.

Cyril:

I only this evening found this thread, and, as always, I am amazed

at the confusion that seems to permeate many of the answers that

your question has generated. One of the aspects of your postulate

raised my eyebrows a little higher than usual:

.

>  How should I present the diode at the lab? I was thinking - as a

>  switch .... as a voltage stabilizer .... or as a non-linear element?

.

A semiconductor junction diode is none of these, although it most

closely deserves the third characterization. A diode - and its close

relative, the bipolar transistor - are very special elements, and it is

the poor presentation of these devices than continually throws the

student into a quagmire of false notions - about their basic modes

of operation and the unique, fundamental, but opaque, equations

which characterize them. Indeed. I would venture to say that a PN

junction behaves in a way that is closer to Nature than any of the

other elements associated with electronics - with the exception of

our common roots - L, C ,R, antennas, and perhaps magnetically-

coupled transformers. 

.

There is no other path of dealing with such devices except through

the door opened by William Shockley:

.

   IF = IS(T) exp (VF/VK) otherwise stated as VF = VK log (IF/IS(T))

.

where VK is the 'thermal voltage', kT/q and IS(T) is a very peculiar,

extremely tiny, massively temperature-dependent, but wonderfully

magical quantity, rooted deeply in the physical world, and diffused

with the integer numbers 3, 32 and such constants as pi; Planck's 

h; the mass of the electron and its ghost, the hole; temperature of

course, and the fundamental band-gap energy of the material.

.

(It is not quite this simple. For one, these are simply the constants

that are implicated in defining ni2 the intrinsic carrier concentration.

Nevertheless, these roots are so deeply important and so beautiful

that they are the right place to start, if you ever hope to experience

the joy of this subject).

.

However, God being Good, there is a much easier way; although if

your students aren't thoroughly familiar with exp and log they need

to go back to school for a while. Or... you take the time to explain it

all to them.  But before getting into that, let me mention something

that's quite unforgivable in advising a student how to "think about a

diode"; namely, that  "its forward voltage is 0.7V" (assuming we are

speaking of a silicon junction). I know very well that Glen Ellis is far 

from naive in suggesting this; but I think this notion gives a student

a very wrong - and even dangerous - idea about junction behaviour

in general. I can't tell you how many times I have cringed when I've

queried a job interviewee:  "What is the forward-biased voltage of a

silicon junction?" to be told "It's 0.7V". 

.

So, how can we capture the essence of such a junction - let's keep

it down to a (somewhat simplified) diode model - without getting on

our knees searching for Planck, or Boltzmann, or even Shockley?

It's very simple. First, we know - from our professorial experience -

that the measurement of IS(T), even at an elevated temperature, is

a fool's errand, for all sorts of reasons. So let's not even try.

.

Rather, take a sample, firmly in your hand, of the general variety of

diode you are considering, and a thermometer in the other.  Find a

power supply, of as high a voltage as you can borrow from the lab.

For example, using a 30 V supply, and a resistor of 2.94 Megohms

(a standard 1%-tolerance value in the E96 set) connected in series

with a correctly polarized diode establishes a current of quite close 

to 10 uA, assuming, as a guess, that the VF of the diode is ~0.6 V.

Check: IF = (30-0.6) V / 2.94 Megohms = 10uA.

.

Say your digital - thus, presumed accurate! - thermometer shows a

temperature in a somewhat chilly environment of 16.95oC. In Kelvin

terms, that's a rather awkward number of 290.1K. Temperature is a

very important factor in determining VF, so we have to be diligent in

this matter. At this temperature, VK = (k/q) x 290.1 = 25.0 mV. Now

you can say that the forward voltage is simply this:

.

      VF = 0.6V + 25mV x log (IF/10uA)  at a temperature of 16.95oC

.

So, if we wonder about VFat IF = 100 uA, it will be 25mV x log(10)

higher than 0.6 V which is 657.6 mV and It increases or decreases

by the same amount (57.6mV) for each decade of change in IF.

.

If this were a scholarly treatise, a great deal more could be said on

this topic. But it's time for my annual bath and then - I trust - dinner

at Chateau Gilbert. 

.

Remember, young ones: If you are not having fun doing electronics,

it's not too late to become a dentist.

.

The Lone Arranger.....

Dear Cyril - from my experiences, it is helpful (and necessary) to show the students the difference between a "quasi-linear" and a non-linear voltage-to-current relationship. In particular, it is important to distinguish between static (R) and dynamic (r) resistances.

More than that, IMO an interesting application of diodes for the purpose of amplitude limiting is a sinewave oscillator (WIEN) - and the question: Which resistance matters? Static or dynamic?

I can hardly recall my university times. From my current point of view, two applications of the humble diode are really important: rectifier (aka switch) and limited (ok - more the TVS sort of diode than the 'simple' one). Anyway - the characteristic curve(s) and how to derive resistance/differential resistance from these are really important for these applications.

BTW: a long time ago I got a green LED emit orange. Just a matter of 20 mA vs. 200 mAs. This one survived, but when I tried to repeat the experiment (some years later) you could always see a minor flash when the bond wire tripped :)

For intro (basic) students, I would show the diode as a switch, at .7V.

