[Solved] Why Doesn't This Resistor Mixer Sum the Input Voltages?

Discussion in 'Analog & Mixed-Signal Design' started by johnyradio, Oct 13, 2018.

  1. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    My understanding is that parallel resistors, tied together at their outputs, sum their input voltages
    Vo= V1+ V2.. VN.

    [​IMG]

    Therefor, why in this circuit is the output peak, 5V, the same as the input peaks?
    These are 2 schmitt oscillators, btw.

    [​IMG]
     
  2. Hymie

    Active Member

    Mar 30, 2018
    514
    115
    If that was to work, I could configure 10 resistors each with an input of 5V and by some miracle there would be 50V at the commoned end.
     
    atferrari likes this.
  3. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    Yep, I hear what you're saying.

    Does that mean the formula is wrong?

    Thx
     
  4. crutschow

    Expert

    Mar 14, 2008
    19,830
    5,547
    No.
    Where did you get that "understanding"?
     
    KJ6EAD and simozz like this.
  5. ebp

    Well-Known Member

    Feb 8, 2018
    1,852
    643
    Back to basics

    in the circuit at #1
    use just two equal value resistors, one with the "input side" connected to 10 volts, the other with the input side connected to 0 volts. Assume no current flows into the "output"
    Calculate the voltage at the output.

    now use three resistors, two connected to 10 V and one to 0

    now use three resistors, two connected to 10 V and one connected to 5 V

    These are all easy to calculate.
     
    Last edited: Oct 13, 2018
  6. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
  7. ebp

    Well-Known Member

    Feb 8, 2018
    1,852
    643
    Something else to do:

    Repeat all of the above, but with the output short-circuited to "ground" and instead of calculating voltages, calculate the currents.

    Try another with different resistor values and three different voltages.
     
  8. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    To calculate, I only need the correct formula. Seems that website has the wrong formula.

    Should be following, correct?

    Vo= (V1+ V2.. VN) / N
     
    Last edited: Oct 13, 2018
  9. Hymie

    Active Member

    Mar 30, 2018
    514
    115
    Perhaps you omitted the all-important op-amp to the summing circuit.

    Summing Op-amp.jpg
     
  10. ebp

    Well-Known Member

    Feb 8, 2018
    1,852
    643
    You should be able to solve the first two-resistor problem with Ohm's law (Kirchhoff's voltage law also applies, but this really should be a "by inspection" case). Don't just look for formulas to plug values into. This is absolutely fundamental stuff and you need to be able to work out at least the simpler problems easily. Sometimes redrawing the circuit can help by making things a bit more visually obvious.

    Ohm's law and Kirchhoff's laws go a long way to solving simple circuits. You MUST understand these laws.

    You also have access to a simulator. Try using it to find the answer and to show you details of voltages and currents. See if you can figure out how to arrive at the same answers.

    Do you know how to find solutions of problems using simultaneous equations? (not required for the problems I suggested, but often very useful)
     
    johnyradio likes this.
  11. MrChips

    Moderator

    Oct 2, 2009
    16,909
    5,214
    The formula is wrong. Why?
    Because the correct formula is the sum of the currents, as in Kirchhoff's Current Law, not voltage.
    For the sum of currents formula to be correct, you are missing something. What is missing?
    You are missing a path for the current to flow. Supply a path for the sum of the currents to flow then the current formula will work.
    Add an op-amp to the circuit to convert from current to voltage. Then the voltage formula will work.
     
    johnyradio likes this.
  12. crutschow

    Expert

    Mar 14, 2008
    19,830
    5,547
    Yes, but only for equal value resistors.
    As an exercise, determine why that's true.
     
    Last edited: Oct 13, 2018
  13. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    Thx for all replies.

    In that case, what if each input is the supply voltage?
     
  14. Plamen

    Member

    Mar 29, 2015
    68
    6
    Petkan:
    Your assumed formula Vout=V1+V2...etc is not correct. Clearly if you applay say 1V to all inputs, you will get 1V at the output.
    Use superposition to calculate the transfer function. Signal sources are considered one at a time with the rest shorted. The total response is the sum of individual responses. Load at the output must be considered too as it affects the result.
     
    johnyradio likes this.
  15. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    I def misread that article. My bad. More careful reading shows (I think) that Vo is the average of all Vin's.

    Solved.
     
  16. Hymie

    Active Member

    Mar 30, 2018
    514
    115
    If all the inputs are at the supply voltage, then the op-amp output will be driven to the supply voltage (actually the negative supply voltage in this case, as it is an inverting summing configuration).

    The same is true if the summed voltages equals or exceeds the supply voltage.
     
  17. ArakelTheDragon

    Active Member

    Nov 18, 2016
    822
    48
    You will need more basic reading before you can do this circuit.
     
    johnyradio likes this.
  18. johnyradio

    Thread Starter Active Member

    Oct 26, 2012
    185
    5
    thx for taking time to reply.

    Thx to @crutschow for suggestions.

    You're right, I've no more questions on this topic . The op has been answered, op amp is off topic.

    Cheers!
     
  19. MrChips

    Moderator

    Oct 2, 2009
    16,909
    5,214
    Wrong. If the signal is AC coupled it can go negative.
     
    johnyradio likes this.
Loading...