0
\$\begingroup\$

enter image description here

In above screenshot V2 is increasing ramp voltage from 0 to 12V, which accordingly increases Ib. Transistor is an NPN type.

Ic(q1) is the collector current.

Vce = V1 is a constant 1V.

So it means Im keeping Vce constant and only increasing Ib.

Many tutorials say in amplification region Ib/Ic is constant and called beta i.e. hfe.

But the blue plot which is Ic/Ib is very different than what I expect. What is wrong?

edit: Here is after I decreased the base current with FakeMoustache's suggestions:

enter image description here

\$\endgroup\$
16
  • \$\begingroup\$ Congratulations, you just figured out why the \$h_{fe}\$ value is meaningless for most situations. Besides that the value varies greatly for a specific product. Have a look at the datasheet for the transistor you are using there and the different values and conditions where that value was measured. \$\endgroup\$
    – PlasmaHH
    Commented Dec 16, 2015 at 15:10
  • \$\begingroup\$ Some remarks: using the "DC sweep" simulation is better suited for this kind of thing. Why force Vce at 1 V while driving the transistor "hard" ? Note that you get ridiculously high Ic currents. Your base resistor has a very low value, why ? In another question you said you use a PWM signal, you use a ramp here, why ? \$\endgroup\$
    – Bimpelrekkie
    Commented Dec 16, 2015 at 15:11
  • \$\begingroup\$ But many tutorials take this as a fixed value \$\endgroup\$
    – user16307
    Commented Dec 16, 2015 at 15:11
  • \$\begingroup\$ I think you're confusing small-signal AC analysis with this [large-signal] DC one. \$\endgroup\$
    – got trolled too much this week
    Commented Dec 16, 2015 at 15:12
  • 2
    \$\begingroup\$ That is because you're "mistreating" this NPN, you force a very large Ib but Vce is not enough to allow the IC that would need to flow when Hfe is large. Try with Ib as a current source from 1uA to 1mA and see what Ic does. Oh, use a DC-sweep this time, not a tran simulation ! See: denethor.wlu.ca/ltspice/#dsweep \$\endgroup\$
    – Bimpelrekkie
    Commented Dec 16, 2015 at 15:20

1 Answer 1

Reset to default
3
\$\begingroup\$

Over a substantial range of Ib (corresponding to V2=1.5 to 10V or Ib around 0.1 to 1.0 mA) you show that hFE for this transistor model varies only between 180 and 205. That is close enough to constant for many applications, especially in the context of all the other conditions that cause hFE to vary.

So some introductory texts may simplify their treatment of a transistor by pretending it is actually constant, in order to teach other things (aka small-signal analysis). A good text will return to this issue later on and explore the actual behaviour in more depth.

The linked article (from comments above) makes this statement without any such qualification. However it is on a webpage which also contains this utter howler. In my estimation, such a page is actually worth LESS than you paid to read it!

\$\endgroup\$
2
  • \$\begingroup\$ I agree with Brian Drummon doubts. The article under discussion contains even more errors. \$\endgroup\$
    – LvW
    Commented Dec 16, 2015 at 19:49
  • \$\begingroup\$ Actually, what I said in my previous comment (now deleted) was incorrect. In this case the degradation in beta is due to high-level injection which for this particular model kicks in early. Speaking of which... LTspice ships with a lot of dubious transistor models, but for this one its Gummel plot looks somewhat plausible. \$\endgroup\$
    – got trolled too much this week
    Commented Dec 17, 2015 at 0:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.