For engineer students,

I would show the Dynamic Impedance of the diode, first. 

Then (1) show how the diode can be a simple 'switch'  

or (2) a dynamic 'logrithmic' device. 

All depends on the perspective

of the design for the circuit. 

For me, 1975, engineering student, 

seeing the simple diode as a dynamic 'logrithmic' devices  

was the beginning of real 'fun' and discovery in electronics. 

The 'myth' of the simple diode faded away 

and I had found a 'real' scientific tool to use in designs. 

IIn my opinion, the Internal Dynamic Impedance 

Depends on the level of impurities in the diode.  

A combination of old-time Resistance 

and Diode-Dynamic-Impedance. 

Experiencing the world of electronics

and the great joy of discovery is important. 

...

As we progress along, we are better able 

to share our sense of wonder and amazement 

to the community of our students

and improve the world around us. 

http://www.geocities.ws/glene77is/articles_ET_files/article_ET_AFX_CW_Limiter.html

http://www.geocities.ws/glene77is/bmp_et_limiter/a-OAD-DDDG-1-B-Zoom3-333-141009-a.jpeg

http://www.geocities.ws/glene77is/bmp_et_limiter/a-OAD-DDDG-1-B-666-141009-a.jpeg

Dear colleagues,

I see that the nonlinear properties of diodes are most interesting to you... and they really deserve to be discussed (maybe in a separate question)... but my question here is about the simplest and most intuitive notion about the diode - thinking of it as of a switch... or, more precisely speaking, as of a one-way valve...

I am not sure if "limiter" is the exact term for this circuit... or maybe "clipper" is more appropriate... But I want to clarify what is the idea of this network of a diode and resistor in series that is "suspended" between the two sources - the input  AC and the reference DC one...

When using the diode as a one-way-valve, the denomination 'rectifier' is commonly used.

Each student of Electronics (Electrical Engineering) should know the properties and uses of diodes in all their glory - DC and dynamic behavior too. Strictly speaking, he should know the behavior of PN junction in all situations - DC, AC, pulsed excitation, heat, light.

It can save a lot troubles in the practical life.

Indeed, it is evident already from what is stated above.

Prof. Cyril,

Hats off to you.

you are a Tireless Researcher and Inventor of Exploring Basic Circuits and Circuit Components.

I am Happy that you are training your Students also to Step into the Unknown and New Realms of Basic Of Circuits.

I wish You and Your Students,who are lucky to Work Under you,Every Success.

Regards.

Prof.P.S.Subramanyam

"...but my question here is about the simplest and most intuitive notion about the diode - thinking of it as of a switch."

Cyril - referring to the pictures you have shown, it is clear that the dc source causes the diode to be "off". Hence, it is up to the sinewave to "open" the diode for amplitudes which are above the DC voltage.

But what will the students see? The function of a switch (the upper part of a sinewave)? No - they will see that the diode is not only a "switch" but a device that is non-linear (distorted upper parts of a sinewave). That means: IMO from the beginning you and the students are confronted with the non-linear characteristics of the diode. 

Hi Cyril

and what about "ideal diode" (DC only) - see annex :-))

Josef

use of Half Wave and Full Wave Diode Rectifiers in Magnetic Amplifiers may also be exposed to the Students.

P.S.

Dear Pisupati,

Which teacher would not be touched by such words of encouragement? Thank you very much, they mean a lot to me... they inspire me and make invest even greater effort in working with my students...

Today I had a lecture exactly with these students  when my phone get a notification to your comment ... I presented you to them as a well known Indian scientist and an experienced educator... and then I read your message; they were delighted...

Thanks also for the examples of half-wave and full-wave rectifiers in magnetic amplifiers. I only fear that these amplifiers are not well known in today's "electronic times".

Dreher,

IMO we can use both "rectifier" and "limiter"; only in the first case, we focus on what this valve passes, and in the second case - what it stops. So, we can say either "this is a rectifier of positive voltage" or "this is a limiter of negative voltage".

Dear Lutz,

Sorry, but I will have to disappoint you because I do not see any sense to involve the non-linear characteristics of diodes in this "pure switching" application.

Just think how much more clearly and intuitively understandable for a human being is to imagine the diode as the well-known from the routine electrical switch or valve (case 1 in the atached picture)... rather than as the so abstract non-linear resistor with its differential and static resistance... It is better to talk about discreetly changing resistor with two resistance values (zero and infinite) than for non-linear and differential resistance...

So that no the diode is what limits the sine wave on the upper part but the reference voltage source, which is connected in parallel to the output during this region of limitation... and determines the output voltage... The function of the diode is only to connect the voltage source to the output.

Of course, everywhere here I mean an ideal diode switch (with VF = 0).

"...superimposing them..."

