2
\$\begingroup\$

circuit Given this circuit I’m asked to determine the transfer function from \$V_i\$ to \$V_o\$, but there are no answers given and my result disagrees with my earlier DC analysis.

I think think the first “stage” is just a voltage follower. By superposition, it’s easy to show that \$V_o’ = V_i\$ (though I'm unsure whether I'm allowed to split \$V_i\$ like that):

schematic

simulate this circuit – Schematic created using CircuitLab

For the second stage

schematic

simulate this circuit

Since \$A_2\$ is ideal, there is no current through \$C_2 \parallel R_5\$ so \$0V\$ drop across it, and the inverting port follows the non-inverting one, thus there is \$-V_i\$ across \$R_3\$, and by Ohm's law \$i = -\frac{V_i}{R_3}\$ running through it.

After which I did KVL (red arrows): $$ V_o + i(R_4 \parallel C_3) + V_i = 0\\ V_o -\frac{V_i}{R_3}(R_4 \parallel C_3) + V_i = 0\\ V_o + V_i[1 - \frac{1}{R_3}(R_4 \parallel C_3)] = 0\\ V_o = - V_i[1 - \frac{1}{R_3}(R_4 \parallel C_3)]\\ \frac{V_o}{V_i} = H(s) = \frac{1}{R_3}(R_4 \parallel C_3) - 1\\ H(s) = \frac{1}{R_3} \frac{\frac{R_4}{sC_3}}{R_4 + \frac{1}{sC_3}} - 1 = \frac{1}{R_3} \frac{R_4}{sR_4C_3 + 1} - 1\\ H(s) = \frac{R_4 - sR_3R_4C_3 - R_3}{sR_3R_4C_3 + R_3}\\ H(0) = \frac{R_4 - R_3}{R_3}\\ $$

However, I've found earlier through DC analysis (\$R_1, R_2, (C_2 \parallel R_5)\$ are short, \$C_3, C_1\$ are open) that the DC-gain should be \$1 + \frac{R_4}{R_3}\$.

I'm not even sure where to look to find my mistake.

\$\endgroup\$
4
  • \$\begingroup\$ Are you certain that \$i=-\frac{V_i}{R_3}\$. Also check the polarities for KVL voltages. \$\endgroup\$
    – RussellH
    Sep 3 at 17:18
  • \$\begingroup\$ Aha, spotted it, it's the very first line in the KVL, it's supposed to be - V_i. I'm not sure what the correct procedure here is, should I answer it myself and pick that answer? \$\endgroup\$
    – Learath2
    Sep 3 at 21:00
  • 1
    \$\begingroup\$ @Learath2 Yes, just write your own answer and select it. Are you supposed to account for input bias currents? Speed-up C2, as well? \$\endgroup\$ Sep 3 at 23:08
  • \$\begingroup\$ I see. Ideal op-amps so no input bias currents. What do you mean by "Speed-up C2"? \$\endgroup\$
    – Learath2
    Sep 4 at 12:33

2 Answers 2

1
\$\begingroup\$

The first line doing the KVL was wrong, leading to the wrong result. It's supposed to be $$ V_o + i(R_4 \parallel C_3) - V_i = 0 $$

\$\endgroup\$
0
\$\begingroup\$

If opamps are ideal, R1, R2, R5 play no role.

A1 is a buffer indeed. Note that due to virtual ground, no current flows through R6, and thus none through R7. So vin is buffered to vout.

A2 is a non-inverting stage, with gain G2 = 1 + R4/R3 = 1 + 27/9 = 4.

So, the TF(0) = 1 * 4. Which agrees with what you found before.

\$\endgroup\$
6
  • \$\begingroup\$ We seem to agree that the DC-Gain should be 4, but at the very end I find H(0) = 2, so I must have made a mistake somewhere. \$\endgroup\$
    – Learath2
    Sep 3 at 20:51
  • \$\begingroup\$ Does the first-stage buffer have any practical use (compared to the "standard" op amp buffer) or is it just an academic problem? \$\endgroup\$
    – Hyp
    Sep 4 at 10:53
  • \$\begingroup\$ @Hyp I'm not completely sure as I'm just learning these right now, but C1 makes this a Miller Integrator (a less than perfect one due to R7). So it does change the frequency response, though not sure what one might use it for :D \$\endgroup\$
    – Learath2
    Sep 4 at 12:36
  • \$\begingroup\$ @Hyp: some resistors may be employed for bias currents comepensations, which are not mentioned here though. So I'd say they're just there to test your knowledge. \$\endgroup\$
    – edmz
    Sep 4 at 17:27
  • \$\begingroup\$ @Learath2: that's not an integrator. Input current is not being integrated. \$\endgroup\$
    – edmz
    Sep 4 at 17:28

Your Answer

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

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