A Beginners Guide To The Complementary Latch

This article begins with a detailed description of how a "complimentary latch" works. Then it examines the use of the latch in a practical door alarm circuit.
COMMENTS SUGGESTIONS	An Introduction To The Complementary Latch	MORE ALARM CIRCUITS
Complementary Latch Circuits	The Rest of Ron's Circuits	Write To Ron	More Free-to-Use Circuits	Circuit Exchange International
This article begins with a detailed description of how a complimentary latch works. It illustrates the operation of the latch by examining its use in a small door-alarm circuit. And finally - it discusses how the design of that circuit might be altered to produce an improved and more practical device.
A Schematic Diagram Of A
Transistor Door-Alarm That
Uses A Complementary Latch
A latch is a switch that - once it's turned on - will stay on all by itself.

And once it's turned off - will stay off all by itself.

The common household light-switch is a good example of a latch.

You supply a brief momentary input - when you turn the switch on. 

Then the mechanism takes over - and locks the switch in the on position.

A similar thing happens when you turn the switch off.

              ========= 

A complementary latch is an electronic switch made from two transistors.

One of the transistors has to be npn - and the other has to be pnp.

It's called a complementary latch because it's usually made from a pair of transistors that are said to compliment one another.

              =========  

Many npn transistors have pnp equivalents - and vice versa.

That is - aside from polarity - the transistors have very similar specifications.

They are said to complement one another.

For example - the BC557 is the complement of the BC547 - and vice versa.

In practice - you'll probably find that any old pair of npn/pnp transistors will work fine.

              =========  

In the diagram - Q1 is the pnp transistor and Q2 is the npn transistor. 

You can tell that Q1 is the pnp transistor from the way its symbol is drawn.

The small arrow is pointing towards the base.

Similarly - you can tell from its symbol that Q2 is the npn transistor.

It's drawn with the small arrow pointing away from the base. 

In each case - the arrow points towards an "n" terminal.

It indicates the direction of conventional current flow through the base junction.

              ========= 

To form the latch - you connect the collector of each transistor - to the base of the other transistor.

The collector of Q1 goes to the base of Q2 - and the collector of Q2 goes to the base of the Q1.

To switch the complimentary latch on - you simply supply one of the transistors with a momentary base current.

It doesn't matter which one.

When that transistor switches on - it uses the other transistor to latch itself on.

When the latch is switched on - Q2 connects the negative lead of the buzzer to ground - and the buzzer sounds.

              ========= 

We'll start by examining what happens when we supply base current to Q2 - the npn transistor.

In standby mode - the current flowing through R1 is diverted to ground by the normally-closed switch.

When the switch is opened - the current is no longer diverted to ground.

Instead - it flows through D1 and into the base of Q2.

This switches Q2 on.

              ========= 

When Q2 switches on - it connects the base of Q1 to ground.

This causes base-current to flow - through R2 - into the emitter-base junction of Q1.

So Q1 switches on.

And current through R2 and Q1 flows into the base of Q2.

              ========= 

Now - it doesn't matter if you re-close the switch and cut off the R1 current.

Q2 no longer needs the R1 current in order to remain switched on.

It has provided itself with an alternative and constant supply of base current - through R2 and Q1.

In other words - Q2 has used Q1 to latch itself on.

While it's on - Q2 connects the negative lead of the buzzer to ground - and the buzzer sounds. 

              ========= 

We can achieve a similar result by connecting the base of Q1 to ground momentarily.

When you close the normally-open switch - base current will flow into the emitter-base junction of Q1- through R2.

So Q1 will switch on.

Current through R2 and Q1 will flow into the base of Q2.

              ========= 

This will cause Q2 to switch on.

When Q2 switches on - it connects the base of Q1 to ground.

Now - Q1 no longer needs the momentary connection to ground in order to remain switched on.

It has provided itself with an alternative and constant connection to ground through Q2.

In other words - Q1 has used Q2 to latch itself on.

And - while Q2 remains switched on - the buzzer will sound.

              ========= 

To switch a complementary latch on - all you need do is forward bias one of the base junctions briefly. 

It doesn't matter which one. 

The result is the same.  

Once the latch starts to switch on - the process is self-sustaining. 