Aparna, I have the feeling that you are very close to the right answer:)

Josef,

Here is my answer (and of my students in 2008):

https://en.wikibooks.org/wiki/Circuit_Idea/Group_66a#Lab_4a:_How_to_make_perfect_circuits_by_series_NFB ("ideal diode" in series)

https://en.wikibooks.org/wiki/Circuit_Idea/Group_68b#Lab_4:_How_to_make_perfect_components_by_parallel_NFB ("ideal diode" in parallel)

So, the "formula" of the "ideal diode" is:

"Ideal diode" = "real diode" + "VF voltage source"

@ Cyril

OK - we obviously have a different understanding of the English word 'limiter':

YOU use it for something I'd name "blocking device".

MY understanding of 'limiter' is the "transient suppressor device" action: limit something to an acceptable level. Thus "limiter" is a word I'd never assign to a rectifying diode - not even when considering the blocking aspect. What about "blocker" ?

Cyril,

When talking about diodes i usually distinguis between AC and DC circuits. In DC circuits i refer to the diode as a nonlinear resistor, whose resistance depends on the applied voltage/current. 

On the other hand when talking about diodes in AC circuits, i use the terms "Ideal diode model" and the "Offset diode model". I suppose that is what you mean.

Check this file:

http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-071j-introduction-to-electronics-signals-and-measurement-spring-2006/lecture-notes/17_diodes1.pdf

"Sorry, but I will have to disappoint you because I do not see any sense to involve the non-linear characteristics of diodes in this "pure switching" application.

Just think how much more clearly and intuitively understandable for a human being is to imagine the diode as the well-known from the routine electrical switch or valve (case 1 in the atached picture)... rather than as the so abstract non-linear resistor"

Cyril - perhaps you misunderstood my comment. It was only related to the rough diagram you have shown (Diode, DC source with reverse biasing, AC source, resistor). And I was asking myself: What do the students see on the scope? The function of an ideal switch? This was my only concern: Discrepancy between measurement and simplified theory.

Sir,

But, it can also change the Inductance Values as Saturable Reactors,just as we can change the Value of the Resistance using Diode Switching..

P.S.

Dear Lutz,

Students will see (have seen) the function of a limiter on the scope's screen. More specifically, they will see in time the signals generated by two sources - an AC source (when the "switch" is open) and a DC source (when the "switch" is closed)... and this is the purpose of this experiment...

And to repeat again: The amplitude of the AC signal is high enough (24 V) compared with the negligible VF (0.7 V).

Cyril - OK, there was a misunderstanding between us. We are speaking about different circuits.

In the diagram I was referring to (chalk on a blackboard) there was no RL and the scope was connected across R1.

Now you have added another resistor RL - and you are measuring the voltage across this new resistor. That means: Now the "diode switch" produces something like a short  across RL (very small current through RL for pos. halfwaves)  - in contrast to the former diagram (chalk on blackboard) where we had a large current through R1 for the pos. halfwaves. 

Really, another viewpoint at this circuit is to think of the reference voltage source as of a Zener diode...

Dreher, as I can see on the web, "diode clipper" is most frequently used:

https://www.google.bg/search?q=diode+clipper&safe=off&espv=2&biw=1366&bih=643&site=webhp&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiO3omP8srLAhUn_XIKHRo6AQEQ_AUIBigB

https://en.wikipedia.org/wiki/Clipper_(electronics)

Maybe I should change the title of the question...

Diode-transistor logic?

And what about the circuit in the annex :-))

Aparna,

What do you think about the logical function of the input diodes - AND or OR?

Josef,

Bridge rectifier + "common" emitter stage... common but ungrounded? Flying input source and floating emitter... but the collector output (Y) is grounded?

The actual output voltage is the voltage across the collector resistor... and it is referenced to VCC... But we use the complementary voltage VC between the collector and ground...

As I can see, you frequently exploit this idea in your questions:)

Hi Cyril

Grounding provide inputs.

I thought the input (floating) voltage was between A and B... Maybe this could be another application:)?

I think OR:)

detailed solutions

Interesting and unexpected...

If we add invertor with one tranzistor we get EXOR function.

Dear Aparna,

The input diodes of DTL are connected in parallel... and this configuration is usually used to implement an OR logical function. But here, you say, they implement AND function. Then what is the trick?

Maybe this explanation will help to see it:

https://en.wikipedia.org/w/index.php?title=Diode_logic&oldid=445500585#AND_logic_gate

... and the diodes are inverse connected and biased... to act as "normally closed switches"... a really clever trick...

Cyril,

Radical thinking, and yet so basic.  

Thanks.

.

Cyril:

I only this evening found this thread, and, as always, I am amazed

at the confusion that seems to permeate many of the answers that

your question has generated. One of the aspects of your postulate

raised my eyebrows a little higher than usual:

.

>  How should I present the diode at the lab? I was thinking - as a

>  switch .... as a voltage stabilizer .... or as a non-linear element?

.

A semiconductor junction diode is none of these, although it most

closely deserves the third characterization. A diode - and its close

relative, the bipolar transistor - are very special elements, and it is

the poor presentation of these devices than continually throws the

student into a quagmire of false notions - about their basic modes

of operation and the unique, fundamental, but opaque, equations

which characterize them. Indeed. I would venture to say that a PN

junction behaves in a way that is closer to Nature than any of the

other elements associated with electronics - with the exception of

our common roots - L, C ,R, antennas, and perhaps magnetically-

coupled transformers. 

