Calculate Current

[xrunner]

Could help myself, video coming shortly!

Oh neat, I will be interested in your results so I can compare ... 

[EEVblog]

00:00 - The Forum question
02:12 - Would this make a good job interview question?
03:36 - Dave's Solution
07:04 - The Solution
08:08 - Other forum responses
11:40 - Kirchoff's Current Law MUST Hold!
12:51 - REBEL against The System!
15:10 - Xrunner's practical test
21:38 - Practical Test

[strawberry]

thermal dependant circuit that reach equilibrium at some specific point
can use LED string as photo cell to lit up one LED

[gbaddeley]

It’s a trick question.

[Nusa]

My guess is poster asked a question about the problem, but did not give us the actual question asked about the circuit. So we're all spinning our wheels for no particular reason.

[not1xor1]

My previous post was quite confuse and unclear, so I'm attaching a simulation screenshot to better describe what I meant.
I adjusted the voltage supply and resistor to get about 20mA through the left LED.
I think transient simulation renders better the non linear behaviour of diodes.

[Capernicus]

How you power it makes a difference,    If I charged it with 2 volts 20ma peak output, but with just a battery,  I still have a theory that all the leds would be off.

[xrunner]

How you power it makes a difference,    If I charged it with 2 volts 20ma peak output, but with just a battery,  I still have a theory that all the leds would be off.

Well why don't you test your theory, wire it up, and tell us what it actually did ... 

[mikeselectricstuff]

A completely literal answer would be that B and C pass no current, as there is no difference between total Vf and the supply.
But as there is also zero resistance it could be infinte current.



 

[TrickyNekro]

Of course, theoretically it can work... And you have to calculate it iteratively based on the I = Is *  ( exp ( Vd / n * Vt ) - 1 ). You know theoretically the n and the Vt so you get the Is. Because you got two LEDs over the single one, it isn´t really sure that you are going to get 20mA through B, probably you are going to get less, going off my head here.

Practically, you wouldn´t even put diodes in parallel because of the differences in manufacturing and temperature ( as in Vt ), which can lead to a runaway effect.

Practically, you also have your average LED lamp committing that crime out there, only take seriously the ones that have many many LEDs in series and hope the lead resistance is enough to balance out.

BTW, just for the extra mind boggling, you also have to make sure no light is shining on the LEDs, or equal amount of light for all that it matters, go have fun now :-P

[Capernicus]

How you power it makes a difference,    If I charged it with 2 volts 20ma peak output, but with just a battery,  I still have a theory that all the leds would be off.

Well why don't you test your theory, wire it up, and tell us what it actually did ... 

I know with 2 leds in series, you need 4 volts to power it.   I swear the initial post was 2 volts in,  and he changed it!!!

[Capernicus]

Think about it what you will,   but I bet if I took a uni course on electronics I would get in the lowest 1% percentile of the class.

Thanks Dave for the vid,   If I had 120 leds Id make a video of me putting a 500 ohm 1/4 resistor on them and just plug it straight into the wall as a thnkyou vid,  but I dont have the leds.

[hadibaria]

Everycircuit simulation results. default LED settings Vf = 2v under 20mA

[Wytnucls]

Your calculations exactly match the results given by the EveryCircuit simulation.
Tomorokoshi (reply 21)

[Tomorokoshi]

The range of answers and results just goes to show:

You never can tell with D's.

[Wytnucls]

I'm not sure the person that formulated that question expected this level of accuracy, but you are obviously on the right track. 
(Note to self : I must eat more corn!)

https://eng.libretexts.org/Bookshelves/Materials_Science/Supplemental_Modules_(Materials_Science)/Solar_Basics/D._P-N_Junction_Diodes/3%3A_Ideal_Diode_Equation

[lacek]

Dave, sorry, but the solution you have shown in the video is wrong. I agree that this question is somehow wrong and it lacks  the U(I) curve for the LED diode.  But you can't assume that one diode drops 2V @ 10mA and other drops 20V @ 20 mA. 

I would say that Tomorokoshi's solution is as good as one can do.  Just assume exponential dependence of U on the current.


This is somehow what you do in your video.  If you assume Ib = 20mA then  you have Ic = 10 mA and the voltage drop Vc is < 2V.  if you assume IC=20 mA, then the IV=40 mA and VB >2V 
in first case you have total voltage < Vcc in the latter you have > Vcc.  So the answer would be that
IB is "between 10mA  and 20 mA"
IC is "between 20mA and 40 mA"

Tomorokoshi's wrote system of equations that give solutions that satisfy the above

I agree that it is not a proper exam questions as it is not possible to solve without U(i) characteristis. If this is "open answer" (so not ABCD) then there is a place to write the above.
I disagree with most of the criticism about idealizations. Yes the circuit is "academical", but the idealizations simplify the problem and allow to consider just the core of the problem, not to dwell on the detail: are the diodes matched or not. This is important, but for learning it is important to start from simple problem, not to dwell on such digressions.

[lacek]

Of course, theoretically it can work... And you have to calculate it iteratively based on the I = Is *  ( exp ( Vd / n * Vt ) - 1 ). You know theoretically the n and the Vt so you get the Is. Because you got two LEDs over the single one, it isn´t really sure that you are going to get 20mA through B, probably you are going to get less, going off my head here.

Practically, you wouldn´t even put diodes in parallel because of the differences in manufacturing and temperature ( as in Vt ), which can lead to a runaway effect.

