De-bouncing circuit

**broken link removed**

this link shows the problem of contact bounce and the two solutions of it( digital and analogue ) at analogue solution he used schmitt triger ic to prevent errors com fron the unknown area.

i don't understand what is the source of errors at this area ?
please help ,very urgent

The source of errors is exactly what it says it is, the logic state in that voltage range is not defined the circuit appears to oscillate when the voltage is the unknown range, the Schmidt trigger provides a clean vertical square wave output even if the output from the analog debouncer causes a soft knee when the switch is released.

It may be curious to note that this will also cause the digital circuit to read logic low for a slightly longer period of time than the switch is actually pressed because of the R/C time constant which is what causes the soft knee when the switch is released.

With a standard logic circuit the positive going and negative going logic thresholds are essentially the same, so if you slowly move the input signal through this logic threshold, as the output switches there may be enough stray positive feedback (either internal or external to the chip) to cause the circuit to oscillate. A Schmidt trigger has significant difference between the positive and negative going thresholds that this feedback can't cause oscillations.

You might try reading this. http://www.ganssle.com/debouncing.htm

It is worth noting that debounce circuits don't always need the later Schmidt trigger such as in the case of micro controllers or other more sophisticated IC's as many of them have built in Schmidt triggers on their inputs. This will be stated in their datasheet.