.

There is no other path of dealing with such devices except through

the door opened by William Shockley:

.

   IF = IS(T) exp (VF/VK) otherwise stated as VF = VK log (IF/IS(T))

.

where VK is the 'thermal voltage', kT/q and IS(T) is a very peculiar,

extremely tiny, massively temperature-dependent, but wonderfully

magical quantity, rooted deeply in the physical world, and diffused

with the integer numbers 3, 32 and such constants as pi; Planck's 

h; the mass of the electron and its ghost, the hole; temperature of

course, and the fundamental band-gap energy of the material.

.

(It is not quite this simple. For one, these are simply the constants

that are implicated in defining ni2 the intrinsic carrier concentration.

Nevertheless, these roots are so deeply important and so beautiful

that they are the right place to start, if you ever hope to experience

the joy of this subject).

.

However, God being Good, there is a much easier way; although if

your students aren't thoroughly familiar with exp and log they need

to go back to school for a while. Or... you take the time to explain it

all to them.  But before getting into that, let me mention something

that's quite unforgivable in advising a student how to "think about a

diode"; namely, that  "its forward voltage is 0.7V" (assuming we are

speaking of a silicon junction). I know very well that Glen Ellis is far 

from naive in suggesting this; but I think this notion gives a student

a very wrong - and even dangerous - idea about junction behaviour

in general. I can't tell you how many times I have cringed when I've

queried a job interviewee:  "What is the forward-biased voltage of a

silicon junction?" to be told "It's 0.7V". 

.

So, how can we capture the essence of such a junction - let's keep

it down to a (somewhat simplified) diode model - without getting on

our knees searching for Planck, or Boltzmann, or even Shockley?

It's very simple. First, we know - from our professorial experience -

that the measurement of IS(T), even at an elevated temperature, is

a fool's errand, for all sorts of reasons. So let's not even try.

.

Rather, take a sample, firmly in your hand, of the general variety of

diode you are considering, and a thermometer in the other.  Find a

power supply, of as high a voltage as you can borrow from the lab.

For example, using a 30 V supply, and a resistor of 2.94 Megohms

(a standard 1%-tolerance value in the E96 set) connected in series

with a correctly polarized diode establishes a current of quite close 

to 10 uA, assuming, as a guess, that the VF of the diode is ~0.6 V.

Check: IF = (30-0.6) V / 2.94 Megohms = 10uA.

.

Say your digital - thus, presumed accurate! - thermometer shows a

temperature in a somewhat chilly environment of 16.95oC. In Kelvin

terms, that's a rather awkward number of 290.1K. Temperature is a

very important factor in determining VF, so we have to be diligent in

this matter. At this temperature, VK = (k/q) x 290.1 = 25.0 mV. Now

you can say that the forward voltage is simply this:

.

      VF = 0.6V + 25mV x log (IF/10uA)  at a temperature of 16.95oC

.

So, if we wonder about VFat IF = 100 uA, it will be 25mV x log(10)

higher than 0.6 V which is 657.6 mV and It increases or decreases

by the same amount (57.6mV) for each decade of change in IF.

.

If this were a scholarly treatise, a great deal more could be said on

this topic. But it's time for my annual bath and then - I trust - dinner

at Chateau Gilbert. 

.

Remember, young ones: If you are not having fun doing electronics,

it's not too late to become a dentist.

.

The Lone Arranger.....

Dear Barrie Gilbert,

Thanks for the so comprehensive and valuable explanations... and also for your wit which enlivens the discussion and makes it colorful.

Cyril

À mon avis, d'une vie dans ce domaine,

Wit is the life-blood of all invention.

Hi dear Cyril 

I think logical operands were well known before the invention of diodes.What i have realized in this regard is, if we have two switches say S1,S2  OR operation is literally either S1 OR S2 is ON (switches connected in parallel) to drive the Load.  similarly for AND operation both S1 AND S2 must be ON i.e series connection of switches.For Not Operation When switch is ON Load is OFF(NOT).This is used even in Pneumatic,Hydraulics and so on wherever controls are necessary.

I know, many text books presented this, but readers grasps it with restricted environmental background. Hope my words are not mistaken.

best regards

vasanth   

Dear Vasanth,

I fully agree with you and second your approach... and am glad to see someone thinking as me...

I have said many times that basic ideas do not depend on their implementations; they are the soul while specific implentations are the body. There are (and there will) many implementations but the basic idea is the same...

Thanks for the contribution,

Cyril 

Thank you dear 

Its my pleasure that i have soul mate somewhere else on this planet and we look at the technical  issues differently but enthusiastically with a new perspective

regards 

vasanth

Maybe we should establish a "club of circuit enthusiasts" inside RG:)

Cyril, 

Sense the message so abruptly , clearly, thoroughly

and humorously presented,  ( see Dr. Lief posted above ).

I found it encouraging to be reminded that

even the 'lowly' diode can be viewed as

a device "closer to Nature"

than we had originally penned here. 

When I try to follow the "yellow brick road" of the electron,

then I am drawn away from simple descriptions

and further into real explanations

which show the wonders of the world

where we have been placed.  