Practically, you also have your average LED lamp committing that crime out there, only take seriously the ones that have many many LEDs in series and hope the lead resistance is enough to balance out.

BTW, just for the extra mind boggling, you also have to make sure no light is shining on the LEDs, or equal amount of light for all that it matters, go have fun now :-P

I am sure this is problem from some sort of Circuits I class or some "electricity for physicists" course. They were taught during the lecture what is the dependence of diode voltage on the current and they are supposed to solve a problem of determining how to solve a simple circuit with a non-linear U(I) dependence.

The formula that Tomorokoshi has used most likely was covered in the lecture and students are assumed to know about it and to use it for a solution. The problem stated as is is not "self-consistent" but it is written for a student group that all attend the lecture. It is not expected to remind all necessary formulas in problem text. The students are supposed to find them.


I see no point in complaining about "unpracticality" of said situation. Its like taking any physics problem that deals with physics of the material point in outer space and complaining that thing almost always have non-zero diameter and are affected by light pressure from sun or relativistic effects that have to do with gravitation that modify the 1/r^2 force. Yes they do, and from the context of class it is obvious does whether they should be included or not.

In Phyisics/Electronics one can always consider "more accurate model", but both "rough models" and "very accurate models" have their uses.

[lacek]

@Dave:  assuming 2V at middle junction is simply wrong.  You run in contradictions as you correctly state in the video.

[bdunham7]

The formula that Tomorokoshi has used most likely was covered in the lecture and students are assumed to know about it and to use it for a solution.

Unfortunately, Tomorokoshi's solution can't account for temperature because the thermal characteristics are unknown--so he appears to just assume the two temperatures are the same. Even if we had a datasheet, we might not know if the I/V curve was done quickly at a constant temperature or perhaps reflects an equilibrium at a standard ambient temperature.  Any analysis of 'ideal diodes' or even real ones without considering temperature is not going to be very helpful.

I understand the idea of progressing from basic to advanced, but I think it is not impossible to put a little thought into questions and examples to make sure that they aren't nonsensical under more advanced analysis.  I can't think of any basic principle or theory that cannot be demonstrated with a circuit that actually works.

[hans]

My guess from an academic standpoint is that a solution should be able to provide the solution as derived by @Tomorokoshi. It is possible to calculate that with Ohm's law, KCL and the ideal diode equation.

All the comments about the 2 parallel diodes should be matched or else they will burn out, is our practical thinking hat. There are always certain levels of theory vs practice you account for. If you're doing precision analog stuff, then you will also know that resistors have a certain tolerance, a temperature coefficient and even a voltage coefficient. Or if you learn about opamps, first you will learn how wonderfully excellent amplifiers they are, but then learn about offset voltage/current, a finite gain bandwidth, supply swing, and that the open-loop gain is finite, therefore even a unity voltage gain opamp stage is not exactly "unity"..

[xrunner]

I am sure this is problem from some sort of Circuits I class or some "electricity for physicists" course. They were taught during the lecture what is the dependence of diode voltage on the current and they are supposed to solve a problem of determining how to solve a simple circuit with a non-linear U(I) dependence.

The formula that Tomorokoshi has used most likely was covered in the lecture and students are assumed to know about it and to use it for a solution. The problem stated as is is not "self-consistent" but it is written for a student group that all attend the lecture. It is not expected to remind all necessary formulas in problem text. The students are supposed to find them.

I would like to know how etnel came across this problem, if he was in a class where it was presented or just stumbled upon it somewhere. The forum indicates his account was active today, so I will PM him and ask him nicely to tell us.

(of course nicely  )

[rstofer]

I am sure this is problem from some sort of Circuits I class or some "electricity for physicists" course. They were taught during the lecture what is the dependence of diode voltage on the current and they are supposed to solve a problem of determining how to solve a simple circuit with a non-linear U(I) dependence.

Or, this is a problem in early DC circuits (Kirchhoff's Laws) LONG before diodes are even introduced. 
I'm not convinced that applying the diode equation is appropriate at the level I assumed for the question.
Yes, the circuit is an abomination but that's one of the reasons I chose the simple model (the LED Vf is always 2V regardless of current).  Nobody would ever actually design something like this.


"All models are wrong but some are useful"

https://stats.stackexchange.com/questions/57407/what-is-the-meaning-of-all-models-are-wrong-but-some-are-useful

[lacek]

Or, this is a problem in early DC circuits (Kirchhoff's Laws) LONG before diodes are even introduced. 
I'm not convinced that applying the diode equation is appropriate at the level I assumed for the question.
Yes, the circuit is an abomination but that's one of the reasons I chose the simple model (the LED Vf is always 2V regardless of current).  Nobody would ever actually design something like this.

I am not so sure. For this problem diode equation is essential. Actually at my university there is an "electronics" course for physicists. There are literary 6 lectures, with crazy pace, and this problem would fit nicely to that course. The audience are people who know basic, ohmic circuits, at this course they are taught some practical aspects of pn junctions, bipolar transistors,mosfets, and some basic of signal processing. They are in their second year, so they have already learned the analysis, mathematical methods in physics etc.


As for "Nobody would ever actually design something like this." this is never consideration for courses for non-engineering students. There are numerous problems on "infinite series of resistors" after all.  For example Problem 2 from IPhO 1967.

[lacek]

The model "(the LED Vf is always 2V regardless of current)" is not bad as a "rough approximation".  you get 10mA instead of 14mA.... which can be still usable for some power dissipation considerations. Just healthy safety margin should be used.