Is it possible to reliably plot a speaker's impedance with a digital oscilloscope and a signal generator, and would a dual-channel oscilloscope have any advantage?
You can do it, but it will be tedious. Take a look at the Arta software package. You can down load and use it for free. You can do all the measurements and analysis in the demo mode. If you buy it, you can save the measurements. It will export impedance measurement files for use in crossover software. Just export as ascii and select .frd file type. It includes a program called Limp that will let you measure impedance with any sound card and a single resistor. The manual explains how to make the connections. You could use the schematic in the manual to do the same with your O scope if you like. https://arta.io/ If you want to do it with the generator and the scope you just need a 100 ohm resistor in series with the speaker and generator to form a voltage divider. The ratio of voltages across the speaker / voltage across resistor should let you calculate the speaker resistance.
Last edited:
I used to only do it this way. A dual channel scope is useful to use Lissajous curves to examine phase, although impedance is minimum phase and modern simulators seem to be able to determine it.
I would measure DCR and multiply by 1.2. If you need more detail than that, you are too fussy.
Yes, I have been fussy. I had a switched 100r series resistor on a low-low-Z signal generator and plotted voltage (ACVM) vs frequency. Works same as a classic Ohmmeter except with AC signal. Accuracy is terrific in the speaker main band. When speaker Z rises to 50r you do more math; it's not so hard. Speaker Z over 100r is probably not stable enough to rely on.
If a moving coil speaker is not minimum-phase at the electric side, it is sick. (Yes, a very high efficiency motor could reflect-back some cone wonkiness, but again that's too fussy.)
Speaker efficiency over 5% it is good to have a "normal load": box, horn, etc. Over 25%, you need your horn or pipe termination or the apparent impedance is too high.
Yes, I have been fussy. I had a switched 100r series resistor on a low-low-Z signal generator and plotted voltage (ACVM) vs frequency. Works same as a classic Ohmmeter except with AC signal. Accuracy is terrific in the speaker main band. When speaker Z rises to 50r you do more math; it's not so hard. Speaker Z over 100r is probably not stable enough to rely on.
If a moving coil speaker is not minimum-phase at the electric side, it is sick. (Yes, a very high efficiency motor could reflect-back some cone wonkiness, but again that's too fussy.)
Speaker efficiency over 5% it is good to have a "normal load": box, horn, etc. Over 25%, you need your horn or pipe termination or the apparent impedance is too high.
I don't believe it is even physically possible for it to be anything but minimum phase. If you find one that is not minimum phase, you have found a perpetuum mobile. If any device with two terminals has a non-minimum-phase impedance, it has zeros in the right half plane, which means its admittance has poles in the right half plane, which means it is active.
+1 for ARTA. Pretty easy. You can even use the measurement results in LTspice afterwards.
https://preamp.org/static/ltspice-model-from-impedance-measurement-data/index.html
https://preamp.org/static/ltspice-model-from-impedance-measurement-data/index.html
I would just use a two-channel oscilloscope, an accurate resistor, two 1:1 passive probes, something that can generate sine waves and a spreadsheet. Measure the magnitudes and the phase difference and use the spreadsheet to calculate the impedance, Libre Office Calc and Excel can handle complex numbers. It takes some time, but who cares when you do these measurements once a year or less?
I put together a simple program some time back to look at speaker impedances.
You can read the story here.
The software is here.
You'll need Python 3 and GNU Radio.
This is a plot of the impedance of a car stereo speaker:
The speaker itself claims to be an 8 ohm speaker. It has a bass/midrange speaker and a tweeter with a two-way crossover.
You can read the story here.
The software is here.
You'll need Python 3 and GNU Radio.
This is a plot of the impedance of a car stereo speaker:
The speaker itself claims to be an 8 ohm speaker. It has a bass/midrange speaker and a tweeter with a two-way crossover.
Yes. You will also need a resistor for reference.
Possibly. It depends on what you want/need.
I am not sure, but given your question I would guess you're new to this stuff, and maybe you don't know how impedance measurement works. Let us know your background and we might be able to help you more.