# Voltage divider types question

I have been studying different schematics ive found online to try to get a better understandment on the subject. Recently I came a voltage divider across that consisted of a transistor and a resistor, it looked something like this. I am only familiar with a voltage divider that consists of two resistors where you can apply the voltage divider formula to know the output voltage. I'm really curious to know whats's the difference in functionality between these two options? Why would you use one type and not the other one? What effect does the transistor have?

Thanks!

• Chalk and cheese mainly. – Andy aka Dec 8 '16 at 12:49

Basically, the transistor here makes a regulator out of your usual voltage divider.

The problem with voltage dividers is that, if you draw current from them, their voltage output drops down (because the load will act as if there was another resistor in parallel with the bottom resistor of the divider). And their output voltage value will therefore largely depend on the load current, which can fluctuate.

Now, with the transistor here, the voltage at the output will be about the voltage at the transistor base minus VBE (about 0.6V). Moreover, the current flowing through the base (so the current flowing out of the divider) will be much less (beta ratio) than the current flowing through the load. Which means that the current drawn from the divider will be much less dependent on the current through the load, so the voltage at the base will be more or less constant. So, overall, the voltage output will be much more stable.

It still doesn't make a very good regulator, however. There will still be some voltage drop depending on the load current, and, if the input voltage changes, the output will change too. But it can be sufficient in some applications.

The first one is an emitter follower circuit with the benefit that it can keep the output voltage constant over a wide range of current drawn at the output.

In contrast, the simple resistor divider will give a voltage that starts to drop as soon as any current is drawn from the output.

The transistor circuit can be analyzed as a voltage divider followed by an emitter follower buffer. The functionality is similar between the two circuits.

Performance-wise there are important differences.

1. The resistors-only circuit (type B) can be almost perfect as far as initial tolerance and changes with temperature go, as close as the two resistors are matched in resistance ratio and temperature coefficient, which could be within ppm or better. And it's not limited in voltage- it works equally well for microvolts or volts, right up to either rail. For example if I have a 5.00V reference and want a 10mV voltage I could use a 10 ohm precision resistor and a 4.99K ohm precision resistor. Output resistance would be about 10 ohms.

2. The output resistance of the buffered (type A) circuit is lower than the B circuit in many cases, without the huge power consumption of a very low resistance divider. In most cases, it's just a bit worse than the divider output resistance divided by the hFE of the transistor, which is quite good. For example, two 10K resistors from 5V to give you 2.5V will consume 0.5mA and have a 5K output resistance. Adding the transistor gives you similar consumption (not counting any emitter resistor) but output resistance that will be more like 50 ohms with an hFE of >>100. You could get something similar with two 100 ohm resistors but it would draw 100x the current.

Note that the transistor emitter can only source current, it cannot sink it, so the net load must result in a current flowing out of the emitter. Resistors don't care, and will work equally well (or badly) regardless of the current direction.

3. There is a diode (Vbe) drop between the divider voltage and the emitter voltage- and that will change with temperature. So with the two 10K resistors in the above circuit, the output voltage will be more like 1.8V (you can sort-of fix this by changing the resistor ratio) and will increase at +2mV/°C with temperature (which can't be fixed easily). So the 10mV example is not going to be a practical application for the A circuit.

Adding an op-amp voltage-follower buffer may be better than the simple emitter follower transistor, but there are other complicating factors- the op-amp may oscillate with a capacitive load, the op-amp will consume some power, it may not be able to source as much current as a transistor, it may not have the output or input voltage range to work reliably, etc.