I have this simple circuit. And from what I calculated the output impedance seen from the OUT node comes to about \$138.765 \Omega\$.

$$ X_c = \frac{1}{2\pi * 890*10^3 * 10^{-9}} = 178.8257\Omega $$ $$ Z_{out} = R2||X_c = \frac{220 * 178.8257}{\sqrt{220^2 + 178.8257^2}} = 138.765\Omega$$ enter image description here

With an input of 890K Hz sine wave of 2v pk-pk, the voltage at OUT node comes to is about 1.2615v pk-pk.

If I connect a resister of the same value of \$138.765 \Omega\$, the voltage at OUT node should be half ed.

enter image description here

But I see a voltage of 697.843mV instead of 630mV. What is the problem, am I calculating the output impedance wrongly?

  • 2
    \$\begingroup\$ Since you have a reactive component (the capacitor) in there, the output impedance is a complex number. In order to test this, you need a similar load impedance -- one that also includes a reactive component. \$\endgroup\$ – Dave Tweed Nov 12 '14 at 16:23
  • \$\begingroup\$ How did you arrive at \$\approx 138\Omega\$? \$\endgroup\$ – Null Nov 12 '14 at 16:26
  • \$\begingroup\$ @DaveTweed Thanks, I see. But why is the output 697mv? \$\endgroup\$ – Arjob Mukherjee Nov 12 '14 at 16:49
  • \$\begingroup\$ Because you have failed to account for the phase shifts, which are different with/without the load resistor. \$\endgroup\$ – Dave Tweed Nov 12 '14 at 17:02
  • \$\begingroup\$ Z= R+Jwc wc= capacitive reactance complex conjugate of impedance \$\endgroup\$ – user58210 Nov 13 '14 at 14:04

The output impedance can be calculated from 1/(1/Xc + 1/R)

Xc = \$1 \over j\omega C \$

R = 220\$\Omega\$

The admittances add in quadrature since one is imaginary and the other is real- It is easy to show(tm) that the magnitude of the impedance Z is as follows:

|Z| = \$ R \cdot (\frac {1}{\omega C}) \over \sqrt{R^2 + \frac {1}{\omega^2 C^2}}\$

That is the reason you are seeing a higher voltage than you expect. The situation will be different again with the added resistor, so you have to consider the above equation with the two resistors effectively in parallel.

  • \$\begingroup\$ I thought |Z| = (Xc*R)/sqrt(R^2 + Xc^2). \$\endgroup\$ – Arjob Mukherjee Nov 12 '14 at 17:16
  • \$\begingroup\$ @ArjobMukherjee Thanks, fixed. I'd prefer not to write Xc here, because the problem is coming up from ignoring that Xc is imaginary, and Xc is used both ways. \$\endgroup\$ – Spehro Pefhany Nov 12 '14 at 17:47
  • \$\begingroup\$ That is how I have calculated the impedance. How do I account for the phase shift? \$\endgroup\$ – Arjob Mukherjee Nov 12 '14 at 18:33
  • \$\begingroup\$ If you're just measuring the voltage, the phase is of no importance, but as I said you have to add the admittances in quadrature, so use the capacitive reactance and the Thevenin equivalent resistance (the two in parallel). \$\endgroup\$ – Spehro Pefhany Nov 12 '14 at 19:20
  • \$\begingroup\$ I am sorry, but I do not understand what In-phase and quadrature components are. Can you suggest a book or something? \$\endgroup\$ – Arjob Mukherjee Nov 13 '14 at 14:20

I finally understood what was wrong.

Without the load resistor, the impedance seen at the OUT node is a complex number. $$\dfrac{1}{Z_{out}} = \dfrac{1}{R2} + \dfrac{j}{X_c} $$ Thus, $$ Z_{out} = 87.527 - j107.678\Omega$$

With that Load resistor of 138.765 \$\Omega\$, the total impedance of the circuit becomes, $$ Z_{total} = 138.765 + 87.527 - j107.678$$ $$= 226.292 - j107.678\Omega$$ $$|Z_{total}| = 250.604 \Omega $$

If the unloaded voltage at the OUT node is 1.26v, then, the voltage at the OUT node with the LOAD resistor is a simple voltage divider, and

$$V_{out} = \dfrac{1.26 \cdot 138.765}{|Z_{total}|} $$ $$V_{out} = \dfrac{1.26 \cdot 138.765}{ 250.604} = 697.68mV. $$

The answer checks out, with acceptable error.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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