How to measure impedance?
James P. Meyer
On 22 Aug 1996, Skip wrote:
> Can someone explain how to make a set-up for measuring output and input
> impedances of a preamp? I bleieve I have the necessary meters, oscilloscope's
> etc... I just need an explanation and a list of parts needed.
Impedance is a complex term. That means that in order to be 100%
accurate, you need to be able to measure inductance, capacitance, and
resistance, and you need to do it at every frequency. I suggest you rely
on the specifications given you by the manufacturer of whatever it is
you're concerned with.
Jim
stuartk
Hello,
To measure output impedance, what I would do, as a first
approximation, is get a scope or RMS volt-meter with a high
input impedance (relative to the device under test, meaning
just about any of them) and a bunch of resistors whose value
(and current rating) is in the ball park of what you think
the output impedance might be. Note - DON'T try to measure
the output impedance of power amps this way unless you
fully understand what you are up against! Remember, they
usually have output impedances of less than an ohm,
sometimes much less, can pump many amps, and will
valiantly attempt to vaporize your resistor...
Hook up the measuring device and play a tune, noise or a tone
of some frequency (say 1KHz) into the measuring device alone.
You will measure an RMS value if your measuring device
has RMS capability. If it doesn't, just go by the approximate
peak. Then load the output with your resistor by connecting
one end of the resistor to the signal and the other to the
ground. The peak or RMS value will drop by some amount. Find
or extrapolate the resistance that cuts the peak or RMS value
in half - that value of resistance is roughly the output
impedance of the device you are measuring.
To find the input impedance, do the same with an external signal
source of low output impedance relative to the expected input
impedance of the device, passing the signal through a series
connected resistor into the device. Measure the strength of the
signal at the *input* before and after the resistor is in place
(ignore the device output). As above, the value of R that cuts
the signal in half will be approximately the input impedance
of the device.
-stu.
Skip
Can someone explain how to make a set-up for measuring output and input
impedances of a preamp? I bleieve I have the necessary meters, oscilloscope's
etc... I just need an explanation and a list of parts needed.
Thanks
-Wes
James P. Meyer
On Thu, 22 Aug 1996, stuartk wrote:
> approximation, is get a scope or RMS volt-meter with a high
> input impedance (relative to the device under test, meaning
> just about any of them) and a bunch of resistors whose value
> (and current rating) is in the ball park of what you think
> the output impedance might be.
Using resistors alone, the only thing you can measure is the
resistive portion of the impedance. That is only part of the picture.
Jim
stuartk
In article
<Pine.SOL.3.91.960823...@godzilla5.acpub.duke.edu>,
"James P. Meyer" <jim...@acpub.duke.edu> wrote:
>
> Using resistors alone, the only thing you can measure is the
> resistive portion of the impedance. That is only part of the picture.
>
> Jim
Of course, but why do you think equipment manufacturers specify
in/output impedance in terms of resistance?
I remain confident that the techniques I described will work quite well
given the intentions of the original poster.
-stu.
James P. Meyer
On Fri, 23 Aug 1996, stuartk wrote:
> > Using resistors alone, the only thing you can measure is the
> > resistive portion of the impedance. That is only part of the picture.
> >
>
> Of course, but why do you think equipment manufacturers specify
> in/output impedance in terms of resistance?
Resistance is usually the major, but not the only, component of
impedance. The question was "How to measure impedance?". If you say what
you mean, and mean what you say, you can't measure impedance with a
resistor. If someone can't take the manufacturer's figures and needs to
do the measurement for themselves, then they probably want measurements
more accurate and complete than what the manufacturer gave them. That's
why I mentioned the fact that impedance is not totally resistance.
Jim
ter...@foothill.net
Ok, not practical.
Richard D Pierce
In article <Pine.SOL.3.91.96082...@godzilla5.acpub.duke.edu>,
James P. Meyer <jim...@acpub.duke.edu> wrote:
>On Fri, 23 Aug 1996, stuartk wrote:
>
>> > Using resistors alone, the only thing you can measure is the
>> > resistive portion of the impedance. That is only part of the picture.
>> >
>>
>> Of course, but why do you think equipment manufacturers specify
>> in/output impedance in terms of resistance?
>
> Resistance is usually the major, but not the only, component of
>impedance. The question was "How to measure impedance?". If you say what
>you mean, and mean what you say, you can't measure impedance with a
>resistor. If someone can't take the manufacturer's figures and needs to
>do the measurement for themselves, then they probably want measurements
>more accurate and complete than what the manufacturer gave them. That's
>why I mentioned the fact that impedance is not totally resistance.
In the case of the impedances presented by the input to amplifiers and
loudspeakers and the like, we find that the values are sufficiently
constyrained that all we need to measure is the impedance magnitude to
derive everything we need to know. You'll be hard pressed to find, for
example, ANY such device whose input impedance does not present,
essentially, a minimum-phase characteristic. Further than that, you'll be
hard-pressed to find any of these devices whose complex impedance is NOT
cmpletely constrained to the first and fourth quandrant of the complex s
plane.
Given this, simply measuring the impedance magnitude using any of the
simple techniques mentioned here (for example, driving it withg a current
cource and measuring the voltage magnitude across the device, etc.) gives
you ALL the data you need to derive complex impedance. Once you have the
impedance magnitude, get the log impedance and from that derive the phase
using the Hilbert transform. Once you have that, magnitude and phase
together can be used to derive both the resistive and reactive parts.
In every case where I have done that, the complex impedance derived by
the above method and that mearured directly has matched to within a few
percent. This is for input impedances to amplifiers (which present VERY
simple loads, looking like notion more than a simple mesh of resistance
and capacitance) or multi-way vented-box loudspeakers (which prevents a
hugely complex combination of resistances, capacitances and inductances,
often requiring an equivalent circuit description of several dozen
components just to describe the impedance reasonably accurately).
Basically, measuring the impedance magnitude vs frequency alone is enought
to get you everything you need to know about the complex impedance. Of
course, an insufficient number of badly measured data points is useless,
but let's at least presume a properly done dataset.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |
stuartk
In article
<Pine.SOL.3.91.96082...@godzilla5.acpub.duke.edu>, "James
P. Meyer" <jim...@acpub.duke.edu> wrote:
If someone can't take the manufacturer's figures and needs to
> do the measurement for themselves, then they probably want measurements
> more accurate and complete than what the manufacturer gave them.
Well, just for example, my CD player has a manufacturers specification
of ">50K ohms" for output impedance. Huh??? That can't really be true,
can it? Or maybe the poster of the original question doesn't have the manual.
In any of these cases, I remain confident that the technique I described is
reasonable and practical, and will deliver a meaningful number.
-stu.
Gary A Howard
Rent a vector network analyzer, such as an HP 4195A, which can sweep and
display the impedance vector on a Smith or Polar chart. Note that this
is a very low power sweep, just barely enough to hear, so the reactance
of the speaker is rather ideal. (As the speaker motion increases, it
will tend to be less ideal, due to its mass.)
You can get scalar SqRt(R^2+X^2) network analyzer results with an
o'scope and a signal generator, but you also need an impedance bridge
with which you can seperate the forward and reflected power.
Gary A Howard
ga...@winternet.com
Gary A Howard
Skip:
I read your question, then other responses, then replied to yours, not
realizing that the other responses had changed my thinking from PreAmp
impedance, your goal, to Amp impedance, to speaker impedance.
There is a spec somewhere for the standardized line-level input and
output impedance. I think it is an EIA spec; Maybe someone on this
group knows.
Gary A Howard
ga...@winternet.com