Simple method for determining Speaker Imp?

[default]

A bridge circuit is an unnecessary complication.

This whole exercise is an unnecessary complication . . . the bridge
removes the subjective element and lends the exercise some scientific
credibility.

Complicated? How? You have a bridge with two identical resistors in
the ground legs, a calibrated pot in the adjustable leg, the device
under test in the opposite leg. Tune for null on a meter across the
bridge and measure the resistance directly. It is too easy.

No. You would only be listening for a volume *difference*. That's the
beauty of the scheme. No matter how nonlinear your ear might be to volume
changes, you can pick out volume differences very accurately. As soon as your
ear cannot detect the difference between the attenuator settings, then the
"bridge" is balanced and the pot setting is equal to the speaker's impedance.

You can pick out volume differences as well as a meter that can
resolve less than a 100 microvolt null? I wouldn't bet I could do
that.

Exactly! Using my suggestion, you could determine the speaker's
impedance at *any* frequency that interests you. In an enclosure or bare.
Isn't that a lot better than trusting the manufacturer's nominal impedance
rating?

Well the bridge does the same thing, more accurately.

My primary point was that it is a wasted effort.

So after all this toil and trouble you do know the speaker impedance
at all frequencies of interest - you plot it out - you have a pretty
graph for your effort, and you impressed the hell out of the
salesperson pushing the speakers - it tells you nothing of the sound
output at all frequencies. The efficiency of the speaker and speaker
in a cabinet will be different at various frequencies.

Then after all that, you will connect it to an amplifier that may have
damping characteristics that color the sound, an amplifier with a
feedback network that will color the sound, a crossover network that
colors the sound, a room that colors the sound, phase differences in
the speakers, diffraction in the speaker cabinets and articles in the
room, Etc. Etc. so you can play music for ears that have some unknown
frequency response, to be interpreted by someone's own psycho acoustic
bias, while listening to a ventilation system with a bad bearing or
some other outside noise.

OK you have golden ears - you live in a cathedral (a very quiet
cathedral designed by an expert acoustician) buried in the woods and
the wind isn't blowing, and there's 15 feet of snow covering
everything in sight. Absolutely optimal conditions for listening . .
.. Just how important is speaker impedance given the breadth and depth
and intricacies and variables that make up a sound system?

Were you planning on listening to Aerosmith, on warped, scratched
vinyl, on a turntable with a slipping belt, from an analog master, or
some really exceptional Dorian classical music digitally mastered?

The final nail - we are probably talking about bargain $10 speakers
here since you don't have the impedance and no plot of the sound
output under laboratory conditions.

This way insanity lies . . .

Have fun; fun might be the redeeming feature to all this.

 

[Arch-lab]

Pretty neat idea Jim, this is what I was looking for some creative suggestions.
In the back of my head I was already thinking of something like this, but
simpler like just a sine audio output from a walkman or something... I suspect
this idea can be simplified a little too... or use a meter to avoid the
sound...

I think in the end the DC meter is all I need because they always tend to
either have 4 or 8 ohm speakers and that is all I want to know. To the guy not
following the thread, yes we all know the impeadance varies with the frequency
;-) .
I just wanted a quick and dirty method to check which is which, there are quite
a few neat little tricks and short cuts out there, sort of like some of the
stuff I've heard from Bob Pease (sp?).

 

[Arch-lab]

"default" - that's a cool idea too...

 

[N. Thornton]

Complicated? How?

Hi. Its unusual here that such complex and unuseful methods are being
suggested for an elementary problem. If you've worked with speakers
long enough you'll know the answer is stick a multimeter on it, job
done. It really is that simple.

As for trying to get accuracy, you can't. Speakers dont operate at the
nominal impedance, they just operate nominally around there. To try
and pin an mc LS to within half an ohm at 8 ohms would be optimistic.


Regards, NT

 

[Ralph & Diane Barone]

This whole exercise is an unnecessary complication . . . the bridge
removes the subjective element and lends the exercise some scientific
credibility.

Complicated? How? You have a bridge with two identical resistors in
the ground legs, a calibrated pot in the adjustable leg, the device
under test in the opposite leg. Tune for null on a meter across the
bridge and measure the resistance directly. It is too easy.

I suspect that the reactance of the speaker would result in an inability to
completely null the bridge.

An easier approach would be to feed the speaker through a 1 k ohm resistor.
Measure the voltage across the 1 k resistor and that gives you the current
(in mA) through the speaker. Measure the voltage across the speaker and
V/I gives you the impedance of the speaker at that frequency.

Or just measure the damn thing with an ohmmeter and multiply the reading by
1.3. Close enough...

 

[Ban]

Ralph & Diane Barone wrote:
|||||| Connect the system to a speaker and set the generator to a
|||||| frequency that is appropriate for the speaker you're testing.
|||||| Switch the attenuator switch back and forth while you vary the
|||||| pot setting. At some point you will be able to hear that the
|||||| volume level at either switch setting will be the same. If
|||||| you've calibrated the pot and added a scale, you will be able to
|||||| read the speaker's impedance on the scale.
||||||
|||||| Jim
|||||
||||| Seems like a lot of effort. Why not use the amp/generator to
||||| feed a balanced bridge and calibrate the pot directly in ohms?
||||| You can do that with your ohm meter.
||||
|||| A bridge circuit is an unnecessary complication.
||| This whole exercise is an unnecessary complication . . . the bridge
||| removes the subjective element and lends the exercise some
||| scientific credibility.
|||
||| Complicated? How? You have a bridge with two identical resistors
||| in
||| the ground legs, a calibrated pot in the adjustable leg, the device
||| under test in the opposite leg. Tune for null on a meter across the
||| bridge and measure the resistance directly. It is too easy.
||
|| I suspect that the reactance of the speaker would result in an
|| inability to completely null the bridge.
||
|| An easier approach would be to feed the speaker through a 1 k ohm
|| resistor. Measure the voltage across the 1 k resistor and that
|| gives you the current (in mA) through the speaker. Measure the
|| voltage across the speaker and V/I gives you the impedance of the
|| speaker at that frequency.
||
|| Or just measure the damn thing with an ohmmeter and multiply the
|| reading by
|| 1.3. Close enough...

I can recommend the software Hobbybox
http://www.audio-software.de/downloads.html
http://www.audio-software.de/hobbybx51.zip
is in German, but has english help-file to download.
with this demo you can use your soundcard to play the MLS test-signal and
record the impedance and when you have a microphon also the frequency and
impulse response.

You need one power resistor of around 10 Ohms and an audio-amp. you can also
determine the Thiele-Small parameters of the speaker-chassis with different
methods.

I think this will solve all points to everybody's contentment.

ciao Ban