and other objects involved. That is the source of the sound waves.
     
  20. Jan 4, 2015 #19

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    There's a standard that says DC resistance is something like 70%-80% of nominal impedance
    so an 8 ohm speaker will read 5 or 6 ohms

    i dont know why they chose that fraction , perhaps to protect today's DC coupled amplifiers ... seems to me that some really old speakers were less.

    Look up "cochlea" which is the organ that translates sound into nerve impulses. There's some minimum frequency difference that a human ear can distinguish.

    Here's an article by an engineer from a guitar amp company that does a decent job of speaker basics.
    http://peavey.com/support/technotes/concepts/impedance.cfm
    the amplifier and interconnecting wires get into the act as well, see last page.

    Thiele's 1971 paper is here:

    http://diyaudioprojects.com/Technic...ed-Boxes-Part-I.pdf#page=10&zoom=auto,-99,285

    and

    http://diyaudioprojects.com/Technical/Papers/Vented-Box-Loudspeaker-Systems-Part-II.pdf

    he's math intensive and i've not yet mastered his technique
    but you fellows will whiz right through it.

    For my first enclosure I used Olson's 1947 book on acoustics - will see if i still have it around. Ever since i just go to an online program for Thiele method , without really understanding why it works.

    old jim
     
    Last edited: Jan 4, 2015
  21. Jan 4, 2015 #20

    dlgoff

    User Avatar
    Science Advisor
    Gold Member

    You seem to be a visual person, so here's a good place for getting animations on Waves and Sound. They have a lot more areas; even things about Loudspeakers (which are really linear motors)
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Homemade speaker Q's: calculate at what freq for impedance?
  1. Impedance matching Q's (Replies: 1)

  2. Speaker's impedance (Replies: 4)

  3. What is impedance? (Replies: 10)

Loading...