I have the following simulated system (ngspice):

enter image description here

The sensor may be something that generates pulses or a slow moving DC voltage, for example. The output could be a coil in a relay or pulses, for example. All of the signals are slower than 500Hz.

I would like to measure the impedance of Vout1 and Vout2 to make sure that it is within the allowed range for the signal conditioning circuit of a diagnostic device I am designing.

My method is as follows:

  1. Run the simulation and measure the average voltage of Vout1
  2. Connect a resistor (R1) from Vout1 to ground
  3. Run the simulation again and vary R1 until the average voltage of Vout1 is half of what it was in step 1

Is this a valid method?


1 Answer 1


Is this a valid method?

It's a valid method but won't really tell you very much.

I would use an external AC excitation method and drive a sinusoidal AC signal onto the output through a known impedance (\$R\$). So, for example, if you choose an excitation level of 1 volt p-p and you see 10 mV p-p on your output node, you can calculate the true output impedance because you know what value \$R\$ is.

Then, you can vary the frequency across the spectrum you require and develop a graph of output impedance vs frequency. This will be far more useful than just trying to measure it at DC.

You can then develop it to measure the complex impedance of your output signal. All really easy in a simulator.

  • \$\begingroup\$ Thanks. Applying the sine wave breaks the simulation unfortunately - it's very brittle, probably at the limits of SPICE. \$\endgroup\$
    – Andy
    Oct 14, 2023 at 21:55
  • \$\begingroup\$ It shouldn't be a problem even for ltspice. Maybe post your simulation circuit so it can be examined. \$\endgroup\$
    – Andy aka
    Oct 14, 2023 at 23:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.