74HC123 and 74HC221 monostables
Two one-shots per chip: a trigger makes one pulse on Q, and a resistor and a
capacitor set how long it lasts.
They stretch a short pulse into a long one, turn an edge into a pulse of a known width, debounce a switch, or (on the 74HC123) watch for a signal that has stopped. The two chips share a pinout and a model; the difference is retriggering. Aliases: 74123, 74221, SN74HC123, CD74HC123, CD74HC221, 74LS123, one-shot, monostable multivibrator, pulse stretcher, retriggerable monostable.
- 74HC123: retriggerable. A trigger during the pulse starts the timing again, so the pulse runs on for a whole width from the last trigger.
- 74HC221: not retriggerable. It ignores triggers until its pulse is over.
Pins
| Pin | Name | What it does |
|---|---|---|
| 1 | 1A |
One-shot 1 trigger, active low (falling edge). |
| 2 | 1B |
One-shot 1 trigger, active high (rising edge). |
| 3 | 1RD |
One-shot 1 reset, active low. |
| 13, 4 | 1Q, 1QN |
One-shot 1 pulse output and its complement. |
| 14 | 1CEXT |
One-shot 1 timing capacitor, one side. |
| 15 | 1RCX |
One-shot 1 timing capacitor, other side, and the timing resistor (REXT/CEXT on the data sheet). |
| 9 | 2A |
One-shot 2 trigger, active low. |
| 10 | 2B |
One-shot 2 trigger, active high. |
| 11 | 2RD |
One-shot 2 reset, active low. |
| 5, 12 | 2Q, 2QN |
One-shot 2 pulse output and its complement. |
| 6 | 2CEXT |
One-shot 2 timing capacitor, one side. |
| 7 | 2RCX |
One-shot 2 timing capacitor, other side, and the timing resistor. |
| 8 | GND |
Ground. |
| 16 | VCC |
Supply, 2 V to 6 V. |
For each one-shot, the capacitor CX goes between nCEXT and nRCX, and the resistor
RX from nRCX to VCC.
Truth table
Inside the chip nA, nB and nRD meet in one gate, NOT A AND B AND RD, and its
rising edge fires the pulse:
A |
B |
RD |
What happens |
|---|---|---|---|
| falling | 1 | 1 | Pulse |
| 0 | rising | 1 | Pulse |
| 0 | 1 | rising | Pulse on the 74HC123; nothing on the 74HC221 |
| x | x | 0 | No pulse; a pulse in progress ends at once, Q low |
| anything else | 1 | Nothing new |
The pulse width
The width is the data sheet's tW = K RX CX: K = 0.45 for the 74HC123 (at 5 V)
and 0.7 for the 74HC221 (at 4.5 V), the CD74HC123 and CD74HC221 typicals. So 10 k and
100 nF give about 0.45 ms on a 74HC123 and 0.7 ms on a 74HC221.
A trigger discharges CX, then RX charges it, and the pulse ends when nRCX climbs
past a reference, placed so the width comes out at K RX CX: 36% of the supply on the
74HC123, 50% on the 74HC221.
Properties
None.
What the model gets right
The capacitor really charges. The model discharges the capacitor on the board and lets the resistor on the board charge it, and asks the analog solver to land on the crossing that ends the pulse. A different resistor or capacitor, or a leaky one, changes the pulse the way it does on the bench.
Retriggering on the 74HC123 and its absence on the 74HC221 are both as the data sheets
describe, and so is the reset that ends a pulse at once. Delays from a trigger to Q:
25 ns on the 74HC123 and 18 ns on the 74HC221 (the CD74HC123 and CD74HC221 typicals at
5 V and 15 pF). The chip's own power-up is not read as a trigger, and both one-shots
power up idle, Q low.
What it does not model
K's variation with the supply and with small capacitors (the data sheets give curves for it), the minimum resistor and capacitor the data sheets ask for, and the discharge transistor's own resistance beyond a 74HC output's. There is no 74HCT or 74LS version.
Common mistakes
Wiring RX to the wrong pin: it goes from nRCX to VCC, with CX between nRCX and
nCEXT. Leaving nRD floating: it is active low, so tie it to VCC, or the one-shot
never fires. Leaving the trigger you are not using floating: to trigger on a falling
A, tie B high; to trigger on a rising B, tie A low. Expecting a 74HC221 to
stretch a burst of triggers into one long pulse: that takes the retriggerable 74HC123.
See it in action
Browse the example circuits at /templates.