Is there easy way to measure speaker impedance

[grogi.giant]

Is there an easy way to, without more sophisticated equipment, measure the speaker impedance at certain frequency?

Let's say I have a frequency generator (read 'a laptop'), a DC/AC multimeter and an amplifier... ??

 

[amirm]

Sure. It is actually fairly simple and doesn't need fancy equipment at all. Here is a quick tutorial from Audio Precision on the setup: https://www.ap.com/technical-librar...edance-using-apx-derived-measurement-results/

 

[Flaesh]

Impedance Measurement

(REW)
No multimeter needed; a resistor and some wires are needed.

 

[Jon AA]

Dayton Audio - DATS V3 Computer Based Audio Component Test System

DATS V3 Computer Based Audio Component Test System

www.daytonaudio.com

So easy a caveman can use it.

 

[notsodeadlizard]

I'll be honest with you, I don't know.

But here's what I know as a longtime non-audio electronics designer: Measuring of impedances is the classic "hard" problem in electronics, it was, it is, and it always will be.

You need two good fast synchronous measurement ADCs to measure both current and voltage at the same time and evaluate the phases,
you generally need a good shunt (because non-contact current measurement is indicative to put it mildly), the shunt can be just a piece of wire to a speaker of known length and resistance,
you you need a very specific voltage drop amplifier on the shunt (a rather funny and elegant little thing),
and all this must be supported by some program.

Above you were given a link to a "simple jig" for measuring impedance, which uses the obvious "wire as a shunt" technique, and everything else is hidden in the implementation of a wonderful audio analyzer,

Unfortunately, there are no other options.
Fortunately, you can do it yourself, we live in a wonderful time of availability of the most unthinkable components.
But there is also a reality in which such a project is not small, it requires a serious investment of time and effort.
Alas...

 

[SIY]

I'll be honest with you, I don't know.

But here's what I know as a longtime non-audio electronics designer: Measuring of impedances is the classic "hard" problem in electronics, it was, it is, and it always will be.

You need two good fast synchronous measurement ADCs to measure both current and voltage at the same time and evaluate the phases,
you generally need a good shunt (because non-contact current measurement is indicative to put it mildly), the shunt can be just a piece of wire to a speaker of known length and resistance,
you you need a very specific voltage drop amplifier on the shunt (a rather funny and elegant little thing),
and all this must be supported by some program.

Above you were given a link to a "simple jig" for measuring impedance, which uses the obvious "wire as a shunt" technique, and everything else is hidden in the implementation of a wonderful audio analyzer,

Unfortunately, there are no other options.
Fortunately, you can do it yourself, we live in a wonderful time of availability of the most unthinkable components.
But there is also a reality in which such a project is not small, it requires a serious investment of time and effort.
Alas...

Wow, I never knew something so simple was so complicated.

Oh wait, it isn't.

Anyone with a soundcard, a small power amp. and a few resistors can get excellent impedance measurement results for loudspeakers.

 

[notsodeadlizard]

Wow, I never knew something so simple was so complicated.

Oh wait, it isn't.

Anyone with a soundcard and a few resistors can get excellent impedance measurement results for loudspeakers.

There are probably miracles in this world, but if they happened, then such a good company as Audio Precision would have gone bankrupt long ago.
But this doesn’t happen to them for some reason I don’t understand, especially considering the specifics of their equipment and the very small market for it.
I think so.

 

[SIY]

for some reason I don’t understand,

That part is correct. The rest of us will do a simple measurement using simple means and get good data.

 

[notsodeadlizard]

Wow, I never knew something so simple was so complicated.

Oh wait, it isn't.

Anyone with a soundcard, a small power amp. and a few resistors can get excellent impedance measurement results for loudspeakers.

By the way, there are almost all of the above units except for the: shunt voltage drop amplifier is avoided and a shunt with the too big resistance is used instead (simply because such the instrumentation amplifier is far from trivial here, especially when it comes to a wide range of amplifier output voltages) and synchronous instrumental (non-audio) ADCs are replaced by the soundcard's audio ADC (that is why calibrations are used, but no one knows how good they are in a metrological sense):

Impedance Measurement

As I understand it, some purely spectral impedance estimation is implemented here in the region of small signals, and despite the rather large shunt resistance (>33 Ohm), there are still some nuances with the signal-to-noise ratio.
It is not forbidden to do this, but you should not idealize low-signal estimates, this is still not quite a solution to the problem of "impedance measurement".

