How Can I Predict Output Current in a Transistor Using Base Current?

[oneamp]

From what I've read about transistors, I thought that I could find the hfe value in the data sheet, then apply a current to base and have the emmitter-collector throughput at a level of base*hfe. It sounded simple, and I was going to have a few different resistors for controlling base current, selected by a switch.

Well, I was confused off the bat, when looking at the data sheet. There was no 'nominal hfe'. There was a max hfe, a minimum hfe, and they were very different values. So how can I do what I wanted to do... know the output current via the base current? It sounds like I'm missing a general principle.

Thanks

 

[analogdesign]

hfe (or beta) varies wildly. The manufacturer will typically guarantee a minimum hfe but it is VERY bad design style to make a circuit that depends on a particular value of hfe.

Also, while physically the BJT is driven by base current, in practice it is more useful to think about it as a voltage amplifier with Ic exponentially related to Vbe via the diode equation.

Your idea for biasing the transistors would be almost impossible to make work because your transistors would slam themselves into saturation. Read this informative link for information on the proper way to bias a discrete BJT.

http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/npncebias.html

 

[oneamp]

Thank you. So, we cannot reliably use a transistor to amplify a signal to a desired value? It's just always an on/off switch in operation?

 

[meBigGuy]

You can use a transistor to reliably amplify a value. One way is to use an emitter resistor. Assume for a moment that Vbe is constant and an NPN transistor. You have 10 ohms in the emitter( to ground) and 1000 ohms in the collector (to Vcc). If you apply a signal to the base it will appear across the 10 ohm resistor causing a current to flow. This current will mostly come through the 1000 ohm resistor so there will be 100 times the voltage drop across the 1000 ohm resistor as across the 10 ohm resistor. There you have inversion and voltage gain. This is extremely oversimplified.

If you want reliable current gain there are ways to do that also.

 

[analogdesign]

Thank you. So, we cannot reliably use a transistor to amplify a signal to a desired value? It's just always an on/off switch in operation?

Of course you can use a transistor to reliably use a transistor to amplify a signal. It's how I make my living. But you can't do it in the way you described.

MeBigGuy is right and gave a simple way to get voltage gain. The gain in the structure he suggested is the ratio between the collector or emitter resistors.

If you need current gain, you should use a current mirror. Look that up in wikipedia. If you need high current gain, you should look at a transimpedance amplifier using a op amp.

 

[oneamp]

Thanks

 

[sophiecentaur]

The point about using any amplifying device (linear and not just a switch) is that you should always use it with feedback, defined by the surrounding components. You end up with a well behaved amplifier at the expense of some loss of gain. But, hey, your average transistor, these days, has loads of gain to spare. It's not like in the old days when they had an Hfe of 10! (and I don't mean Factorial )

 

[psparky]

The point about using any amplifying device (linear and not just a switch) is that you should always use it with feedback, defined by the surrounding components. You end up with a well behaved amplifier at the expense of some loss of gain. But, hey, your average transistor, these days, has loads of gain to spare. It's not like in the old days when they had an Hfe of 10! (and I don't mean Factorial )

Some day I'm going to really learn about transistors.
Today is not that day.