some questions about hFE of transistors ...

[kypo]

Hello everyone,

First of all, i would like to thank everyone in this forum for making the life of electronics hobbyists so much easier! THANK YOU ALL !

I am having some problems in understanding how the hFE parameter of transistors really work... the eqn. is: hFE = Ic / Ib .. right?
Now the confusing point is: i've read in soo many places that to calculate the value of Ic, you simply multiply Ib by hFE, which mathematically seems logical from the above eqn. However, i found that in transistor datasheets they are giving specific hFE values for each specified Ic current !! That was a real surprise for me !! If hFE itself is dependent upon Ic, and Ic is what i want to calculate, then according to the above eqn. : Ic = Ib x hFE >>> TWO UNKNOWNS !?? because Ic is what we want to find at the first place, however, hFE depends on Ic ... it is really confusing :cry:

Another point, my digital multimeter have an "hFE" option. When use it for the transistor i have, it gives a SINGLE value !! then, what is the Ic of the measured hFE ?

Thank you all in advance,
-- kypo --

 

[Nigel Goodwin]

Hello everyone,

First of all, i would like to thank everyone in this forum for making the life of electronics hobbyists so much easier! THANK YOU ALL !

I am having some problems in understanding how the hFE parameter of transistors really work... the eqn. is: hFE = Ic / Ib .. right?
Now the confusing point is: i've read in soo many places that to calculate the value of Ic, you simply multiply Ib by hFE, which mathematically seems logical from the above eqn. However, i found that in transistor datasheets they are giving specific hFE values for each specified Ic current !! That was a real surprise for me !! If hFE itself is dependent upon Ic, and Ic is what i want to calculate, then according to the above eqn. : Ic = Ib x hFE >>> TWO UNKNOWNS !?? because Ic is what we want to find at the first place, however, hFE depends on Ic ... it is really confusing :cry:

Another point, my digital multimeter have an "hFE" option. When use it for the transistor i have, it gives a SINGLE value !! then, what is the Ic of the measured hFE ?

Life is complicated, isn't it? :lol:

Your meter will give the hFE at a specific collector current, and databooks will normally do the same as well - you need a graph of Ic against Ib if you want to find it at a precise point, and you would need to plot it for the individual sample that you are using, as it varies considerably between samples. Usually databooks will give a minimum value, at a specific current.

You should design circuits to make them independent of hFE variations, this is one of the reasons for negative feedback in designs.

BTW, it's often been common place in older commercial designs to have some resistor values marked as 'select on test', which means it's selected for the particular transistor fitted during manufacture. Replacing the transistor (if it fails) means changing the value of the resistor, usually to obtain a specific voltage - mostly on the collector of the transistor.

If you care to post a diagram of the circuit you are trying to calculate, perhaps we can help further?.

 

[kypo]

... Your meter will give the hFE at a specific collector current ...

Yes, it should be that way. However, what is the value of that specific current ? i guess the manual of my DMM should specify it, right?? well, guess what ... it didn't :lol:

If hFE varies with Ic, then what people usually say "to calculate Ic, multiply hFE by Ib" is wrong, isn't it ? If that is wrong, and i guess so, then what is the mathematical equation for Ic ?? Actually i don't have a specific circuit that i want to calculate the current for .. i just want to know how to do so in general.

Also, i've tried to make the Ic - Ib plot. What i did is that i've started gradually from a zero Ib upto about 10 uA, measuring Ic for each Ib. My design sets Ic(sat) to 10 mA (to prevent the transistor from oversaturation, as there is no hFE of 1000 ) .. my first Ib was 1uA, the corresponding Ic was 148uA, which means that hFE is 148. Then, i've tried the rule that says "to get Ic, mutiply Ib by hFE" so i did that for the next value of Ib .. i've set the Ib to 2uA multiplied it by 148 .. and, tada: Ic = 296uA and when i measured it, Ic was exactly 296uA ... A SHOCK (for me at least :lol !! why hFE didn't change?? is it possible that the hFE for Ic = 140uA is the same as for Ic=296uA ?? Well, experimentally, yes. Theoretically, i don't know! Maybe there is no change in hFE when Ic is in uA's scale ?! Because if you look at it, most datasheets that i've read indicate hFE values for Ic's in the mA range .. could that be a reason ?! I don't know .. help me out everyone :cry:

... You should design circuits to make them independent of hFE variations ...

Interesting!! How could i do that (using transistors only) ? :roll:

Thanks again Nigel..

 

[Nigel Goodwin]

Interesting!! How could i do that (using transistors only) ? :roll:

Here's one way, in the 'Bad' example you would have to select the resistor values for each and every transistor. The 'Good' example has negative feedback from collector to base, this will allow a wide spread of transistor gains without changing values.

However, you get this at the expense of lower gain, the 'Bad' example will have more gain.

 

[fat-tony]

Another thing to remember is that Ib * hFE = Ic is an approximation. You really don't want to depend on hFE, since it changes with current, temperature, etc.

 

[kypo]

yes, i know that the eqn. is an approximation. But it is a fair approximation i think ...

Hmmm .. so, let's reconsider it from the beginning. How would one design the transistor part of a certain circuit, considering the relations between Ic, Ib and Ie ?! I mean, if hFE was bad to consider, and i know it is. Then, what is the best technique to calculate and find out the relations between currents ?? in another way: what is the optimum technique to find out the gain of the transistor ?! i hope i made it clear .. :lol:

I know there are other techniques using other devices. However, i just want to know how various functions could be done using a transistor ..

 

[Nigel Goodwin]

yes, i know that the eqn. is an approximation. But it is a fair approximation i think ...

Hmmm .. so, let's reconsider it from the beginning. How would one design the transistor part of a certain circuit, considering the relations between Ic, Ib and Ie ?! I mean, if hFE was bad to consider, and i know it is. Then, what is the best technique to calculate and find out the relations between currents ?? in another way: what is the optimum technique to find out the gain of the transistor ?! i hope i made it clear .. :lol:

I know there are other techniques using other devices. However, i just want to know how various functions could be done using a transistor ..

Show us what sort of circuit you are considering, then perhaps we could make suggestions. It would be rather impractical to give advise on every possible circuit!.

 

[Russlk]

This graph of the 2N3904 may shed some light on the subject: