Maker Pro
Maker Pro

amplifier stability and compensation

That's great rich,
I'm fairly ok with the lesson plan, but i'm short on practical
aids and examples. I could just scrible up a few circuits on pspice but
i'd love to get a look at how other people have attempted it and then
take what i like about their views, and ideas of my own, and meet
somewhere in the middle.

For the lesson i'm going to stick to a simple style;
Introduce
Explain
Demonstrate
Get the students to imitate
Get the students to practice

Would anybody do anything differently?
 
P

Pooh Bear

Jan 1, 1970
0
Ok, to clarify;
I know the amplifier stability criterion, all i'd like to see is
what labs you think would be useful in teaching the subject.
The project concerns the METHOD of instruction of this topic, so
any links to university sites or completed projects would be spiffing.

The best method IME of instruction of pretty much anything is to cover the
theory in a practical manner without reams of needless equation that send
the students to sleep followed by *practical* demonstrations of how it
works.

A good example of the interesting requirements of amplifer stability and
compensation can be found in just about any competent discrete audio hi-fi
amplifer btw. It's remarkably easy to make a great oscillator.

As for demonstrating the importance of phase margin - why not devise a
simple unity gain op-amp circuit that's stable - but with an additional
passive phase lag stage from the output, of lowish impedance, that the
feedback can be switched to which should then oscillate - then show how
adjusting the gain should be able to recover stability. You should also be
able to calculate the frequency at which it'll oscillate.

Graham
 
J

Jim Thompson

Jan 1, 1970
0
That's great rich,
I'm fairly ok with the lesson plan, but i'm short on practical
aids and examples. I could just scrible up a few circuits on pspice but
i'd love to get a look at how other people have attempted it and then
take what i like about their views, and ideas of my own, and meet
somewhere in the middle.

For the lesson i'm going to stick to a simple style;
Introduce
Explain
Demonstrate
Get the students to imitate
Get the students to practice

Would anybody do anything differently?

(1) Have your students had passive circuit theory including Laplace
analysis, and understand "phase"? You did say "third year"?

(2) Get a copy of Middlebrook's paper that discusses measuring phase
margin. (Referenced on my website.)

(3) Drop the "dr" from your E-mail address. I'm very leery of people
who are "title-pompous". I knew a fellow from not-a-few years back, a
co-instructor of mine, who insisted on being addressed as "mister"...
his son is now on death row.

...Jim Thompson
 
The e-mail is a joke, i'm only 20, so i hardly have a doctorate.
I couldn't just use dr_quinn for obvious reasons.
Thanks for the continuing input lads, keep it comin...
 
G

Genome

Jan 1, 1970
0
Hi guys, I've a project on amplifier stability and compensation. I'm
about half way through it now, anyboody got any info or anything
interesting. It's a fairly simple project, what i have to do is complie



notes that will be used to teach 3rd year electronics students about
amplifier stability.


The next stage is where i create four one-hour lab session to
demonstrate how to;


calculate the loop-gain,
show how the loop gain varies with frequency,
show how a amplifier becomes unstable,
demonstrate how to stabilise it.


I haven't completely decided on these labs so if anybody thinks they'd
do it differently please let me know.


I've already decided that i'm going to use simple op-amp circuits in
these labs. What i need to do is create these circuits. Does anyboddy
have any examples / suggestions/ ideas on what circuits would suit the
labs?

I would be inclined not to start out with amplifiers or op-amps. They are
designed to be stable in the first place.

Why spend time taking something that is stable and then break it to prove a
point?

What you, and they, should be more interested in is 'feedback systems'. The
whole point of stability is to take an open loop system and then close a
loop around it to achieve a stable system whose performance is predictable.

The op-amp is a tool used to close such a loop. It's response, as an error
amplifier, is tailored according to the gain response of the original system
to achieve this.

But I wouldn't start there.....

The first thing I compensated was a phase locked loop, CD4046, op-amps not
included. I monkeyed about with the piece of shit trying to make it work,
even read AoE but did so superficially.

Did it want to play..... no it did not.

For the first time in my life I sat down and read through the words
properly. Then I did the sums. Then I applied the results and it worked.

I was, sort of, taught about loop stability on my Applied Science degree.
They used op-amps and bode plots and wein bridge oscillators but it didn't
sink in. Well perhaps some of it did subconciously.

Things started to come together when I had to do it myself.

So, that's your first lab.

Two weeks before.

Tell them they are going to be using a 74HC4046. Give them a copy of the
relevent section from AoE (I won't tell Win or Paul if you don't..... just
mention where it came from).

Tell them they are going to multiply the frequency of an input signal by a
number between 1 and 16.

Ask them to deliver a written report one week before the lab that says how
they are going to make it work and how they would change it so it doesn't.
Say why it does and why it doesn't.

Tell them they can cheat and talk to each other about it. The reports can be
singular or a collaboration, list the contributors.

Make sure that they understand it's a 'hands up if you didn't understand'
situation.

Any questions?

Tell them that four days in there will be a meeting for
questions/clarification.

Do the meeting.

Take the report(s) at one week and set another meeting for three days.

Read what they have written. Learn from 'your' mistakes.

Take the next meeting and tidy up loose ends, whatever they may be...

Do the lab.

Ask for reports one week later.

Read them, where did you go wrong?

Meet again for a revue.

DNA
 
The project is a to set the syllabus and labs for a hypothetical class.
It's not a practically excercise, and i won't ever actually be teaching
this. However your input is quite interesting, thanks Genome. I'm
actually gonna use what you wrote in my interview. It'll earn me plenty
of brownie points with the examining body.

Again, i'm looking for the "pot of gold" that has a few examples of
what lecturers actually use in the real world. It's be fantastic to
relate my situation to to how a professional goes about it, and that's
the info i figure you guys would have. So far i've got loads of food
for thought, i've a foolscap page of scribbled ideas down now and i'll
get kickin on them this week.

Cheers again guys, and keep the ideas coming!
 
The project is to set the syllabus (including labs) for a hypothetical
class. It's not a practically excercise, and i won't ever actually be
teaching this. However your input is quite interesting, thanks Genome.
I'm actually gonna use what you wrote in my interview. It'll earn me
plenty of brownie points with the examining body.

Again, i'm looking for the "pot of gold" that has a few examples of
what lecturers actually use in the real world. It's be fantastic to
relate my situation to to how a professional goes about it, and that's
the info i figure you guys would have. So far i've got loads of food
for thought, i've a foolscap page of scribbled ideas down now and i'll
get kickin on them this week.

Cheers again guys, and keep the ideas coming!
 
R

Rich The Newsgroup Wacko

Jan 1, 1970
0
On Sat, 07 May 2005 14:23:29 -0700, Jim Thompson
(3) Drop the "dr" from your E-mail address. I'm very leery of people
who are "title-pompous".

Heh. :)
 
R

Rich The Newsgroup Wacko

Jan 1, 1970
0
The e-mail is a joke, i'm only 20, so i hardly have a doctorate.
I couldn't just use dr_quinn for obvious reasons.
Thanks for the continuing input lads, keep it comin...

Hey, Dr. Quinn was one hot babe! (leer, snort)
 
Top