EMRFD Message Archive 817

Message Date From Subject
817 2007-06-02 09:03:30 KF4YYD New with Questions
Hello all,

New to group and wanted to say hi before I ask my question.

I'm busy reading chapter two of EMRFD, in particular the section on
biasing the bipolar transistor; I am also reading the books
Principals of Electronic Devices and Circuits by LaLond and Ross,
this one is more technician orientated, and W1FB's Design Notebook.
Of the three texts the first two always seem to use the tranistors
Beta when calculating values. So if you are trying to evaluate a
circuit what value of beta do you use?

For instance, I'm starting out with a basic audio amp, such as one
described in W1FB's design notebook. I can calulate the the basic DC
voltages I should expect to find, but if I want to refine this answer
I need to use beta. If my circuit uses a 3904 NPN transistor, I have
a beta spread of 100 to 300; since I have no way of knowing what the
value may be what should I use?? Or if I am trying to troubleshoot a
circuit what value would you use?.

I starting to understand the different biasing schemes, base bias,
negative feedback biasing, and voltage divider biasing; and
understand that the emitter resistor helps to limit the effect that
the tranisitor beta has on the circuit but just don't understand what
values to use.

Tom kf4yyd
818 2007-06-02 11:26:02 Thomas S Knutsen Re: New with Questions
Hello Tom and welcome to the group.

At page 2.9 in EMRFD beta is explained. I can also recomend
Introduction to RF design from ARRL, this book has the equation for beta.

An 2N3904 has an low frequency beta around 100 and Beta cutoff around 3MHz.
This is different for every transistor type, and has to be calculated.

73 de Thomas

KF4YYD wrote:
>
> Hello all,
>
> New to group and wanted to say hi before I ask my question.
>
> I'm busy reading chapter two of EMRFD, in particular the section on
> biasing the bipolar transistor; I am also reading the books
> Principals of Electronic Devices and Circuits by LaLond and Ross,
> this one is more technician orientated, and W1FB's Design Notebook.
> Of the three texts the first two always seem to use the tranistors
> Beta when calculating values. So if you are trying to evaluate a
> circuit what value of beta do you use?
>
> For instance, I'm starting out with a basic audio amp, such as one
> described in W1FB's design notebook. I can calulate the the basic DC
> voltages I should expect to find, but if I want to refine this answer
> I need to use beta. If my circuit uses a 3904 NPN transistor, I have
> a beta spread of 100 to 300; since I have no way of knowing what the
> value may be what should I use?? Or if I am trying to troubleshoot a
> circuit what value would you use?.
>















>
>
> Tom kf4yyd
>
--
______________________________________________________
Best Regards
Thomas S. Knutsen
LA3PNA
Mail: Thomas@la3pna.net
MSN: snknutse@online.no
Cellphone: +47 99244910
820 2007-06-02 12:54:38 Steven S. Coles Re: New with Questions
Welcome Tom,

In "typical value design" you pick a value near the middle to use in
the design. In "worst-case design" you demonstrate that parts with
extreme values (100 and 300 for the 2N3904) will produce results in
the acceptable range. (Usually worst case also takes temperature into
account.) In "specific part design" you measure the hfe (beta current
gain) of the particular transistor you will use and design to that value.

73,

Steven, KD7YTE
>
> KF4YYD wrote:
> >
> > Hello all,
> >
> > New to group and wanted to say hi before I ask my question.
> >
> > I'm busy reading chapter two of EMRFD, in particular the section on
> > biasing the bipolar transistor; I am also reading the books
> > Principals of Electronic Devices and Circuits by LaLond and Ross,
> > this one is more technician orientated, and W1FB's Design Notebook.
> > Of the three texts the first two always seem to use the tranistors
> > Beta when calculating values. So if you are trying to evaluate a
> > circuit what value of beta do you use?
> >
> > For instance, I'm starting out with a basic audio amp, such as one
> > described in W1FB's design notebook. I can calulate the the basic DC
> > voltages I should expect to find, but if I want to refine this answer
> > I need to use beta. If my circuit uses a 3904 NPN transistor, I have
> > a beta spread of 100 to 300; since I have no way of knowing what the
> > value may be what should I use?? Or if I am trying to troubleshoot a
> > circuit what value would you use?.
> >
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> >
> >
> > Tom kf4yyd
> >
> --
> ______________________________________________________
> Best Regards
> Thomas S. Knutsen
> LA3PNA
> Mail: Thomas@...
> MSN: snknutse@...
> Cellphone: +47 99244910
>
821 2007-06-02 18:12:25 Thomas Butchers Re: New with Questions

Steven & Thomas,

 

Thanks for the reply.