In school, I was told to focus on the description,  even though I was graded on my attention to the explanation. 

I think there is a Life (of its own being) Living in this Topic. 

Creo que hay una vida (de su propio ser) que viven en esta zona.

On the other hand,

I tend to agree with your classroom approach. 

I do understand that

if the instructor presents the basic ideas

("soul") of the device too soon 

then that student will be overloaded ,

and he will be without language. 

Thus, in tutorial practice, 

You  do well to present

the implementation ("body")  of the device, First, 

and let the student grow into the larger experience. 

I look forward to seeing where you can lead this topic. 

Glen

Cyril,

Regarding the "Club for Circuit Enthusiats" ,

Vasile Surducan has a "Question"

on "Do Exotic Amplifiers Even Exist?" 

as linked below. 

His Question is not about a "circuit collection", 

but an "examination of circuits" , 

seeking unique properties. 

https://www.researchgate.net/post/Do_Exotic_amplifiers_even_exist#56f3f61e48954c2495263be0

"Thus, in tutorial practice,

You do well to present the implementation ("body") of the device

and let the student grow into the larger experience."

Glen - I respect, of course, that every tutor/teacher/instructor/lecturer has its own method for presenting explanations for technical phenomena to the audience. But I must admit, in this case, I do not agree with you. 

(Explanation: As soon as I introduce the diode simply as a switch - active at app. 0.7V - the students will ask: What is the "secret" behind this switch? What is internal? As a consequence, I start with the pn junction and its properties).

Always, I try to follow A. Einsteins rule:

"Try to make your explanation as simple as possible - BUT NOT SIMPLER!"

IMO "generalization" does not mean "simplification"... and it is always useful to reveal and show the general idea on which the particular implementation is based.

So, what is the "pure" idea behind this circuit solution particularly implemented with a diode switch? Let me do it consecutively as a scenario of 4 steps as I did it with my students last week:

Step 1. The first idea that can occur to us is to assemble the output signal by two input signals - AC and DC.

Step 2. To implement this initial idea, we should commutate two input voltage sources - the input AC source and the reference DC source (i.e., to disconnect the one source from the output and to connect the other source to the output). For this purpose, we need a 3-terminal switch.

Step 3. But in electronics we have only 2-terminal switches (diodes, transistors...) Then how do we commutate the two voltage sources? If we connect them in parallel, a conflict (short connection) will appear since they are perfect ("ideal") voltage sources with zero internal resistance.

Step 4. Thus the final powerful idea occurs to us - to deliberately weaken the one source ... to artificially increase its internal resistance by connecting a resistor in series to it.

Now, when connecting the other (stronger) voltage source in parallel, it will force the weak source and will determine the output voltage.

As you can see above, I managed to show the most general idea of the clipper without mentioning details like "pn junction", "forward voltage", etc... at this stage they are not necessary...

See for comparison a more conventional story about diode limiters that tells what has been done but does not say why it was done so.

http://www.wikihow.com/Understand-Basic-Diode-Clipping-Circuits

In simplistic visual model perspective, I would describe (zener,regular) diode as a (two way, one way) spring valve inside water tube/channel.  If water pressure is low, the valve remains closed which means there is no water flow in the channel.   If water pressure is high, the valve is open and therefore water will flow through the pipe.  Water flow is compared to current flow and water pressure to voltage.  To understand the 'secret' of diode clippers is to understand the purpose of the circuit and then its components functions in simplistic visual perspective.  The rest will follow.

Exactly, Adam... the diode is not simply a switch; it is a one-way switch, i.e., a valve.

But this metaphor explains what that element is. In addition, we must also explain what this circuit (built on this element) does and how it does it...

.

Cyril:

.

I have read through the note 

.

www.wikihow.com/Understand-Basic-Diode-Clipping-Circuits

.

and must say it is very incorrect and could be misleading,

Treating the diode naively as a "short circuit" without even 

mentioning its forward-biased voltage is to simplify matters

to such an extreme as to turn genuine learning into a sort

of chocolate box of nice spot-answers to hard problems.

,

It might be okay to start with this explanation spending just

a few minutes; then say "Now, let's refine this explanation

one step further", at which point you explain the diode will

only turn on when the voltage across it reaches a certain

threshold, to which we can give a fixed value of "typically

0.7V for a silicon diode". Having sounded down that idea,

and thus accounted for the "clipping level" to be 700 mV

higher than VREF, you take tour students one simple step

further to see that, in fact, the diode has what can naively

be called "an internal resistance", the result of which is

that the output does not "clamp" nicely, but continues to

increase with the input voltage. 

.

But you had better quickly "bite the bullet" and introduce

the concept of a log/exp element and explain that this is

going to make the actual calculations very much harder,

.

This sort of step-wise refinement of one's explanations

of circuit behavior I call "FOUNDATION DESIGN".

.

The Lone Arranger

Dear Barrie Gilbert,

I fully agree with your sequence of presentation of the diode... and follow it in my lectures and labs devoted to this semiconductor device...

