Anyone knowing how to get the input impedance of a device under large AC signal

The voltage is large AC signal, not small AC. I never deal with this kind of biasing...
I am doing some simulations to extract the impedance of the device. I heard about some ways to measure it, like load pull, S parameter. But I can't not do measurement now...

The device is quit nonlinear, when I input sinωt signal, do transient response simulation and then Fourier transformation. The output current contains DC component +sin(ωt+Φ1)+ sin(ωt+Φ2)+... How do I know the impedance for this case ??

The reason that I want to know the device impedance is that I want to calculate the voltage drop when I connect the device with other circuit component.

Anyone know about this ? Thanks !!!

Dear soul_ls
Hi
Can you give me more explanation please? i confused that what do you need clearly !
Best Wishes
Goldsmith

In reality, one would feed a "extremely non-linear" device with a high impedance, so the output current is more independent of the non-linearities. Which bit of the output current are you interested in?
Frank

I simplify the device configuration in the following plots.
=====ITO electrode===
-----------------------

water

-----------------------
Nonlinear device
-----------------------
=====ITO electrode===


My goal is to get the electrical field distribution within the water. I want to use the equivalent circuit to solve the problem. If I know the input impedance of the nonlinear device, I can treat the circuit as the voltage divider. So I can get the voltage drop on the water layer to get the electric field. However, the voltage used between the two electrode is a large AC signal, say Vsinωt. I did transient simulation for the device to get the output current of the device - Io +I1sin(ωt+Φ1)+ I2sin(ωt+Φ2)+...

But I don't know how should I define the input impedance this case. Since V/I or dV/dI seems meaningless. Some of my friends tell me to define Z= Vsinωt/I1sin(ωt+Φ1), i.e., the impedance at a certain frequency. But I checked the other components, they are not so small compared with I1sin(ωt+Φ1). I guess I can not neglect them.

I am wondering whether should I treat the circuit as voltage divider since the signal is large AC and the device is nonlinear. I don't know if my direction is right or wrong

Calibrate out your non linearities:- Set the probes up to some point, increase the voltage between them and measure the increasing amplitude of the fundamental at the output. Now you have a relationship between the amplitude of the fundamental as an input and as an output. move your probes around and measure the fundamental output. Apply your correction factor. . .
Frank

chuckey said:

Calibrate out your non linearities:- Set the probes up to some point, increase the voltage between them and measure the increasing amplitude of the fundamental at the output. Now you have a relationship between the amplitude of the fundamental as an input and as an output. move your probes around and measure the fundamental output. Apply your correction factor. . .
Frank

Click to expand...

Hi, Frank,

Thanks for your reply. I am not able to measure the device at this stage ... I only do the simulations and try to get a model to characterize its characteristics. I don't understand what you mean by saying " increase the voltage between them and measure the increasing amplitude of the fundamental at the output. " I got DC I-V curve. I did transient simulations at different voltage and frequency and get the current in time domain. But does that give me some info of the impedance of this device ? How should I deal with current harmonic components. Thanks

You can't use DC, it electrolysis the water and the electrodes get covered in a layer of insulating gas, reducing their surface area. You say you have done simulations, what have you taken the conductance of the water as?
Frank

Gnerally, the impedance is a complex frequency dependent quantity. In the nonlinear case, it's additionally voltage dependent. It's however unclear, which quantity you want to measure. Is it the impedance for a particular frequency, or the full "impedance spectrum"?

If the input signal is a sinewave, the impedance can be quite easily determined by synchronous demodulation of the acquired current, either using anlog or digital means.

chuckey said:

You can't use DC, it electrolysis the water and the electrodes get covered in a layer of insulating gas, reducing their surface area. You say you have done simulations, what have you taken the conductance of the water as?
Frank

Click to expand...

Hi, Frank,

Yes. You are right. The experiments use AC signal. The conductance that I use is the equivalent conductance of water at that frequency, i.e., Z(w)_water. If I know the impedance for the device Z(w)_device, I can calculate the voltage drop on within the water layer so as to get the electric field. But the output current of the device has a lot of harmonic component at frequencies. So I am not sure if I can use the voltage divider to calculate

---------- Post added at 16:44 ---------- Previous post was at 16:25 ----------

FvM said:

Gnerally, the impedance is a complex frequency dependent quantity. In the nonlinear case, it's additionally voltage dependent. It's however unclear, which quantity you want to measure. Is it the impedance for a particular frequency, or the full "impedance spectrum"?

If the input signal is a sinewave, the impedance can be quite easily determined by synchronous demodulation of the acquired current, either using anlog or digital means.

Click to expand...

Hi, FvM,

Thanks for your reply. I would like to know the Z(w, V) for the device so I can know the voltage drop on the water layer when I apply different sinusoidal voltage at different frequencies on the electrodes.

----V------
|
Z(water)
|
|
Z(device)
|
----GND-----

So V(water)=V* Z(device)/(Z(water)+Z(device))


But I am not sure if the other harmonic current (now sinwt, but Dc or sin2wt term) will cause a voltage drop or not? Or if it is right to calculate using the voltage divider in this case. If yes, how to calculate the voltage drop...

Assuming that the involved impedances are not purely resistive, the voltage divider or whatever calculation you want to perform has to use complex numbers. As said, it should only consider the fundamental, filter out the harmonics. If you don't have phase information with your measurements, you need to aquire all three magnitudes Vin, Vdevice and Vwater to determine the complex impedances.