[sdiy] Speaker Impedance

Bob Weigel sounddoctorin at imt.net
Mon May 23 20:51:21 CEST 2005


The speaker impedence with a transistor amp won't matter usually unless 
you go below.  I mean in terms of safety.  You will obviously get less 
peak power out of a speaker of higher impedence since it restricts the 
flow of charge more.
    If on the other hand you are using a tube amp, it's not recommended 
that you use higher impedences.  You can actually short out a tube amp's 
output totally without anything blowing up but on a transistor amp the 
outputs will blow up if you do that.  The typical tube amp uses an 
output transformer to isolate the high voltages from the speaker that 
are used on the plate of the output tubes.  (On a push/pull amp for 
instance the HV of 300-700V usually depending on the type of output tube 
used and how much headroom the designer wanted, is tied to the center 
tap primary of the output xformer and the other two outer legs of the 
primary are tied to opposing plates of output tubes...or tube pairs or 
however many they decided to paralell.  The tubes are then biased so 
they are in a more linear response zone and a phase inverter drives one 
side opposite the other to get the 'push pull' effect.)   In such 
circuits, the output xformer secondary is usually very low dc impedence 
(tenth of an ohm or so).   The tubes can cycle completely on without 
burning out anything; being internally limited by the availability of 
excited electrons on the cathode and often even further by naturally 
sagging power supply when tube rectifiers are used.  However when NO 
load is present on the output, the back emf created can do some bad things.
      Ok all that said, let's talk about just how much sound you'll get 
out of a given speaker.  There is a vast difference in speaker 
efficiency which winds up being often more material than the impedence 
believe it or not.  Measure the dc resistance of an 8 ohm speaker for 
example.  What do you get? If you say 7 ohms you are probably looking at 
a speaker that isn't designed for much efficiency.  Why?  Because that 7 
ohms you are measuring isn't capable of making ANY SOUND (except 
sizzling if you get it too hot! :-) ).  That's right.  No inductive 
effect results from what you just measured and it WILL make heat.  100% 
of that 7/8ths of the voltage drop will go into creating nothing but 
heat.  1/8th if it is TRULY an 8ohm speaker, will go into making sound 
and the efficiency of how that is used is also contingient on things 
like the strength of the magnetic field it is immersed in.
       Speaker efficiencies in rating charts vary from 85db at 1W/1M to 
102db or so for woofers and 112db or so for tweeters down to 90db.  
Double the power, bump up 3db.  So in other words what a 100W amp will 
do on a higher end speaker at a given peak frequency it might take a 
2500W amp or so to get that same level out of the lower end speaker!  
:-)  Thing is, those ratings are often at a peak frequency and the more 
efficient speaker might tend to be less flat so there's another factor.
        Anyway if you double the impedence on a transistor amp you will 
get approximately half the power (plus any relief in power supply sag).  
-Bob

Rafael_Cohen at prusec.com wrote:

>Hey everybody,
>Guess what... it's another newbie question!!
>
>I have some schematics requiring 8 ohm speakers, and in some of my
>DIY/scavenging I come across speakers of various other impedance, usually
>higher than 8ohm.  Everything from 16ohms to 45 ohms... sometimes computer
>speakers or speakerphones, etc...
>
>My question is, what do I need to do in order to adapt the circuit to
>speakers of different imedance?  Also, is there a simple way to test the
>imedance of a speaker?
>
>Thanks again,
>Rafael
>
>
>
>  
>



More information about the Synth-diy mailing list