# Selecting a transistor - Reading transistor datasheets

I am trying to select a transistor part and figure the actual current calculations in my circuit.

I have 30mA of collector (load current). This is the transistor I am choosing for my design : https://www.onsemi.com/pub/Collateral/MMBTA05LT1-D.PDF

MMBTA06

So, Once I have figured out the collector current through my transistor, I need to calculate my required base current. But while calculating the required base current, how will I choose the Hfe value? In the datasheet it is given as 100 (For Ic=100mA and Vce=1.0V)

The graph in fig 5, gives Ic vs Hfe, but the Vce is 1.0V. And in fig 6, Ic vs Vce(sat), the gain is fixed as 10. How will I take values from this graph when the conditions are different from my application.

My actual Ic is just 30mA. How will I fix my gain now?

I am going to operate the transistor in linear region. It would really help me if one could tell me which graphs we need to consider in the transistor datasheet for my current calculations.

And I am calculating Min, typ and max current values for my circuit. So, I am considering -40,+25 and +85degC temperatures. Can someone help me with the calculations.

Thanks.

• What circuit are you using the transistor in? If you edit your question, you can use the built in schematic tool to add a schematic to your question. Jan 16, 2019 at 2:08
• If you want to go it alone, look at fig. 5 in the datasheet (and realize that you won't always operate your circuit at exactly the same temperature). Jan 16, 2019 at 2:10
• You say "fix" your gain, as in "hold it steady" -- Figure 5 is what you want to use, but keep in mind that the $H_{FE}$ figure is a guaranteed minimum. Transistor characteristics vary all over, from part to part and from temperature to temperature. You need to design your circuit to work regardless. Jan 16, 2019 at 2:14
• You must define acceptable Zin , Zout, Vo_pp, Vcc, gain, BW, & THD then we'll see what works with 1 transistor Jan 16, 2019 at 4:07
• Gain can be based on impedance or R ratios include Rbe but you want to minimize sensitivity to hFE so define the above then it can be expressed Jan 16, 2019 at 4:46

The saturation curves are for switching applications, so you can ignore those numbers if you are indeed operating in the linear region, say with Vce >= 1.0V. If what you wrote is incorrect, and it's not in the linear region, then all the below is garbage.

hFE is stated to be a minimum of 100 at 10mA and at 100mA (and 25°C). Those are firm guarantees. There is no maximum hFE given. This transistor won't be the best to use if your circuit is heavily dependent on the value of hFE- it's more specified for switching applications.

We can, however, make some guesses. From the typical curves (Fig 5), the gain at 30mA in the linear region is between 90 and 270 or so, depending on temperature, with it typically at about 210 at 25°C. Since the gain can be as low as 100 at 25°C, it may be reasonable to guess that the gain at -55°C can be as low as (100/210)* 90 = 42. Similarly, we might guess (by risky extrapolation) it could be as high as perhaps 450 at 150°C. We might want to allow for a bit more than that on both sides, since we're extrapolating so maybe 40 to 500, at 30mA collector current, over the full temperature range of -55 to 150°C. You can judge how much safety margin to add on those numbers.

With a 10:1 or more range in hFE you'll have to use something like emitter degeneration to properly bias the transistor in the linear range. Below 1V Vce the hFE will drop as the transistor saturates.

Note you can buy 'binned' transistors with a better controlled and characterized hFE but even with that the variation between parts may still be 2:1 before temperature is taken into account (eg. 120~240 hFE). 45mA is ideal for maximum bandwidth or GBW product but poor for large signal linearity.

So it depends on your specs for THD or peak +/- asymmetry at max swing.

Generally you want to keep Vce>2V for high linear in H bias at high currents or > 1V at low currents as hFE drops rapidly to 10% down near Vce=Vce(sat)

So decide on your output power or swing, voltage gain, linearity or asymmetry or THD before you choose a Q point. Then decide your input and output impedance.

Fundamentally you want to to minimize the dynamic range in current to 1 decade as 2 decades of current swing will add ~>10% THD unless you use negative feedback bias, which is suggested if you need this.

A single bipolar transistor, given a resistive load, has maximum gain of VDD / 0.026 or 39 * VDD; 10volts allows a maximum gain of 390.

Given the need for bias stabilization, and for large voltage swings, you should plan on about 1/3 of that, thus expect a gain of 10x per volt of VDD. Why? simulate this circuit – Schematic created using CircuitLab