Measuring AC Impedance with a LCR meter

[Mike Sulzer]

Description of IV impedance measurement technique

The description of this pickup impedance measurement technique needs to cover the following parts:
1. The hardware (IV circuit)
2. The raw data collection and spectral analysis (Electroacoustics tool box)
3. The back end data processing (NLLS fitting and simpler procedures derived from better understanding of the problem).


This post concerns the IV hardware; Joe Gwinn suggested the use of the IV circuit on this forum some time ago. The schematic of the specific circuit used here is attached. It is slightly different fro the original. I use an LM 837; this is obsolete and might not be available, but there are many op amps that will do the job. You want about 20 MHz unity gain frequency or more, and a slew rate of several volts per microsecond. The basic IV circuit is the simplest op amp circuit possible. The feedback resistor Rf maintains the negative input at very near ground potential when the op amp output adjusts itself to have some voltage Vo resulting in a current io which is the negative of the current flowing through the test impedance. The value of the test impedance is the negative of the ratio of the voltage across the impedance (the input voltage) to the current flowing through it (-vo/io). The input signal could be a sign wave, and then a series of measurements needs to be made with a high quality source at different frequencies. If a broad band signal is used followed by spectal analysis, the quality of the source is less important and the entire band can be analyzed at once.

I use +/- nine volts from batteries. There is nothing wrong with a supply derived from the power line if you are careful with ground loops to the analysis system. This can be difficult sometimes.

An inverting output buffer is used so that the software does not need to invert the sign. A follower is used on the voltage sample derived from the input impedance. The resistor Rs is used to improve the dynamic range of the measurement. The magnitude of the pickup impedance can vary almost 100 to 1 over the frequency range. The proper use of Rs allows the variation to be shared between the input voltage sample and the output voltage sample, decreasing the input voltage when the magnitude of the test impedance is low. This makes it easier to avoid saturation or too low an input level in either channel. A good but not perfect rule is to make Rs and Rf equal. Rf should be set between the minimum and maximum expected impedance magnitudes, perhaps somewhat closer to the minimum.

[hasan hadi]

The description of this pickup impedance measurement technique needs to cover the following parts:
1. The hardware (IV circuit)
2. The raw data collection and spectral analysis (Electroacoustics tool box)
3. The back end data processing (NLLS fitting and simpler procedures derived from better understanding of the problem).


This post concerns the IV hardware; Joe Gwinn suggested the use of the IV circuit on this forum some time ago. The schematic of the specific circuit used here is attached. It is slightly different fro the original. I use an LM 837; this is obsolete and might not be available, but there are many op amps that will do the job. You want about 20 MHz unity gain frequency or more, and a slew rate of several volts per microsecond. The basic IV circuit is the simplest op amp circuit possible. The feedback resistor Rf maintains the negative input at very near ground potential when the op amp output adjusts itself to have some voltage Vo resulting in a current io which is the negative of the current flowing through the test impedance. The value of the test impedance is the negative of the ratio of the voltage across the impedance (the input voltage) to the current flowing through it (-vo/io). The input signal could be a sign wave, and then a series of measurements needs to be made with a high quality source at different frequencies. If a broad band signal is used followed by spectal analysis, the quality of the source is less important and the entire band can be analyzed at once.

I use +/- nine volts from batteries. There is nothing wrong with a supply derived from the power line if you are careful with ground loops to the analysis system. This can be difficult sometimes.

An inverting output buffer is used so that the software does not need to invert the sign. A follower is used on the voltage sample derived from the input impedance. The resistor Rs is used to improve the dynamic range of the measurement. The magnitude of the pickup impedance can vary almost 100 to 1 over the frequency range. The proper use of Rs allows the variation to be shared between the input voltage sample and the output voltage sample, decreasing the input voltage when the magnitude of the test impedance is low. This makes it easier to avoid saturation or too low an input level in either channel. A good but not perfect rule is to make Rs and Rf equal. Rf should be set between the minimum and maximum expected impedance magnitudes, perhaps somewhat closer to the minimum.


[ATTACH=CONFIG]13404[/ATTACH]

Thankyou masterMike Sulzer
1-this mean device LCR meter model 380193 cannot measuring impeadance(Z)
but the symbols Z and angle is founding in the screen why not active?
2-how using key PAL\SER and key FREQ 120HZ and 1KHZ
for accuracy for reading
3-When key FREQ 120HZ the reading is diffreance when using key 1KHZ why?

[Mike Sulzer]

Data sampling and spectral analysis

If one wants to use the IV circuit one frequency at a time or with a swept generator, this should work fine. One does need a high quality generator. I not tried this, preferring to use a broad band signal combined with cross spectal analysis in order to measure the entire response at once.

The measurement requires two channels, one for the voltage that appears across the test impedance, and the other for the voltage that is proportional to the current through it. You want the complex impedance, expressed as amplitude and phase or real and imaginary parts. This is best done with cross-spectral analysis, which is somewhat more involved that just measuring power spectra. Such software is not hard to write, but I am using a commercially available package for the Mac, primarily because I do not want to write the real time data collection parts of the software. The package I use is Electroacoustics toolbox {Electroacoustics Toolbox). It is not cheap, but it is good.

When using a broadband signal rather than sweeping, the frequency response of the source needs to cover the entire audio bandwidth, but it does not need to be perfectly uniform since the measurement is the ratio of the voltage and current, and so the magnitude of the signal cancels out. You do want to keep the frequency response reasonably uniform; otherwise, the signal to noise ratio of the measurement could be affected at frequencies where the amplitude is low. However, a random noise-like signal is perfectly OK and does not significantly increase how long it takes to get a good measurement since most of the randomness simply drops out.

The linearity of the D/A converter and output amplifier in the sound hardware is also not important. They just alter the signal a bit before it is measured, and that matters very little. The linearity of the input amplifier and D/A converters is more of an issue, but even here many of these errors drop out when using a random signal. The cross-spectral technique involves multiplying the ffts of the two channels and accumulating. This means that only signals that are coherent between the two channels survive the accumulation. Non-linearities that shift a portion of the signal to a different frequency produce a signal that is not coherent with the input since the input signal lacks such coherency.

You could make the measurement with a series of short pulses, but such a waveform would have low duty cycle and a less than optimum signal to noise ratio. Also such a signal can have coherency between different frequencies and you might lose some of the error rejection gained from a random signal.

The dual fft tool in Electroacoustics toolbox does the cross spectral analysis and presents the information in the form of a transfer function. (Multiplying by Rf later converts the results to impedance). It exports the accumulated data in ascii format, and I use this capability to make files for reading into my own software for further analysis.

[Achiles]

Any usege for this?
www.doc-diy.net :: simple inductance meter

[Joe Gwinn]

Any usege for this?
www.doc-diy.net :: simple inductance meter

It might work, if the inductor self-resistance isn't so high as to prevent oscillation, but the simple equation given won't be all that accurate. You would be better to build an impedance bridge.