3
\$\begingroup\$

I'm not sure how to solve the following problem (I'm not asking or complete solutions, just for a hint on where to start!). I'm asked to find the values for R1, R2 and R3. I know it is a voltage divider and depending on the switch's position, R1 and R3 or R2 and R3 are parallel. But just plugging that in gives me two equations for three unknowns.

The problem has the following specifications: V1 = 10V DC, R1 + R2 = 10kΩ, when the switch is in Pos.1, voltage at VM1 should be 6V, in Pos.2 4V.

Thank you for your help!

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
3
  • 1
    \$\begingroup\$ That should give you a degree of freedom, as there are infinite possible resistor combinations giving the same voltages. The ratios are important. \$\endgroup\$
    – Eugene Sh.
    Commented Jun 15, 2017 at 18:30
  • 1
    \$\begingroup\$ You know R1 + R2 = 10k, so you can reduce your 3 unknowns to 2 by substitution. Replace every instance of R2 with (10k-R1) or the other way around if you prefer. \$\endgroup\$
    – brhans
    Commented Jun 15, 2017 at 18:45
  • 1
    \$\begingroup\$ @Pilotf4 You really should read Spehro's comment to M.Ferru. His first point about the symmetry (4 V and 6 V output values are symmetrically arranged relative to the upper and lower rails) is probably the most important simplification and it really helps a lot. I'll provide a less directly useful, but far more general approach in order to make the Spehro's point clearer by showing what happens if you don't take advantage of symmetries you find, in practice. It also shows how a simple three-resistor problem can bloom, algebraically. \$\endgroup\$
    – jonk
    Commented Jun 15, 2017 at 22:45

3 Answers 3

3
\$\begingroup\$

Spehro's comment elsewhere here (to M.Ferru) about symmetry (the two input voltages and output voltages are arranged symmetrically) provides a very large and obvious simplification -- that it must be the case that \$R_1=R_2\$. This immediately tells the values for those two, given your other input to the problem (\$R_1+R_2=10\:\textrm{k}\Omega\$.) From there it should only take you a moment or two to work out the value of \$R_3\$.

But I want to provide a less directly useful, but far more general approach in order to make Spehro's point all the clearer to you by showing what happens if you don't take advantage of symmetries you find (or if they don't exist.) It also shows how a simple three-resistor problem can bloom, algebraically.

Let's look at this in its most general form:

schematic

simulate this circuit – Schematic created using CircuitLab

There are two equations in play:

$$\begin{align*} V_{OUT_{MAX}}&=\tfrac{V_{IN_{MAX}}\cdot R_{HI}\cdot R_{LO}+V_{LO}\cdot R_{HI}\cdot R_{IN}+V_{HI}\cdot R_{LO}\cdot R_{IN}}{R_{HI}\cdot R_{LO}+R_{HI}\cdot R_{IN}+R_{LO}\cdot R_{IN}}\\\\ V_{OUT_{MIN}}&=\tfrac{V_{IN_{MIN}}\cdot R_{HI}\cdot R_{LO}+V_{LO}\cdot R_{HI}\cdot R_{IN}+V_{HI}\cdot R_{LO}\cdot R_{IN}}{R_{HI}\cdot R_{LO}+R_{HI}\cdot R_{IN}+R_{LO}\cdot R_{IN}} \end{align*}$$

And this solves out as:

