I know this question has been raised previously but as a newcomer to electronics there is still a bit missing for me. As I understand it the small base current allows a larger current to flow through the transistor. So if we need the base current to flow to allow the transistor to operate as a switch, what turns the base current on? If that is another switch, what is the point of the transistor as a switch in the first place, why not just use a mechanical switch?

 

One common case would be switching a large load with a DIO (Digital Input/Output) pin of a microcontroller or even discrete logic chips. The pin can handle very limited current so the transistor provides a way to control practical loads.

All the other cases are similar. An existing switch can't handle the load to be switched, or, we want to translate the output of a device, for example an LDR (Light Dependent Resistor) into a switching action for a load. In the latter, the device output varies but the goal isn't necessarily proportional response rather it is threshold switching.

 

If that is another switch, what is the point of the transistor as a switch in the first place, why not just use a mechanical switch?

If that were true, relays would not need to exist!

Another point - transistors can switch more cleanly. Switch contacts bounce, which causes arcing, which damages the contacts and causes interference.
A triac can switch a mains load exactly at the point the mains crosses zero, again avoiding interference.
Switches wear out - transistors don't, but, on the other hand, they can fail more easily than switches if overloaded.
Transistors are less efficient - they get warm while passing current.
Transistors can switch a load on and off a million times a second - switches can't.
A good designer knows how best to switch a load - transistor, triac, relay or switch.

 

One common case would be switching a large load with a DIO (Digital Input/Output) pin of a microcontroller or even discrete logic chips. The pin can handle very limited current so the transistor provides a way to control practical loads.

All the other cases are similar. An existing switch can't handle the load to be switched, or, we want to translate the output of a device, for example an LDR (Light Dependent Resistor) into a switching action for a load. In the latter, the device output varies but the goal isn't necessarily proportional response rather it is threshold switching.

Hi Yaakov - thank you - this makes it clear and fills in the bit I was missing. Regards Claude

 

what is the point of the transistor as a switch in the first place, why not just use a mechanical switch?

2 good reasons, no contacts to wear out and switches much faster than an electromechanical relay.

 

The transistor is switched by an electronic signal. The switch is switched by a mechanical motion. Completely different uses, though sometimes overlapping.

Bob

 

Speed control of a motor, or brightness control of a lamp, can also be done by pulse width modulation with a logic chip or CPU, for example.

PWM = Pulse Width Modulation

To put it simply, at a given frequency the logic could switch on at the beginning of the cycle, but after x percent of the cycle time, it turns off until the next cycle. You'll get approximately x percent of the speed or the brightness, within certain limits (i.e. below 20 percent the motor may not turn, there may be no visible light from the lamp, or whatever)