In this circuit a non-locking push switch is used to activate a load. The load remains switched on until power is removed from the circuit.
The load is represented by R5 and D1, but could be a lamp, a relay or another circuit. S2 breaks power to the circuit but could be omitted altogether. If S2 is left out, then reset would be by disconnecting the power; this would mean unplugging the battery if battery powered or disconnecting from the electrical outlet.
When first plugged in (or S2 is operated) C1 charges via the base emitter junction of Q1 and hence a brief positive pulse is applied. Q1 will switch on and be saturated, its collector emitter voltage being close to zero volts. Q2 is therefore off, and the full supply voltage is applied to Q1 base via D1, R5 and R1. The circuit is now in a permanent off state.
If S1 is momentarily pressed, a high voltage is applied to Q1 collector and also Q2 base via R3. Q2 now becomes saturated and the full power to the load is applied. At the same time Q2 collector voltage is now low, and so the volatge at Q1 base, applied via R1 is also low and Q1 switches off. As Q1 is off, bias for Q2 is obtained via R2 and R3 and the circuit is now permanently latched on. Even if S1 is pressed again, this has no effect. The only way to reset is to use S2 (if fitted) or remove power source.
The transistor choice depends on the load. For low currents up to 100mA QN2222 transistors or any other general purpose transistor may be used. For higher voltages and currents, the load can be a relay, its contacts rated for the chosen load.