...But when I reveal the secret of how to make an "ideal diode" by using an op-amp, I am tempted to present this network of two elements in series ("real diode" + "op-amp") as a short connection... or just a piece of wire...

https://en.wikibooks.org/wiki/Circuit_Idea/Group_68b#Lab_4:_How_to_make_perfect_components_by_parallel_NFB

https://www.researchgate.net/post/Can_we_virtually_decrease_the_forward_voltage_drop_across_a_diode_up_to_zero_How_do_we_make_such_ideal_diodes_with_zero_voltage_threshold2

Hi Cyril,

but the last circuit is "dynamic lazy" because the OPA output passes to saturation for one input polarity. Josef

Well... but if we insert another (high-ohmic, decoupling) resistor before the inverting input and close the negative feedback during this "input polarity" by connecting another diode (in an opposite direction) between the op-amp output and the inverting input?

.

Yes... the customary answer is to add a second diode in

opposite polarity directly from the op amp output back to

the inverting input.  But there's more to the behaviour of

this innocent circuit than may meet the eye.

.

The Lone Arranger

Fellows,

BGHS,  this is not an "innocent circuit" . 

(a)

At low hand-switchable rates, this circuit may act like a switch.

But at much greater speeds, the diode capacitance will begin to reduce the DUT Diode's internal dynamic impedance.  

( I am not including the OpAmp chip feed-through leakage. ) 

(b)

Does a more perfect Active-Diode circuit

require a forward bias on each of these diodes ?    

There is an OpAmp chip transient delay  driving the diodes

from

(1) zero-crossing

to

(2) acceptable diode forward conduction. 

In a long-ago-school paper,

I split the dual active-diode circuit into two stages, 

Positive and Negative half-waves, 

with Both stages having a small forward bias to the diode.

Then summed them together

for a final ( almost ) full-wave output. 

The Transition Time was still measurable 

... but the benefit was that

the OP responded measurable faster, 

and with measurable less distortion. 

Maybe, these considerations make things more difficult

than you want for an intro class . 

Glen

Dear Barrie Gilbert,

I just want to note that if we simply "add a second diode in opposite polarity directly from the op amp output back to the inverting input", the output voltage of this circuit will be always zero (virtual ground). The negative feedback loop will be always closed excluding only the moment of transition through zero (between -VF and +VF) when both diodes will be cut off.

That is why I have proposed above to insert a decoupling resistor between the inverting input and the input resistor. I am not sure if this is the perfect solution but it just occurred to me then.

Regarding your note about "accounting for the "clipping level" to be 700 mV higher than VREF" in the simple diode clipper, it is interesting to note that the total clipping voltage can be considered as composed of two parts - one voltage drop of 700 mV (across the diode) and one reference voltage (produced by a source).

For a comparison, in the circuit of a Zener diode clipper (ordinary + Zener diode), the clipping voltage is composed of two voltage drops.

In the circuit of the active diode clipper, the reference voltage can be applied to the non-inverting input. The op-amp will compensate the forward voltage drop of 700 mV and the clipping voltage will be composed only of the reference voltage.

Dear Glen,

Sorry, I have not read your insertion when writing my comment.

Yes, I also think that the biasing is a remedy like in the case of the op-amp buffered by a push-pull emitter follower included in the negative feedback loop.

.

Cyril:

.

I'm not sure what we are trying to achieve, now. What is meant'

by "[we] managed to reveal the secret of diode clippers (limiters)".

What is the "secret" that I am missing?

.

A "ideal diode" generally means one that conducts powerfully in

forward voltage bias, and not at all when this presumed voltage,

connected directly across the device, is of the opposite polarity.

I visualize this as the "right angle diode".

.

As for "diode limiters", there are numerous ways of applying the

basic element, depending on whether one is dealing with some

type of RF circuit - where the speed requirements generally will

not allow the use of amplifiers for the function - or perhaps when

accuracy is the objective, and then it's usually more a matter of

of preserving the (log/exp) voltage/current relationship, in which

case, one usually has to prevent the final (voltage) output from

changing its polarity - in fact, becoming exactly zero - when the

input is of the opposite polarity to the desired mode of operation,

.

There's hardly any secret about any of these operations. But it's

very late - about midnight - and I have to shut down this machine

- as well as the computer - and put them both in the sleep mode..

.

The Lone Arranger.

IMO the basic circuit idea behind diode clippers is to change the voltage produced by a real voltage source (with some internal resistance) by connecting an ideal voltage source (with zero internal resistance) in parallel.

In the case of a serial clipper, the input voltage source is ideal and the reference voltage source - real; in the case of a parallel clipper, the input voltage source is real and the reference voltage source - ideal.

I have exposed this idea on Page 6 in a form of a 4-step scenario:

We can observe this general idea in many situations in life when the strong impose their will on the weak without complying with them... or two forces - large and small, are superimposed... and the resulting force is equal to the greater of them...

.

Aparna:

.

I am also "at the output"of this pointless thread.

.

The Lone Arranger

.

Cyril,

Hoping that this is not really a 'pointless thread' 

and that you are  simply  'in the process' 

of developing your own ideas.  

