VERY wide range of hFE on 2N4401 by fairchild!! Is it a problem?

[Willen]

I have 5 pieces of 2N4401 genuine product from Fairchild (USA). I tested their hFE on multimeter and got VERY large range on same model, same manufracture date:-

-two pieces of 2N4401 has around 175
-two pieces of 2N4401 has around 255
-one piece of 2N4401 has around 230

To compare, I tested 10ths of 2N3904 and it has a range from 195 to 210. It's narrow range. 10ths of BC547 has 345 to 365, also narrow range.

Won't it cause little distortion while designing a certain circuit or while replacing to each other cause of this wide range of hFE on 2N4401? Is it acceptable?

 

[crutschow]

The hFE spec in the 2N4401 data sheet is from 100-300 at 150mA collector current so a 3:1 range of hFE is normal. This is less spread than what you measured, so there's nothing wrong with the transistors.

A good circuit will be designed so that the normal range of hFE has only a small affect on circuit performance. You should always design your circuits to accommodate the normal manufacturing spread of all the device parameters.

 

[kubeek]

The datasheet says Hfe should be between 100 and 300. Yours are actually in the center, so what is the problem?
You should allways design a circuit in such way that it will work with any combination of hfe that is within the specified range. Or you could select your transistors if you really want to.

 

[Willen]

Did you mean- if hFE is 100-300, then I have to design a circuit for 200?

 

[Willen]

very-wide-range-hf

Digital multimeter shows just hFE only. It shows 195 - 210 for 2N3904. How can I explain? I mean at how much 'V' and how much 'Ic'?

 

[MrAl]

Hi,

If the min spec is 100 then you design your circuit so that it can work with a transistor that has a Beta of 100 or less and still do what you want it to do. But you cant design a circuit that has a gain of 200 if the transistor only has a gain of 100 for example, it has to be less than 100. Typically you might design it with a gain of 10, but the type of designs are quite varied so the main point is that the circuit must work with a transistor with a gain of 100 or less, and it must operate correctly.

The Beta spec of the transistor is not an exact specification, and it can vary over temperature and operating point too.

What you also might want to show is what current level you made these measurements at. What collector current did you use when making the Beta measurements.

 

[Willen]

What collector current did you use when making the Beta measurements.

The multimeter has pin socket for transistor, so I just insert a transistor on it. So I don't know about V and Ic. So asking.

 

[crutschow]

The multimeter has pin socket for transistor, so I just insert a transistor on it. So I don't know about V and Ic. So asking.

It depends upon the multimeter design and should be stated in the multimeter manual. If it's a cheap multimeter with no manual then you would have to measure the collector voltage and current to determine the exact measurement point.

I looked at the manual for a Tenma Part #: 72-7765 multimeter and it measures hFE at ≈10μA base current and ≈1.5V collector-emitter voltage, so that may be typical for other multimeters also.

 

[crutschow]

Did you mean- if hFE is 100-300, then I have to design a circuit for 200?

No. You design the circuit so that it will work with the full range of hFE from 100-300. Typically the circuit uses negative feedback so that the range of hFE has only a small effect on the circuit performance. How small an effect is determined by the requirements of the circuit, which is defined by the system performance requirements. That is generally an engineering judgement.

For example, suppose the circuit gain must be within a value of 50±10%. Then you design the circuit so that the transistor hFE variation due to manufacturing tolerance and over the system operating temperature range stays within the value of 45-55. Generally this involves some iterative design and simulation/testing with high and low gain transistors until the circuit meets the requirements.

 

[unclejed613]

first of all, the trick is to design circuits where the individual characteristics of a device are minimized as much as possible. you will see that a lot of designs are made this way. it doesn't matter which extreme of the transistor beta range a particular device has, because the individual characteristics are swamped into insignificance by the design of the circuit.


second, if you look carefully at transistor data sheets, you usually find a curve that shows Hfe vs Ic. this means that the collector current when the measurement is done is very important. the Hfe range in the data sheet will probably have a notation showing what particular collector current used when the measurement was made.

 

[Willen]

Typically you might design it with a gain of 10, but the type of designs are quite varied so the main point is that the circuit must work with a transistor with a gain of 100 or less, and it must operate correctly.

And won't it affect if I used 10 hFE to design circuit but transistor gave 300 or more hFE?

 

[kubeek]

That really depends on how the circuit is designed and what is the purpose of the transistor.
But generally the hfe is similar to open loop gain of an op amp. You almost never use an opamp without feedback which sets the total gain of the circuit and diminishes the influence of the of the open loop gain.

 

[crutschow]

And won't it affect if I used 10 hFE to design circuit but transistor gave 300 or more hFE?

Of course it will. That's why you don't use a single value of hFE to design the circuit, you design for the full range of hFE. You test/simulate the design to insure that it meets your requirements under all conditions. As noted, you use local negative feedback in the circuit to minimize the effect of transistor variations.

 

[Willen]

As noted, you use local negative feedback in the circuit to minimize the effect of transistor variations.

Emitter resistor? Or collector to base resistor?

 

[unclejed613]

in an effort to "simplify" terminology, a lot of engineers are using similar terminology whether they're talking about op amps or transistors. for instance i've been seeing a growing number of transistor data sheets that use the term Gain Bandwidth Product to describe the Ft (transition frequency, i.e. the frequency at which the beta of a transistor drops to 1).

if you have a transistor with a nominal Hfe of 300, and being used in a circuit with a gain of 10, the wide variations in beta between randomly selected devices won't have very much effect on the operation of the circuit.

 

[crutschow]

Emitter resistor? Or collector to base resistor?

Either or both.

 

[audioguru]

An emitter resistor increases the input resistance of a transistor.
A collector to base resistor decreases the input resistance of a transistor and needs a signal source to the base that has some resistance.

 

[Willen]

An emitter resistor increases the input resistance of a transistor.
A collector to base resistor decreases the input resistance of a transistor and needs a signal source to the base that has some resistance.

Then Is 'collector to base R' known as Positive feedback?

 

[unclejed613]

no it's negative feedback because the signal fed back is 180 degrees out of phase with the input.

 

[audioguru]

no it's negative feedback because the signal fed back is 180 degrees out of phase with the input.

When the voltage at the base goes up then the voltage at the collector goes down then the negative feedback reduces the input resistance.