Editing Schmitt trigger (section)

=== Transistor Schmitt triggers ===
==== Classic emitter-coupled circuit ====
[[File:Schmitt trigger with transistors.svg|right|thumb|327px|Schmitt trigger implemented by two [[Emitter-coupled logic|emitter-coupled BJTs]] stages.]]

The original Schmitt trigger is based on the [[#dynamic threshold|dynamic threshold]] idea that is implemented by a [[voltage divider]] with a switchable upper leg (the collector resistors R<sub>C1</sub> and R<sub>C2</sub>) and a steady lower leg (R<sub>E</sub>). Q1 acts as a [[comparator]] with a [[differential input]] (Q1 base-emitter junction) consisting of an inverting (Q1 base) and a non-inverting (Q1 emitter) inputs. The input voltage is applied to the inverting input; the output voltage of the voltage divider is applied to the non-inverting input thus determining its threshold. The comparator output drives the second [[common collector]] stage Q2 (an ''emitter follower'') through the voltage divider R<sub>1</sub>-R<sub>2</sub>. The emitter-coupled transistors Q1 and Q2 actually compose an electronic [[Switch#Contact terminology|double throw switch]] that switches over the upper legs of the voltage divider and changes the threshold in a different (to the input voltage) direction.

This configuration can be considered as a [[differential amplifier]] with series positive feedback between its non-inverting input (Q2 base) and output (Q1 collector) that forces the transition process. There is also a smaller negative feedback introduced by the emitter resistor R<sub>E</sub>. To make the positive feedback dominate over the negative one and to obtain a hysteresis, the proportion between the two collector resistors is chosen so that R<sub>C1</sub> > R<sub>C2</sub>. Thus less current flows through and there is less voltage drop across R<sub>E</sub> when Q1 is switched on than in the case when Q2 is switched on. As a result, the circuit has two different thresholds in regard to ground (V<sub>−</sub> in the image).

===== Operation =====
'''Initial state.''' For the NPN transistors shown on the right, imagine the input voltage is below the shared emitter voltage (high threshold for concreteness) so that the Q1 base-emitter junction is reverse-biased and Q1 does not conduct. The Q2 base voltage is determined by the divider described above so that Q2 is conducting and the trigger output is in the low state. The two resistors R<sub>C2</sub> and R<sub>E</sub> form another voltage divider that determines the high threshold. Neglecting V<sub>BE</sub>, the high threshold value is approximately

:<math>V_\mathrm{HT} = \frac{R_\mathrm{E}}{R_\mathrm{E} + R_\mathrm{C2}}{V_+}</math>.

The output voltage is low but well above ground. It is approximately equal to the high threshold and may not be low enough to be a logical zero for subsequent digital circuits. This may require an additional level shifting circuit following the trigger circuit.

'''Crossing up the high threshold.''' When the input voltage (Q1 base voltage) rises slightly above the voltage across the emitter resistor R<sub>E</sub> (the high threshold), Q1 begins conducting. Its collector voltage goes down and Q2 starts toward cutoff, because the voltage divider now provides lower Q2 base voltage. The common emitter voltage follows this change and goes down, making Q1 conduct more. The current begins to steer from the right leg of the circuit to the left one. Although Q1 is conducting more, it passes less current through R<sub>E</sub> (since R<sub>C1</sub> > R<sub>C2</sub>); the emitter voltage continues dropping and the effective Q1 base-emitter voltage continuously increases. This avalanche-like process continues until Q1 becomes completely turned on (saturated) and Q2 turned off. The trigger transitions to the high state and the output (Q2's collector) voltage is close to V+. Now the two resistors R<sub>C1</sub> and R<sub>E</sub> form a voltage divider that determines the low threshold. Its value is approximately

:<math>V_\mathrm{LT} = \frac{R_\mathrm{E}}{R_\mathrm{E} + R_\mathrm{C1}}{V_+}</math>.

'''Crossing down the low threshold.''' With the trigger now in the high state, if the input voltage drops enough (below the low threshold), Q1 begins cutting off. Its collector current reduces; as a result, the shared emitter voltage drops slightly and Q1's collector voltage rises significantly.  The R<sub>1</sub>-R<sub>2</sub> voltage divider conveys this change to the Q2 base voltage and it begins conducting. The voltage across R<sub>E</sub> rises, further reducing the Q1 base-emitter potential in the same avalanche-like manner, and Q1 ceases to conduct. Q2 becomes completely turned on (saturated) and the output voltage becomes low again.

