Transistor output resistance and thermal voltage

[tindel]

A couple questions:

I'm in the process of making some small amplifiers and using bjt's in the small signal realm. I have used bjt's as switches for quite a while, so I am quite familiar with their basic operation. I was reviewing small signal analysis trying to refresh my memory about how to do the analysis when I came across a couple variables that I'm not sure how to incorporate them into my analysis in a practical sense.

First, thermal voltage of a pn (diode) junction - V_T - which my book describes as a 'constant', which is a function of temperature. V_T is defined as

V_T= k*T / q

where
k is Boltzmann's constant = 1.38 x 10^-23 joules/kelvin,
T= temperature kelvin,
q= magnitude of electronic charge = 1.6 x 10^-19 coulomb

Is this true for all pn junctions - or does this change with doping, diode types, etc?


Second - small signal output resistance, r_o - my book does an awful job explaining how to arrive at this value, practically speaking. They give the following equation for the output resistance r_o = (V_A + V_CE) / I_C. I have never seen r_o or V_A in a datasheet. I'm not sure what V_A even is. I do understand that r_o is a function of collector current due to v_ce, so I understand why it's used, I just don't understand when I have a circuit I'm analyzing, how to come up with a value of r_o!

True to form - my old electronics book tells me how to do all the calculations, but doesn't explain how to arrive at the value in a practical sense. A quick google search also turned up nothing.

 

[uart]

The term kT/q is not really a parameter of the diode (or transistor) itself, but is actually just the thermal energy of the carriers in the semiconductor, expressed in electron volts. The electrical properties of diodes and transistors do depend upon this quantity, but also on many other parameters of the device. So in summary, yes this quantity is the same for all semiconductors at a given temperature, but the characteristics of those different devices can still be different due to other factors like physical dimensions, doping levels and the intrinsic semiconductor material itself.

 

[yungman]

A couple questions:

I'm in the process of making some small amplifiers and using bjt's in the small signal realm. I have used bjt's as switches for quite a while, so I am quite familiar with their basic operation. I was reviewing small signal analysis trying to refresh my memory about how to do the analysis when I came across a couple variables that I'm not sure how to incorporate them into my analysis in a practical sense.

First, thermal voltage of a pn (diode) junction - V_T - which my book describes as a 'constant', which is a function of temperature. V_T is defined as

V_T= k*T / q

where
k is Boltzmann's constant = 1.38 x 10^-23 joules/kelvin,
T= temperature kelvin,
q= magnitude of electronic charge = 1.6 x 10^-19 coulomb

Is this true for all pn junctions - or does this change with doping, diode types, etc?

I cannot answer about the doping, but yes V_T=\frac {kT}{q}\; holds for all transistor and diodes.

Second - small signal output resistance, r_o - my book does an awful job explaining how to arrive at this value, practically speaking. They give the following equation for the output resistance r_o = (V_A + V_CE) / I_C. I have never seen r_o or V_A in a datasheet. I'm not sure what V_A even is. I do understand that r_o is a function of collector current due to v_ce, so I understand why it's used, I just don't understand when I have a circuit I'm analyzing, how to come up with a value of r_o!

True to form - my old electronics book tells me how to do all the calculations, but doesn't explain how to arrive at the value in a practical sense. A quick google search also turned up nothing.

A lot of transistor data sheets provide collector curves for different Ib. To find the early voltage, extend the straight portion of the curve ( the saturation region) to the left side where Vce going to negative. Keep extending all and they all will meed at Ic = 0. Read off the Vce and that's the early voltage.

r_0\; is defined as change of collector current to change of Vce. That is the slope of the collector curve when the transistor is in saturation region ( the straight part of the curve). As you can see, the slope is different for every Ib. Therefore r_0\; is collector current dependent. Normally r_0\; is not that important as you have collector resistor R_L\; which is usually much lower. The total load resistance is r_0\; parallel with R_L\; which essentially just R_L\; when it is a lot lower than the output resistance. It is important when you use it as a current source where you need very high impedance at the output. In IC, most circuit use active load, then it is more important as it affect the gain of the transistor if this is the load on the collector.

 

[uart]

o[/SUB] - my book does an awful job explaining how to arrive at this value, practically speaking. They give the following equation for the output resistance r_o = (V_A + V_CE) / I_C. I have never seen r_o or V_A in a datasheet. I'm not sure what V_A even is. I do understand that r_o is a function of collector current due to v_ce, so I understand why it's used, I just don't understand when I have a circuit I'm analyzing, how to come up with a value of r_o!

Apart from using the "Early voltage" (as explained by yungman above) many datasheets will specify the common emitter r_o indirectly via the parameter h_{oe}. This parameter is in fact the reciprocal of r_o.

Remember that in an actual circuit that the (external) collector resistor will be in parallel with r_o in the small signal model. Very often this external R_C is much lower than r_o and so dominates the parallel combination (meaning that you can often ignore r_o without introducing too much error in the analysis).

 

[tindel]

Thanks uart and yungman - The data you gave on both the early and thermal voltages is very useful. I understand that r_o is very large, and with small output resistances, relative to r_o I understand that it is not much of a concern unless you have large output resistances.