You wrote: 

I have exposed this idea on Page 6

in a form of a 4-step scenario:

Difficult to follow your previous answer. 

It seems to devolve into an analogy to real life . 

So, 

I am looking for the above mentioned reference.    

Is this 'Page 6' in a publication of yours ?  

or here in this problem on "Answers page 6 " ? 

Since I have been through experiments 

with various limiter/clipper schemes  

( linked previously )

I would like to review your material. 

Thank you for the diagram of Cutcher OPA, which I used 

without the diodes bias  in my own project.  

I eliminated the diodes bias in order to increase 

the odd harmonic content of the narrow pass-band signal. 

( an effort to increase comprehensible communication ).  

*****

Now, back to exploring your presentation on Diodes 

the apprentice Glen

Hello Glen,

Nice to hear you!

Two comments for your circuits:

- about BJT class B output stage.

Adding just one resistor of few hundred ohms to make feed-forward connection from BJT bases to their emitters will greatly reduce the generated crossover distortions.

- about diode limiter

I do not see the purpose of "logarithmic negative control" network.

It only reduces two times the value of resistor R91.

Dobri

Dobri,

You are right on target both times. 

The Cutcher audio BJT pair was a minimalist approach 

which may only suitable for communications grade audio

which I find in radio work.  Normally, the very narrow pass-band signal will become only a sine shaped dit / dah and sound very soft to the ear.   I wanted to have more odd harmonic , more crispness ,  in the very narrow band audio I use for CW reception.  The OPA rings more when it jumps from one BJT to the other in the Class B transition.   

The double control on the limiter was only an experiment which proved NOT useful.  Just tinkering.  

K7MEM advised me years back, that if the ear/brain cannot detect a difference in the processed audio, then the circuit should be kept simpler.  

I am involved with communications grade audio, CW operations on the radio bands.  Rough sounds and high noise levels are the standard fare.  

On the other hand,  

I realize that in music audio, there are no short-cuts

... a prelude must sound like music from the heart.   

Your excellent graphs remind me that the world

has gone digital processing for a reason

... unlimited ability to calculate

... and revise and calculate again. 

Analog us an art-form for some of us geezers

who want to wear the robes of the electron,

who want to feel the natural world. 

Study Well... Life is a beautiful experience,

Glen 

Dear Glen,

Thank you for your attention to my question. I admire your sophisticated circuit solutions that solve important technical problems. But I want to clarify that my question is about the humble passive diode clippers - devices that cut off a part of a large signal above/below a reference level... and particularly, about the case where the reference voltage is produced by a separate voltage source (not by a Zener diode)...

In these applications, especially when the two voltages are high enough, the diode charactersitics (the shape of the IV curve, the forward voltage VF, etc.) are not crucial and can be ignored... and we can think of the diode simply as of a valve (a switch controlled by the current through it or by the voltage across it).

In all these implementations, we can discern four elements connected in series: an input voltage source, a reference voltage source, a diode and a resistor. And if we are of this category of people who not only want to know that it is done so but also want to find out why it is done exactly the way, we should be able to reveal the role of each of these four elements... and then explain it to other people... In particular, best teachers should be able to do this for their students...

And because I believe I belong to this category, at the beginning of each of my courses, I firmly promise to my students that I will not draw any element in any circuit diagram without explaining its purpose...

That is why, I have fabricated the story above: first, to show the general idea, and then, how it is implemented by these specific elements...

And, as you can see, the trick is extremly simple and intuitive... and can be seen everywhere in our routine. Let's see it again in these specific electrical implementations...

Here, at the moment when the input voltage becomes equal to the reference voltage, we have to commutate the two voltage sources... to replace the one by the other... If both they are ideal sources, we have no other way to do this than to disconnect the one from the output... and then to connect the other to the output...

But as we are clever and inventive enough, we solve this problem in an unusal way - we worsen the one ideal voltage source (make it real by adding some "internal" resistance) and constantly connect it to the output. Thus initially it determines the output voltage. Then, when the input voltage reaches the treshold, we just connect the other ("good") voltage source in parallel to the "bad" one... Thus, as though we disconnect the "bad" voltage source from the output and replace it with the "good" one...

As you can see on the blackboard, both voltage sources try to set the common output voltage (point Y). But while the right (reference) source does it via the low resistance (forward biased) diode, the left (input) source does it via the high ohmic resistor; so the output voltage will be determined by the right source. We can think of the RD network as of a parallel voltage summer with weighted inputs where the weight coefficient of the right input is 1 and of the left input is 0.

So once learned this circuit trick, we can implement it with any switching elements acting as valves - from the past, present and future (tubes, diode, some other...) since we now the idea... we have the "key"... This is the power of "pure" ideas...

Cyril,

Thank you for elaborating.  

I follow the presentation. 

I have used a idea similar (link)  

but have not ever explored it so rigorously as you have. 

Certainly, I did not follow it through further,

as I had a specific goal in mind at that time.  

It just occurred to me that my HeathKit HW8 tranceiver

uses a single version of this,

as a Band-Switch circuit controllers. 

Back-Biasing Diodes can turn a point on/off 

allowing RF signals to pass one path or the other. 

