VLSI Switched capacitor not working as planend

Discussion in 'General Electronics Chat' started by Robin Mitchell, Dec 17, 2013.

  1. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    Hi everyone!

    Im currently working on an analogue assignment where you take a square wave and convert it into a sine wave. So the method that is suggested and looks nice is to use a sallen-key low pass amplifier.

    Now I have it done....if the world was ideal. My circuit works fine if I use resistors but because the second half of the assignemnt is to intergrate this thing into VLSI a 10Mega ohm resistor is not a good idea, unless you use a switched capacitor.

    So I have designed and done the math!
    R = 1/fC
    R = 10 Meg
    C = 10fF
    Therefore f should be 10Mhz. Thats fine, I made a ring oscillator to do this. So I wire it all up (in S-Edit) and it just goes

    It looks like that it has a considerbly lower resistance that it should. I checked the ring oscillator output, its fine. I replaced it with a function generator, nope, still same problem. If I remove the switch capacitors with resistors it works no question. What the darn is going on?

    The program uses T-Spice to run (its the same as regular SPICE except it has thyme in it as well).

    Here is the filter
    [​IMG]

    Here is the switched capacitor
    [​IMG]
     
  2. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    Btw i changed it to 6pf just to see the difference, still no working!
     
  3. crutschow

    Expert

    Mar 14, 2008
    13,009
    3,233
    What are the MOSFET stray capacitances?
     
  4. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    The only capacitances im aware of is the gate capacitance. Which should be around 1fF for an NMOSFET and 2fF for the P-type
     
  5. crutschow

    Expert

    Mar 14, 2008
    13,009
    3,233
    The drain and source have some capacitance to the substrate.

    Can you try it at a lower switching frequency so see if it works proportionally there?
     
    Robin Mitchell likes this.
  6. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    I think I have an idea on what the problem is.
    Assuming that SPICE is taking into account the stray capacitances, the sizes of this capacitance could be comparible to the capacitance of the switched capacitor. I will try a switched capacitor of 200fF and see if it follows the standard equation.
     
  7. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    Figured it out >:|

    Example:
    Wanted R = 2Mohm
    Used C = 500fF @ 1Mhz

    Results:
    R Actual = 1.63Mohm
    Calculated C = 611fF
    Stray C = 111fF

    Now it works fine ;)
    I have done other tests taking the 111fF into account and it works fine ^_^

    Thank you so much crutschow for your help, i would have never though about stray capacitance!
     
  8. crutschow

    Expert

    Mar 14, 2008
    13,009
    3,233
    The upper frequency limit of digital ICs is generally determined by the circuit stray capacitances.
     
    Robin Mitchell likes this.
  9. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    200
    Learn something new everyday ;)
     
Loading...