 

[Galliardist]

There are probably miracles in this world, but if they happened, then such a good company as Audio Precision would have gone bankrupt long ago.
But this doesn’t happen to them for some reason I don’t understand, especially considering the specifics of their equipment and the very small market for it.
I think so.

Haven't several of their rivals pulled out of the audio analyser business in the last few years, though?

Rohde & Schwarz for example...

 

[notsodeadlizard]

That part is correct. The rest of us will do a simple measurement using simple means and get good data.

I'm not even going to argue here.
It's just that in my practice I'm used to understanding well why measurements are taken and what exactly is measured, because these determine the method of measurement.
In the case of "impedance measurements" (to be precise, we are talking about impedance estimation because there are no direct methods for measuring impedance at all), the very principle of obtaining raw data for evaluation is fundamental and it is difficult to come up with something new.

But here's what's really interesting - the speaker impedance scanner is not some monstrously expensive device and is quite releasable for use in the "slightly above budget" price segment.
If such a scanner has an open digital interface, the evaluation result can be obtained by a third party DSP for use as a crossover.
This is already interesting to me, so thanks for the dialogue, sometimes interesting things are too obvious to notice.
I think I found my entertainment for a few months.

 

[Flaesh]

measure the speaker impedance

Why\for what purpose do you even need these measurements?

 

[notsodeadlizard]

Haven't several of their rivals pulled out of the audio analyser business in the last few years, though?

Rohde & Schwarz for example...

The audio market is too small for R&S, this is a segment where a few specialized companies will remain, simply because there are enough of them in this segment.
I think so.

 

[grogi.giant]

Impedance Measurement

(REW)
No multimeter needed; a resistor and some wires are needed.

It never crossed my mind that driving the speakers to audible level is not needed and even the miniscule power a soundcard can generate will be sufficient. In the end the impedance will not change with volume

I've dig up a pile of reclaimed resistors - should be enough to build the measuring jig.

There are probably miracles in this world, but if they happened, then such a good company as Audio Precision would have gone bankrupt long ago.
But this doesn’t happen to them for some reason I don’t understand, especially considering the specifics of their equipment and the very small market for it.
I think so.

Where AP excels is noise levels etc. -120dB is pretty common for this equipment...

Impedance measurement however does not need that kind of pureness and even if you get 1/100 noise during your measurement (which would be catastrophic when measuring noise or distortion), it won't matter much. Impedance (both amplitude and phase) acts linear, not logarithmic.

We are also talking about measuring in 20-20000Hz range. That is considered DC by many electrical engineers Measuring impedance in GHz range will be much more difficult...

Why do you even need these measurements?

Endless curiosity... In the context of TPA325x based amplifiers I want to know how much "too bright" my speakers are

 

[Flaesh]

In the end the impedance will not change with volume

Nope). At least the coil, with accurate measurements, will heat up measurably before the end comes.

 

[Flaesh]

One more manual https://www.artalabs.hr/AppNotes/LIMP_Tutorial_Version_2_4_English.pdf (LIMP is part of ARTA)

 

[grogi.giant]

One more manual https://www.artalabs.hr/AppNotes/LIMP_Tutorial_Version_2_4_English.pdf (LIMP is part of ARTA)

That's effectively the same device as for the REW manual you quoted before

However, there are much more additional consideration in the further parts of the publication. Good read

 

[grogi.giant]

Sooo... It seems my wire-jig is working

Thank you all for contributing

 

[egellings]

Wow, I never knew something so simple was so complicated.

Oh wait, it isn't.

Anyone with a soundcard, a small power amp. and a few resistors can get excellent impedance measurement results for loudspeakers.

You need to measure the amplitude of the voltage & current and the angle between them at the same instant if the signal is varying over time. If the signal is unchanging over time in all characteristics, then you could measure voltage, and then current, and after getting a beer, catch the phase angle. For something like an audio signal, you have to snap them all at once.

 

[grogi.giant]

You need to measure the amplitude of the voltage & current and the angle between them at the same instant if the signal is varying over time. If the signal is unchanging over time in all characteristics, then you could measure voltage, and then current, and after getting a beer, catch the phase angle. For something like an audio signal, you have to snap them all at once.

But it is also you who controls the source... You might not change the signal for whole afternoon if that's what you need to capture all three values...