This is a good presentation. 

Will think about it more, the apprentice Glen 

While waiting for some response of the other participants in the forum to my "philosophy" about diode limiters, I will elaborate it a bit ...

Here we change the voltage across a real voltage source by connecting in parallel an ideal voltage source. Similarly, we can change (divert, steer) the current through a constant-voltage non-linear resistor (voltage stabilizer) by connecting another constant-voltage non-linear resistor with close but lower threshold. This trick is known as "current steering" and is widely used to make current switches.

Here is an implementation of this idea...

In 1982, I obtained a patent for a LED voltage indicator consisting only of three LEDs, two transistors and three resistors...

It consisted of a green LED0 (VF = 2.5 V) driven by a current source... so the diode lighted... 

This was the situation when the input voltage was zero.

When I connected a red LED1 (VF = 1.5 V) in parallel to the green one, the latter extinguished... (why?)

I implemented this idea with an n-p-n transistor... that turned on when the input voltage became positive...

Then I added another but p-n-p transistor... that turned on when the input voltage became negative...

Thus I assembled the full circuit of this bizarre voltage indicator...

If you are curious enough, you can observe where currents flow in this circuit for the three cases (-VIN, 0, +VIN)...

In this request for grant of a patent, I proposed also a "diamond" version of this circuit. Here is a scanned picture from the patent description...

Cyril,

"""When I connected a red LED1 (VF = 1.5 V) in parallel to the green one, the latter extinguished... (why?)"""

* The LED(1.5V) acts as a shut regulator

holding the Vin to 1.5V.  

* Last dual-Transistor example is

     V(+) , zero-crossing, V(-) . 

I think the V(zero) will have a +/- 0.6V window

because the transistor bases require bias. 

This is an interesting Step-Wise presentation.

I want to enjoy a  walk-through  carefully. 

This may seem to be a 'basic' subject, 

and the 'secret' may be in varying the back-bias, 

but you are making this interesting

by doing an orderly survey . 

As is the nature of these discussions, 

you appear to be building the ideas as you go along.

the apprentice Glen

Dear Glen,

My motto has always been "Build to understand circuits!"

http://www.circuit-fantasia.com/tutorial/intro/welcome.html

...or even better, "Understand, build and invent circuits!"

http://www.circuit-fantasia.com/circuit_stories/combined_list_of_circuits.htm

Cyril,

In examining the browser loaded source from the URL , I see that it requires a MacroMedia player,

which I do not have. 

Here, I run Linux, from flash drives,

and some custome code. 

The whole server presentation is off-limits to me. 

But ...

"Build to Understand Circuits!" is plain as daylight. 

IMO, these are examples of Applied Understanding.

In that vein of thought,

I come closer to Barrie's insight

that the student is not seeing enough

of the Physics and Theory of the Diode.  

However, I am patient,

and that physical theory may be covered

in the next courses at your school. 

Glen 

Glen : "* The LED(1.5V) acts as a shut regulator

holding the Vin to 1.5V.  

* Last dual-Transistor example is

     V(+) , zero-crossing, V(-) . 

I think the V(zero) will have a +/- 0.6V window

because the transistor bases require bias."

Dear Glen,

At last I could see concrete questions about specific circuit ideas ... rather general and vague qualifications... how much I love this...

Exactly, "The LED(1.5V) acts as a shut regulator" that deprives the current initially flowing through the 2.5V green LED. Note that there is a base resistor that expands the input voltage range... The LEDs fade in/out and actually this is an analog indicator...

In 1983, I published it in the BG magazine "Radio and TV". I regret that the article is in Bulgarian, but I think the figures well illustrate the ideas ...

Cyril,

Yes, you call it 'analog' , as it is obviously.  

The base bias is achieved gradually.  

For 'student' students the 'concrete' examples

may be most easily presented and best understood.

( I would have jumped ahead and used a quad OPA 

and made five levels which switched. ) 

That belies my biggest problems in school,

as I was always the 'engineering' student

who was in love with the ideas and concepts.  

Glen

Maybe the 3-LED voltage indicator above seems too simple?

The multiple LED indicator below is more sophisticated... It is based on the same powerful current steering idea...

Cyril,

Do I see this right ?  

You have I(in) from an adjustable Vee ? 

You call it "Current Steering". 

Thank you for the interesting example. 

It is amazing to see a good engineer's mind at work. 

Glen,

In these arrangements of parallel-connected diode elements with slightly different forward voltages, the current is steered to the element with the lower voltage threshold...

Here is another linear LED indicator where the forward voltages of LEDs are artificially increased by connecting ordinary diodes in series.

Cyril,  "the Tutorial Engineer"

These OPA versions I have used before. 

Your several diagrams, all together, 

form a progressive "set" for presentation. 

Today, I need to start some soldering 

on my own project.  

It is an "advanced AFX" narrow filter  

having double-notches at -60dB

and SPICE stop-band at -98dB. 

After fabrication I expect not-so-good figures. 

the apprentice Glen

http://www.geocities.ws/glene77is/bmp-et-3RLFADI/AFX_3RLFADI-v9-S-8-a.png