It's a sort of avalanche effect.

              ========= 

To turn the latch off - you simply need to switch off one of the transistors momentarily. 

Again - it doesn't matter which one.

A transistor will switch off if you remove the forward bias from its base-emitter junction.

One way to do this is to connect the base and emitter pins together.

              ========= 

Sw1 and Sw2 are momentary action push-to-make switches.

When Sw1 is pushed - it connects the base and emitter pins of Q1 together.

When Sw2 is pushed - it connects the base and emitter pins of Q2 together.

In both cases - the result is the same.

              ========= 

Because the base and emitter pins are connected together - they are both at the same potential.

So - the base junction is no longer forward biased.

Consequently - the base current ceases - and the transistor switches off.

              ========= 

If you switch Q1 off - it can no longer provides base-current to Q2. 

So Q2 will switch off - and no longer provide the path to ground for the Q1 base current.

If you switch Q2 off - it can no longer provides the path to ground for the Q1 base current. 

So Q1 will switch off - and no longer provide base current to Q2. 

Either way - both transistors are switched off - and the buzzer is silent.

              ========= 

The values of R1 and R2 were chosen so that a current of up to about 50mA could be drawn from the buzzer terminals.

This would allow a relay coil to be connected to the buzzer output.

The relay contacts could then be used to switch any type of device. 

              ========= 

R1 sets the standby current.

It also limits the base current reaching Q2 from the normally-closed trigger.

It doesn't have to switch Q2 fully on.

It just has to cause the Q2 collector voltage to fall by about 0v7.

That's enough to forward bias the base junction of Q1.

Then Q1 switches on - and it supplies the additional current required to turn Q2 fully on.

R2 limits the base current reaching Q2 from Q1. 

It also limits the Q1 base current.

              ========= 

D1 allows the R1 current to reach Q2 - but it prevents the normally-closed switch from connecting the base of Q2 to ground.

Without D1 the normally-closed switch would act like Sw2.

Q2 would turn off every time the trigger switch was re-closed.

In other words - the complementary latch would never latch.

              ========= 

Some small alarm type switches have a maximum current rating of 20mA or less.

Without R4 - the normally-open switch would have to carry the current drawn by the load.

This could be up to 50mA.

R4 limits to less than 10mA - the current through the normally-open switch.

              ========= 

We've seen that it only takes a momentary input pulse to operate the latch.

High frequency electrical interference could trip the alarm.

The wires going to the trigger switches act like a radio antenna.

They pick up high frequency electrical interference.

              ========= 

To prevent false alarms we need to make sure that - in standby mode - at least one of the transistors is switched firmly off.

We also need to extend the length of the impulse required to switch it on.

These two jobs are done by R3 and C1.

              ========= 

R3 connects the base of Q2 to ground.

It keeps the transistor switched off - until required.

It also keeps C1 discharged.

To the high frequency signals picked up by the switch wires - C1 represents a short circuit.

The capacitor shorts the signals to ground - before they can switch Q2 on.

              ========= 

R1, D1 and R3 form a potential divider.

If the value of R3 were too low - the base of Q2 wouldn't reach the 0v7 it requires to switch on.

The value of 27k ensures that Q2 is free to switch on.

              ========= 

C1 slows down the speed at which the transistor will switch on.

The trigger switch has to be open long enough for R1 to charge C1. 

The bigger the value of C1 - the longer the switch has to be open.

Since the transistors operate as a pair - slowing down one - slows down both.

              ========= 

This circuit was designed to illustrate the operation of the complementary latch.

It differs from a practical alarm circuit in a number of ways.

For example - the practical circuit is unlikely to include Sw1 and Sw2. 

The latch is designed to power-up in the off state. 

So it's much simpler to switch off the latch by using Sw3 to interrupt the power supply.

              ========= 

There's also the issue of battery life.

The standby current is less than 100uA. 

That is - the current flowing to ground through R1.

A standby current of 100uA is low - but we can do better than that.

Below you'll find a link to several - more practical - complementary latch circuits.

              =========
Practical Transistor Latch Circuits	The Rest of Ron's Circuits	Write To Ron	More Free-to-Use Circuits	Circuit Exchange International
An Introduction To The Complementary Latch