$$\begin{align*} \tfrac{R_{HI}}{R_{LO}}&=\tfrac{V_{HI}\cdot\left(V_{IN_{MAX}}-V_{IN_{MIN}}\right) + V_{OUT_{MAX}}\cdot\left(V_{IN_{MIN}}-V_{HI}\right)-V_{OUT_{MIN}}\cdot\left(V_{IN_{MAX}}-V_{HI}\right)}{V_{OUT_{MIN}}\cdot\left(V_{IN_{MAX}}-V_{LO}\right)-V_{LO}\cdot\left(V_{IN_{MAX}}-V_{IN_{MIN}}\right) - V_{OUT_{MAX}}\cdot\left(V_{IN_{MIN}}-V_{LO}\right)}\\\\ \tfrac{R_{IN}}{R_{LO}}&=\tfrac{V_{HI}\cdot\left(V_{IN_{MAX}}-V_{IN_{MIN}}\right) + V_{OUT_{MAX}}\cdot\left(V_{IN_{MIN}}-V_{HI}\right)-V_{OUT_{MIN}}\cdot\left(V_{IN_{MAX}}-V_{HI}\right)}{\left(V_{HI} - V_{LO}\right)\cdot\left(V_{OUT_{MAX}} - V_{OUT_{MIN}}\right)} \end{align*}$$

Plugging in your values:

$$\begin{align*} V_{OUT_{MIN}}&=4\:\textrm{V}\\\\ V_{OUT_{MAX}}&=6\:\textrm{V}\\\\ V_{LO}=V_{IN_{MIN}}&=0\:\textrm{V}\\\\\ V_{HI}=V_{IN_{MAX}}&=10\:\textrm{V} \end{align*}$$

You get:

$$\begin{align*} \frac{R_{HI}}{R_{LO}}&=1&\therefore R_{HI}= R_{LO}\\\\ \frac{R_{IN}}{R_{LO}}&=2&\therefore R_{IN}= 2\cdot R_{LO} \end{align*}$$

Note that all this work has led to the same point that Spehro made, right away: that \$R_1=R_2\$. You need to develop an eye for these symmetries.

\$\endgroup\$
0
1
\$\begingroup\$

There many solutions to your problem. And by saying "many", I mean a lot ! A way to solve it is just to fix a resistor value. By doing so, you will find the two other resistors value.

If you add "resistor has to be standardize one" in your problem input, you will find just few solutions.

\$\endgroup\$
6
  • \$\begingroup\$ Thanks for your answer! If i get (infinetely) many solutions, is it possible to make it very simple and just assume that R1 = R2 ? \$\endgroup\$
    – Pilotf4
    Commented Jun 15, 2017 at 19:00
  • \$\begingroup\$ Yes you can because you will have 1 unknown resistor for 2 equations. \$\endgroup\$
    – M.Ferru
    Commented Jun 15, 2017 at 19:04
  • \$\begingroup\$ 2 unknown for two equations * \$\endgroup\$
    – M.Ferru
    Commented Jun 15, 2017 at 19:13
  • 1
    \$\begingroup\$ If you assume R1 = R2 then you already know R1 = R2 = 10k/2 = 5kOhm hence, R1/(R2||R3) = 6/4 = 1.5---> (R1||R3)= 5kOhm/1.5 = 3.33kOhm and from there we have R3 = 1/(1/3.33 - 1/5) = 10kOhm \$\endgroup\$
    – G36
    Commented Jun 15, 2017 at 19:18
  • 3
    \$\begingroup\$ From symmetry you know R1 = R2. From given info R1 = R2 = 5K. So R1||R2 = 2.5K, so the common point is a Thevenin equivalent 5V source with 2.5K source resistance and therefore R3 must be 4 * 2.5K (using either position to solve). But do it the hard way, because not every problem will have an easy solution. And remember this because it might be on an exam and you want to be the guy or gal that finishes the proof in 30 seconds and has time to check all the answers 3x. \$\endgroup\$ Commented Jun 15, 2017 at 19:46
1
\$\begingroup\$

You have three unknowns and three equations. The rest is just linear algebra.

The three things you know are:

  1. R1 + R2 = 10 kΩ

  2. The output voltage when the switch is in postion 1:

      V1 * R2 / (R1//R3 + R2) = Vpos1

  3. The output voltage when the switch is in postion 2:

      V1 * (R3//R2) / (R1 + R2//R3) = Vpos2

Get crack'n.

\$\endgroup\$

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.