Forbidden Region

Discussion in 'Homework Help' started by RdAdr, Dec 20, 2015.

  1. RdAdr

    Thread Starter Member

    May 19, 2013
    214
    1
    In digital circuits, we have the voltage thresholds VIL, VIH, VOL, VOH.
    And the noise margins are NM0 = VIL-VOL, NM1 = VOH-VIH

    Also, there is a forbidden region. What is the use of this forbidden region?
    What are the causes that led to defining this region?
     
  2. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,346
    Hello,

    The "forbidden" region will give undefined levels.
    Read the attached PDF for more info.

    Bertus
     
    absf likes this.
  3. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    You want your input signal to swing through the "forbidden" region as quickly as possible, and not dwell there.
     
  4. RdAdr

    Thread Starter Member

    May 19, 2013
    214
    1
    And I suppose this is because of noise, right?
    For example, maybe the digital component has an NMOS transistor. And the forbidden region will be around the threshold voltage. So if the input signal is around Vt, then a little noise can bring it either to over Vt or below Vt, and this gives unpredictable output. So we have a forbidden region at the input.

    And because we want noise margins we must also have a forbidden region at the output. Right?
     
  5. WBahn

    Moderator

    Mar 31, 2012
    17,743
    4,795
    It's not just a matter of noise, but of tolerances. In fact, even in the lack of noise, the forbidden region would still be necessary.

    Let's say that there was no forbidden region (a better name would be the "indeterminate" or the "unspecified" region). That would mean that a manufacturer would have to design their parts so that their switching threshold, for every part, was exactly at that boundary. And it would have to stay at that boundary regardless of temperature variation, or supply voltage variation, or age, or anything else. Not possible to do.

    V_IL is the highest voltage that the receiver guarantees to recognize as a low logic level. But in any given receiver the actual upper limit that it will recognize as a LO will be something above this. If you are a manufacturer of receivers, you know that your parts are ALL expected to recognize an input voltage that is equal to V_IL as a LO, and you know that you have production tolerances, so you design your parts to recognize anything below V_IL + V_something as a LO. You do the same for V_IH except that your receivers are designed to recognize anything below V_IH - V_something_else as a HI.

    For a given part, there is only a very small input range that is truly indeterminate, But the location of this range within the forbidden region is unpredictable and it moves with temperature, supply voltage, age, and everything else. The wider the forbidden region, the easier it is to design parts and, hence, the lower the cost of those parts.
     
    absf likes this.
  6. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,346
    Hello,

    In the PDF I gave you they call it "undefined":

    cmos_levels.png

    Bertus
     
    Sinus23, absf and atferrari like this.
  7. RdAdr

    Thread Starter Member

    May 19, 2013
    214
    1
    Ok. Thanks for answers.
     
  8. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    The term is probably more along the lines of; "guaranteed transfer characteristic".

    Whenever the logic level is in that forbidden region, the manufacturer makes no GUARANTEE what the output state will be.

    Its a bit symantics - but that definition is good to keep in mind when fault finding "down to the metal".
     
    atferrari likes this.
Loading...