hello

I am no transistor expert and i do have very limited experience with them. I am trying to understand how the base current is sized to a transistor. After doing some searching on google i understood that one have to look at the DC current gain (BETA) in order to calculate the (lb). But on the specification sheets like the 2n3055 they mention values like these, DC current gain= 20-70. I am not getting the picture here, should one chose a value between 20-70.

Would be thank full if some one could explain.

 

DC current gain is not a fixed value. It can vary from device to device. It can also change with Ic. A value of 20-70 gives you a range of possible values.

 

hello

I am no transistor expert and i do have very limited experience with them. I am trying to understand how the base current is sized to a transistor. After doing some searching on google i understood that one have to look at the DC current gain (BETA) in order to calculate the (lb). But on the specification sheets like the 2n3055 they mention values like these, DC current gain= 20-70. I am not getting the picture here, should one chose a value between 20-70.

Would be thank full if some one could explain.

Beta is a widely varying specification for a transistor. It depends on lots of things like process, temperature, Ic (collector current), and so on. Expect that each transistor will have a different beta within those values stated on the spec sheet, under the conditions expressed.

We use transistors for at least two things: switches and amplifiers.

In switches, we want to drive the base hard with respect to the expected collector current and beta. This way, under all circumstances, the transistor is guaranteed to be always on.

In amps, we use circuit topologies and negative feedback to limit the design's sensitivity to variations in beta.

Look up "common emitter amplifier" for an example.

 

Any parameter that is described by a range should be viewed as a random variable whose value is likely to be within the range. There is however a small non-zero probability that the parameter will be outside the range for a single particular unit chosen at random. It should be a wakeup call for engineers and circuit designers that components are not 100% tested

 

So what is the logic behind this, i mean how do you chose a value between the minimum and maximum ?. I have even seen ranges like 90-1000 .

 

So what is the logic behind this, i mean how do you chose a value between the minimum and maximum ?. I have even seen ranges like 90-1000 .

I don't know that logic has anything at all to do with this. It is inherent in the semiconductor fabrication process. What you are getting is the best available information that statistical sampling and partial testing can provide. The upshot of all this for a design engineer is to never, under any circumstances, design a circuit that depends on beta, or any other parameter that varies from device to device.

 

I don't know that logic has anything at all to do with this. It is inherent in the semiconductor fabrication process. What you are getting is the best available information that statistical sampling and partial testing can provide. The upshot of all this for a design engineer is to never, under any circumstances, design a circuit that depends on beta, or any other parameter that varies from device to device.

Sorry.. but i am not getting the point. There has to be a method/way to determine the required base current, right ?.

 

Required current to do what specifically? You first need to know what you want that transistor to do, and then you decide the required base current.

 

a simple example would be a switch

 

So what is the logic behind this, i mean how do you chose a value between the minimum and maximum ?. I have even seen ranges like 90-1000 .

You choose a value you can be guaranteed to expect in your circuit. If the gain can vary between 90 and 1000 you can be sure to get at least 90, right? So without any other information you pick that number. That may seem low but my circuits professor would pick 30, as when he was "in the field" making integrated circuits when the transistors on the wafer had a gain of 30 (or more) the chip was deemed acceptable. Less then 30 it would be tossed.

Depending on the application you may even choose a lower number. One example is a saturated switch where the transistor is on very hard. A typical number there is just 10.

 

Required current to do what specifically? You first need to know what you want that transistor to do, and then you decide the required base current.

You choose a value you can be guaranteed to expect in your circuit. If the gain can vary between 90 and 1000 you can be sure to get at least 90, right? So without any other information you pick that number. That may seem low but my circuits professor would pick 30, as when he was "in the field" making integrated circuits when the transistors on the wafer had a gain of 30 (or more) the chip was deemed acceptable. Less then 30 it would be tossed.

Depending on the application you may even choose a lower number. One example is a saturated switch where the transistor is on very hard. A typical number there is just 10.

So its simply if you have an LED that runs on 20mA you just (20/90 (minimum beta value)) = 333microamp

Thanks for making this clear to me.

 

Actually, no.

When a BJT is used as a switch you want to turn on the transistor hard into saturation.
The rule of thumb is to use a beta of 10.

 

The rule of thumb is to use a beta of 10.

Does this apply for all BJT when used as a switch ???

 

Yes. What it really means is that you really don't care what transistor you use as long as the beta is greater than 10 and it can handle the voltage, current and power dissipation.

So what if the transistor's actual beta is 50 and you had designed the base current @2mA?
You are not going to get Ic = 100mA. Why? Because you had set a limit on Ic to 20mA using a current limiting resistor in the collector circuit.

 

Yes. What it really means is that you really don't care what transistor you use as long as the beta is greater than 10 and it can handle the voltage, current and power dissipation.

with respect to the maximum Ib

 

I struggled with this Beta problem allot too when starting out, looking at all these transistor equations that include Beta- then actually building circuits based on the assumption that Beta is a hard parameter got me very frustrated- things did not add up.

Beta is a parameter that the transistor manufacturing processes cannot control accurately, people have to design circuits that always work, even if Beta varies all over the place.
That's how you need to look at it, given the worst case Beta, your circuit still needs to work- design accordingly.

 

...given the worst case Beta, your circuit still needs to work- design accordingly.

...and your circuit must also work if you happen to grab a transistor with the highest Beta.

 

Beta is a parameter that the transistor manufacturing processes cannot control accurately, people have to design circuits that always work, even if Beta varies all over the place.
That's how you need to look at it, given the worst case Beta, your circuit still needs to work- design accordingly.

This sounds like trail and error

 

Why? If your circuit requires certain minimal beta, then you need to calculate with the minimal value.
If yor circuit requires certain maximal beta, then design with the maximal value in mind.
In other words, you need to design the circuit in such way that it will work with ANY beta in that range. And all that is at 25°C, but you need to consider the operating temperature range also.

 

This sounds like trail and error

Not really (although some do design with that method .) You just need to design the circuit so that it will work properly with both the minimum and maximum Beta values.