January 1, 1998

Slow op amp makes fast multivibrator

Lukasz Sliwczynski, University of Mining and Metallurgy, Krakow, Poland

A classic one-shot multivibrator using an op amp (Figure 1a) needs a relatively long-duration triggering pulse, even though the circuit is in theory a slope-triggered configuration. This limitation can be a problem when you want to use a "spare" op amp from a dual or quad package (LM324, LM358, LM6482, or LT1013, for example) triggered from fast logic circuitry. The op amp's limited slew rate (SR) creates the problem. A duration in the order of t=V_LH/SR is necessary for the positive-feedback loop through R₁ and R₂ to capture the output in the high state. (V_LH is the op amp's output swing.) This value of t is an approximation, because the exact value depends on the amplitude of the triggering pulse, its shape, and the form of the op amp's output transition. A simple modification (Figure 1b) of the basic circuit allows you to trigger the one-shot with a narrow pulse; for example, 5 nsec.

The method uses the nonlinear properties and speed of a basic emitter follower. When the triggering pulse appears, Q₁ turns on and the low-value capacitor, C₁, rapidly charges to the voltage level V_C1=V_T­V_BE, where V_T is the amplitude of the triggering pulse and V_BE is Q₁'s base-emitter voltage, approximately 0.7V. When the triggering pulse ends, Q₁ turns off and C₁ slowly discharges through R₁ and simultaneously charges through R₂ from the linearly increasing voltage from the op amp's output. The circuit acts as a simple pulse stretcher, giving some extra time for the other, slower parts     of the circuit. Q₁ replaces diode D₁ in Figure 1a's basic circuit to separate the trigger source from the loading effects of capacitor C₁. The one-shot's pulse width is

 (1)

(2)

provided that

 (3)

Failing the condition of Equation 2, the circuit forms a free-running oscillator with an approximate 50% duty cycle. The low-value resistor R₅ may be necessary to ensure that C₁ discharges faster than C₂ when the R₁||R₂C₁ time constant is higher than R₃||r_DC₂, where r_D is the dynamic resistance of D₂. Figure 2 shows the triggering waveforms. V_C1(t) and V_C2(t) are the voltages on C₁ and C₂, respectively. Calculating the precise value of C₁ entails solving some complicated nonlinear equations. It's much easier to test the circuit empirically with various capacitor values. The circuit in Figure 1b generates a pulse of approximately 700-µsec width with a 5-nsec, 2V triggering pulse. You can lower the triggering-pulse amplitude by approximately 0.7V by using optional resistor R_B1 to forward-bias Q₁'s base-emitter voltage. (DI #2141)

| EDN Access | Feedback | Table of Contents |