2
\$\begingroup\$

Are these two transistors equivalent? I could not find a model for the BC238BP so I wonder if it is ok to simulate the circuit with BC238 to get accurate results.

I found some datasheets online for both of them but they do not seem to be official at all.

If they are equivalent why does the BC238BP have the BP suffix?

Below is my circuit:

enter image description here

Desired output is given below where X-axis is time(s) and Y-axis is amplitude (V)

enter image description here

\$\endgroup\$
5
  • \$\begingroup\$ Edit your question to show application, linear, switch and give measurable results desired. If you can. hFE can make a difference in most cases unless you use negative feedback for resistor ratio design control instead of without or depend on high input impedance. Almost all transistors have a 4:1 range in hFE except those that are binned then 2:1 range which affects price slightly. All transistors are 100% tested and therefore higher hFE tend to cost more but no binning is most common unless you get better stock from Rohm etc. \$\endgroup\$ Nov 5, 2020 at 21:37
  • \$\begingroup\$ As @Bipp... says a good designer makes it less sensitive to hFE by choice such as defining the I/O tolerances with parameters then using the best design. Ask if you don’t know \$\endgroup\$ Nov 5, 2020 at 21:44
  • \$\begingroup\$ Nabla, why don't you modify your question to show us a schematic you are thinking about? Perhaps we can help you understand better how to design specific values so that the BJT variations (and ambient temperature variations) don't significantly impact the design goals. \$\endgroup\$
    – jonk
    Nov 6, 2020 at 0:49
  • \$\begingroup\$ Edited and added circuit diagram. \$\endgroup\$
    – Nabla
    Nov 6, 2020 at 6:28
  • \$\begingroup\$ <now seeing the schematic> So your NPNs are just switching on/off. In this case as long as they do (switch on/off) this circuit should work. If that BC238BP in this circuit would not switch (assuming the opamps / comparators are working) then it is broken. In this circuit it doesn't matter much what NPN you use, almost any general purpose NPN will do the job so stop worrying about the NPN's specifications. They simply do not matter. If you have an 2N2222 or BC547 or 2N3904 lying around, try them and you'll find that all will work in this circuit. \$\endgroup\$ Nov 6, 2020 at 8:39

1 Answer 1

5
\$\begingroup\$

You're over-worrying about the NPN transistors' specifications.

A "bc238" or "bc238pb" are basically all BC238.

In 999 our of 1000 circuits that use a BC238 you can use any BC238 from any manufacturer with whatever suffix and you will not notice the difference.

As a circuit designer I dare to say that in that one out of 1000 circuits where you do notice the difference, that that one circuit is badly designed.

Look at this BC238 datasheet from Fairchild and look op what the \$h_{fe}\$ value (the current amplification factor) is of this BC238. These are the values:

BC238A: 120 - 220

BC238B: 180 - 460

BC238C: 380 - 800

Wow, those are huge ranges! How can we ever design accurate circuits with that?

Well, a properly designed circuit just assumes that \$h_{fe}\$ has the lowest value. The circuit should work with that so here that would be \$h_{fe}\$ = 120. When \$h_{fe}\$ is higher than that, less base current should flow and that should not be an issue.

This sounds harder to do that it is, designers have been making circuits that behave like this since the NPN transistor was invented. If you make the circuit that provides the base current such that only 10% of the current flows into the base (and 90% is "wasted") then you will have such a circuit. If the base current then doubles, the ratios become 20% vs 80 %. That should be no big deal.

My point: forget about the actual suffixes and focus on a robust circuit that will work with almost any NPN transistor.

As a designer the parameters I pay attention to when choosing an NPN transistor are:

  • minimum \$h_{fe}\$
  • maximum collector current \$I_C\$
  • maximum collector - emitter voltage \$V_{CE}\$
  • maximum power dissipation

For high frequency designs I will check the gain-Bandwidth product \$f_t\$ as well.

If those values are close enough to what I need, the transistor can do the job.

\$\endgroup\$
3
  • \$\begingroup\$ Generally I agree with this. almost any NPN transistor is somewhat of a stretch. Surely there are pre-amp sections that are very sensitive to hfe and won't just work with a general-purpose BJT. \$\endgroup\$
    – Reinderien
    Nov 5, 2020 at 21:10
  • \$\begingroup\$ Going well beyond suffixes, many (most?) circuits that specify a 2N3904 will work perfectly well if you use a 2N2222 instead. The 2N2222 is generally more expensive, but if you have a few dozen lying around. Going the other direction often works as well, but you have to be more careful (the 3904 has a lower limit collector current). \$\endgroup\$ Nov 5, 2020 at 21:21
  • 1
    \$\begingroup\$ @Reinderien pre-amp sections that are very sensitive to hfe I agree but then you would need to use NPNs which have an \$h_{fe}\$ that falls within a certain (small) range. You would need to "hand pick" these NPNs and/or buy pre-selected ones, all that is quite expensive. My preference would be to try to avoid that so that \$h_{fe}\$ doesn't become so critical. \$\endgroup\$ Nov 6, 2020 at 8:32

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.