 

Thomas I do have introduction to RF design but the math seems to be a bit over my head at the moment. I have had the book since around 1995 and just recently broke it out. Unfortunately, it has a 3.5” disk and I have no way of reading it, I need to find a way to get it put on CD.

 

Steven, that makes since, I guess somehow I kinda knew that would be the correct answer I just wasn’t sure. Just make three calculations and see how it goes.

 

Thanks,

 

Tom

 

  

 


822 2007-06-02 18:53:03 Wes Hayward Re: New with Questions
Hi Tom, et al,

Yes, IRFD is a more mathematical book than most written for the radio
amateur. There are enough engineers and scientists within the ham
ranks that it is still useful. We have tried to be careful about
telling the buyer that it is a different, more analytic text.

The latest version of IRFD no longer has software distributed with
the book. Rather, the League decided they would keep the price low
and put the software on their web site. So you can get the
software, including MicroSmith, by going to the ARRL site and then
entering IRFD. That should get you to a download, including
manuals. Remember that this is software in support of the book and
not stand alone stuff. And it is DOS based. (Remember that?) It
works just fine from a DOS Window in MS Windows up through XP. There
are problems in getting a printout from XP. I have no idea what the
programs will do with Vista!

Now regarding the beta question: It is not as bad as you might
guess. If you go to the amplifier discussion in Ch 2 of emrfd, you
will see that we can do gain calculations without using beta at
all. This is because we can often regard the bipolar transistor as
a voltage driven device with a transconductance of Ie/26 where Ie is
the DC emitter current in mA. This is not always true. However,
the better circuit designs are often those that are not beta
dependent. If you are ever around engineers who are designing
integrated circuits, you will note that beta rarely come up in the
discussions.

Have fun.

73, Wes
w7zoi
823 2007-06-03 05:54:16 David Stone Re: New with Questions
Dear Tom

I think the answers you have received whilst OK don't really address your problem.

If I design a real circuit for production, the last thing I want is ANY dependence on the transistor beta, or anything else as far as is possible! So real circuits use several techniques to avoid device dependence. The first is negative feed back. A circuit block, say an amplifier is needed with a gain of 10. I choose a circuit and device which can provide a much higher gain, then reduce the gain to the required value with NFB. If my device were ideal, this would mean an undecoupled emitter resistor with 1/10 the value of the collector load (including the next stage input impedence). The classic case of this technique is the op-amp, open loop gain as close to infinite as possible, all other characteristics ideal as possible! The effect of the NFB is to almost remove the device beta from the gain of the circuit. It also increases the input impedence and reduces the output impedence, all desirable characteristics,  reduces the distortion or increases the linearity of the stage.

Real transistor parameters begin to bite the designer where the available devices are not ideal, such as at high frequencies where the available gain may be low, or the current gain available at large collector current may be poor, or the feedback capacitance means that negative feedback is already there at too large an amount! Then we use clever circuits to overcome the problems, like cascode connection or current mirrors or darlington devices or similar tricks.

Spend some time understanding NFB, and study some real circuits, (not too RF) then you will see what is done and why.

Ask again if you need more help

Regards
David GI8FNR

----- Original Message ----
830 2007-06-04 21:19:11 KF4YYD Re: New with Questions
Hi David,

Thanks for the reply; it is starting to make a little more sense to me
now.

I will try and follow your advice and see if I can't start working on
understanding the negative feedback type of circuit and see if i can
make sense of it. As far as studying real circuits I guess I'll have to
see what I can find. Most of what I have available to me at present are
textbooks and I'm sure what they provide are perfect examples instead
of real world ones. Then again I have EMRFD and I am sure they are
practical.

Tanks again, Tom kf4yyd


> Spend some time understanding NFB, and study some real circuits, (not
too RF) then you will see what is done and why.
>
> Ask again if you need more help
>
> Regards
> David GI8FNR