How to Measure High and Low Impedance Accurately?

[EE4life]

Hi All,

I am measuring the impedance response of a piezoelectric component. It has a unique impedance response having a resonance portion (low impedance ~10-30ohmns) and large impedance (100k-200kohmns) within a 2kHz bandwidth. See the webpage for a figure showing the impedance behavior. The frequency is around 20kHz.

I am driving the piezo component using a high power amplifier and a function generator. I am measuring the voltage and the current to determine the impedance.

I am using a unique drive control: constant vibration level. I am monitoring the vibration level and adjusting the voltage to keep it constant. Therefore, we have large voltage (50VRMS) and low current (1mARMS) in antiresonance, and low voltage (0.2VRMS) and high current (50mARMS) in resonance

The problem is that I want to use one setup to measure both the high and low impedance cases (so I can do a computer controlled frequency sweep while measuring). In the high impedance case, there is a voltage drop across the probe, which causes error, therefore I use the voltage monitor from the amplifier. But, in the low voltage resonance case, the voltage monitor has a lot of noise and is jumpy due to the low voltage level, therefore more error.

For current, I decided to use a textronix current probe (TCP305Amplifer w/ TCPA300 probe) and wrapping the wire 10x around the probe to measure the low anti-resonance currents. I also have used a shunt resistor with a voltage probe for current.

What are your suggestions on how I can measure voltage and current accurately? I am most concerned about the current probe working at such low currents ( I do get a somewhat coherent signal) in the antiresonance and finding a voltage probe I can use for high impedance measurments.

The phase is important for me because I want to calculate power. I have found that using a probe in antriresonance increases the power significantly (~20%), even with a 100x probe with ~3pF capacitiance.

Any suggestions/help/comments are greatly appreciated.

 

[zoki85]

Looks like you want to measure impedances in a 1Ω-10⁵Ω range. Resonant/antiresonant frequencies are relatively easy to find but measuring (determining) impedances very preceisely in vicinity of these frequencies can be challenging for high-Q electro(mechanical) systems.You say you have problems with messy wave forms readings near resonances. Sharp signal jumps is typical for high-Q circuits near reasonance. Perhaps, improvement of measurement and noise reduction in readings can be achieved by employment of appropriate active low pass filters... But, are you sure your piezoelectric component has just two resonant peaks over frequency range you consider? Needless to say, a good network impedance analyzer would come in handy.

 

[EE4life]

Thanks for your comment. I am sure I have the anti-resonance and resonance.

Using the voltage monitor at resonance (low voltage), the signal get noisy, meaning random DC biases. Using AC coupling does not solve this. Using the oscilloscope averaging and AC coupling, we get a random DC bias imposed on the AC signal, which makes it look like a dancing sine wave.

I can still get relatively stable reading from the monitor, but it is not near as good as the voltage probe stability. Voltage probe is perfect around resonance.

I would ideally like to use a probe the whole way. Is there a way to do this?

 

[zoki85]

Try to feed it by sine generator exact 50V via 100-1000 Ω variable resistor in series and monitor wave forms of resistor to see what happens at resonant frequency.

 

[EE4life]

I figured out that the voltage probe used to measure current through a resistor is sucking significant power (5mW depending on the phase).
I have decided to use the current probe with a 10x wire wrap. However, I get a imposed high frequency(3times my signal freq) sine wave on my signal. How do I get rid of this?

 

[The Electrician]

Are you doing this for your employment at a company? Is this measurement needed for QC on a product your company manufactures, or is it just for R&D purposes? How many of these do you need to measure per month? Do you have an equipment budget?

 

[rafik]

Are you driving your piezoelectric component with a sine wave?

 

[zoki85]

I figured out that the voltage probe used to measure current through a resistor is sucking significant power (5mW depending on the phase).
I have decided to use the current probe with a 10x wire wrap. However, I get a imposed high frequency(3times my signal freq) sine wave on my signal. How do I get rid of this?

Using 10x more wrap wire may or may not reduce measurement consumption, but it may also cause more troubles than benefits, especially if you get close to fundamental resonant frequency of measuring system. You may get rid of higher frequency imposed to the signal with adequate filter mentioned before, but I suspect this will not improve measurement accuracy much. Regardless of that, let's consider hypothetical situation where you can make perfect measurement but you use signal generator which is not perfect. And let's just stick to the stability of it's frequency (impedance is another issue). If your piezocomponent has quality factor Q≈2000, resonant frequency exactly 20 000 Hz, and the generator's sine wave frequency varies between 19990 Hz and 20010 Hz during measurement, how will that affect circuit current? To estimate these effects use formulae listed in these notes . It should be mentioned measuring high Q circuits with network anlayzers isn't trivial either, but there are some tricks to improve on measurement (see http://home.comcast.net/~w1qg/qmeas.pdf ). Finally, if there are nonlinearities and possible saturation issues of the piezocomponent itself that can affect measurement as well.