===== Variations =====
[[Image:Schmitt trigger inverted symbol.svg|thumb|right|200px|Symbol depicting an inverting Schmitt trigger by showing an inverted [[hysteresis]] curve inside a [[buffer amplifier|buffer]]. Other symbols show a hysteresis curve (which may be inverting or non-inverting) embedded in a buffer followed by a bubble, which is similar to the traditional symbol for a [[inverter (logic gate)|digital inverter]] that shows a buffer followed by a bubble. In general, the direction of the Schmitt trigger (inverting or non-inverting) is not necessarily clear from the symbol because multiple conventions are used, even with the same manufacturer. There are several factors leading to such ambiguity,<ref group="nb">One factor contributing to ambiguity is that one simple transistor-based realization of a Schmitt trigger is naturally inverting, with a non-inverting Schmitt trigger sometimes consisting of such an inverting implementation followed by an inverter. An additional inverter may be added for buffering a stand-alone inverting configuration. Consequently, inverting configurations within an integrated circuit may be naturally inverting, while non-inverting configurations are implemented with a single inverter, and stand-alone inverting configurations may be implemented with two inverters. As a result, symbols that combine inverting bubbles and hysteresis curves may be using the hysteresis curve to describe the entire device or the embedded Schmitt trigger only.</ref> These circumstances may warrant a closer investigation of the documentation for each particular Schmitt trigger.]]

'''Non-inverting circuit.''' The classic non-inverting Schmitt trigger can be turned into an inverting trigger by taking V<sub>out</sub> from the emitters instead of from a Q2 collector. In this configuration, the output voltage is equal to the dynamic threshold (the shared emitter voltage) and both the output levels stay away from the supply rails. Another disadvantage is that the load changes the thresholds so, it has to be high enough. The base resistor R<sub>B</sub> is obligatory to prevent the impact of the input voltage through Q1 base-emitter junction on the emitter voltage.

'''Direct-coupled circuit.''' To simplify the circuit, the R<sub>1</sub>&ndash;R<sub>2</sub> voltage divider can be omitted connecting Q1 collector directly to Q2 base. The base resistor R<sub>B</sub> can be omitted as well so that the input voltage source drives directly Q1's base.<ref>[http://www.datasheetcatalog.org/datasheets/400/334439_DS.pdf 7414 datasheet]</ref> In this case, the common emitter voltage and Q1 collector voltage are not suitable for outputs. Only Q2 collector should be used as an output since, when the input voltage exceeds the high threshold and Q1 saturates, its base-emitter junction is forward biased and transfers the input voltage variations directly to the emitters. As a result, the common emitter voltage and Q1 collector voltage follow the input voltage. This situation is typical for over-driven transistor [[Differential amplifier#Long-tailed a nice pair|differential amplifiers]] and [[Emitter-coupled logic#Operation|ECL]] gates.

==== Collector-base coupled circuit ====
[[Image:Schmitt parallel.svg|thumb|200px|[[Bipolar junction transistor|BJT]] bistable collector-base coupled circuit can be converted to a Schmitt trigger by connecting an additional base resistor to one of the bases]]

Like every latch, the fundamental collector-base coupled [[Latch (electronics)#Basic bistable circuit|bistable circuit]] operates with hysteresis. It can be converted to a Schmitt trigger by connecting an additional base resistor R to one of the inputs (Q1's base in the figure). The two resistors R and R<sub>4</sub> form a parallel voltage summer (the circle in  the block diagram [[#Fundamental idea|above]]) that sums output (Q2's collector) voltage and the input voltage, and drives the single-ended transistor "comparator" Q1. When the base voltage crosses the threshold (V<sub>BE0</sub> ∞ 0.65 V) in either direction, a part of Q2's collector voltage is added in the same direction to the input voltage. Thus the output [[#modified input|modifies]] the input voltage by means of parallel positive feedback and does not affect the threshold (the base-emitter voltage).