Boolean algebra, A AND 1 = 0 WTF ?

[Simon]

The problem is that this country is an educational shit hole  it is easier to find courses on makeup and hair styling than engineering and serious subjects. My local college only does wishy washy courses their branch in a town 20 miles away used to do electronics but that has gone. There are far fewer courses in electronics than any other subject and engineering courses as a whole seem to have been reduced, hence i could hardly find one by distance learning. I've got a maths book on order that amazon are making me wait 2 weeks for just because i didn't use them as the carrier that I'll use to get a head start but maths in itself is boring and to understand it it helps to have it related to something practical.

[Simon]

Ah yes of course, what does the short lived section on opamps tell me to do at the end ? go to this address (wikipedia) to read all about opamps and read about all the things we just pretended to teach you !

[vk6zgo]

Yes but the explanation implied that you would be expected to run an LED straight off a voltage source but for manufacturing tolerances, they failed in the explanations of diodes, zeners and LED's to explain that a diode based device cannot be connected to a non current regulated voltage source or it would short out and blow, the explanations were of no practical use and need more information before I would be able to use the components but for the fact that I know this stuff already.

Their explanation is essentially correct on the whole. I don't think you should be so ready to dismiss it. You are being presented with information from a slightly different perspective than you have thought about before. Rather than dismissing it, you should consider whether your current understanding is incomplete.

You can, of course, connect an LED to a non-current regulated voltage source. An LED has a voltage/current curve just like any other device. If you apply the right voltage to it the expected current will flow and it will light up. It will not short out and blow...it will simply pass as much current as the applied voltage induces.

In reality an LED does have variations in manufacturing tolerances and other causes of variation in its voltage/current curve and that makes it difficult to reliably apply the right voltage. That is exactly why you need to stabilize the current with a series resistor in a practical circuit.

In many circuits an LED needs to be operated from a supply much higher in voltage than that which appears across it in operation.
An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow.

In practice,it is seldom necessary to match the current through multiple LEDs in such simple circuits,as they are usually panel indicators.

As the human eye is quite insensitive to small variations in light level,the difference between say,20mA,18mA,& even 12mA,is vanishingly small.

[Simon]

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

I am now reading through 25 pages of opamp explanations on wikipedia as directed by the course that spent a mere 5 wishy washy pages on it.

[Bored@Work]

Wikipedia is not really a trustworth source. Try http://www.ti.com/lit/an/slod006b/slod006b.pdf

[GeoffS]

After watching Dave's videos on op amps, I've been browsing through Op Amps for Everyone from TI, I've even been understanding some of it!
I've had to take a few breaks to brush up on my maths over at the Kahn Academy.

EDIT: Beaten to the punch 

[IanB]

An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow.

Why would it need a series resistor? If you apply the correct voltage you will obtain the desired current flow.

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

The explanation given was correct.

Physical systems can often be looked at in different ways, and that means that there can be different, but equally correct, ways of explaining their behaviour.

There seems to be some misconception that a diode without a series resistor will turn into a short circuit, pass infinite current and blow up. This is not the case. A forward biased diode will pass current according to its I/V curve. Since the slope of the I/V curve is very steep in the normal operating region, a practical circuit needs some means of stabilizing the current to keep it at the value desired, but this does not undermine the fact that in principle you could apply the correct voltage and get the current desired.

[Simon]

Wikipedia is not really a trustworth source. Try http://www.ti.com/lit/an/slod006b/slod006b.pdf

Tell that to the asshole that wrote the course

[vk6zgo]

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

I am now reading through 25 pages of opamp explanations on wikipedia as directed by the course that spent a mere 5 wishy washy pages on it.

I would suggest you go to your local University & have a look through their secondhand bookshop.
You may well find some "Treasure trove" in the form of old Electronics books.

Boolean algebra has been around forever,as have Op Amps,so you should find something useful.
The very best pickings is when the Uni has a mass sell off of books,which may occur every few years.

Misprints  & "glossed over" subjects in Math & Technical Books have been around for years,but if you have another reference you can check whether it's you,or the text.
Unfortunately,Internet sites,including Wikipedia although they have an aura of authority,are equally vulnerable to typos & errors.

[Simon]

An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow.

Why would it need a series resistor? If you apply the correct voltage you will obtain the desired current flow.

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

The explanation given was correct.

Physical systems can often be looked at in different ways, and that means that there can be different, but equally correct, ways of explaining their behaviour.

There seems to be some misconception that a diode without a series resistor will turn into a short circuit, pass infinite current and blow up. This is not the case. A forward biased diode will pass current according to its I/V curve. Since the slope of the I/V curve is very steep in the normal operating region, a practical circuit needs some means of stabilizing the current to keep it at the value desired, but this does not undermine the fact that in principle you could apply the correct voltage and get the current desired.

firstly we are talking about a simple course, what you refer to which I understand is somewhat advanced for this course and seldom used in practice. Take a look at your current source driving an LED, what is being used to sample the current for making adjustments ? oh let me guess a resistor......

[Simon]

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

I am now reading through 25 pages of opamp explanations on wikipedia as directed by the course that spent a mere 5 wishy washy pages on it.

I would suggest you go to your local University & have a look through their secondhand bookshop.
You may well find some "Treasure trove" in the form of old Electronics books.

Boolean algebra has been around forever,as have Op Amps,so you should find something useful.
The very best pickings is when the Uni has a mass sell off of books,which may occur every few years.

Misprints  & "glossed over" subjects in Math & Technical Books have been around for years,but if you have another reference you can check whether it's you,or the text.
Unfortunately,Internet sites,including Wikipedia although they have an aura of authority,are equally vulnerable to typos & errors.

You are under the misapprehension that there are universities around here that teach electronics, that is the problem, there is very little available in the Uk in the way of electronics courses of either type. Even the renown open university does not do electronics.

[vk6zgo]

An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow.

Why would it need a series resistor? If you apply the correct voltage you will obtain the desired current flow.

my point exactly, the whole explnation given was redundant and should have been a different and correct one.

The explanation given was correct.

Physical systems can often be looked at in different ways, and that means that there can be different, but equally correct, ways of explaining their behaviour.

There seems to be some misconception that a diode without a series resistor will turn into a short circuit, pass infinite current and blow up. This is not the case. A forward biased diode will pass current according to its I/V curve. Since the slope of the I/V curve is very steep in the normal operating region, a practical circuit needs some means of stabilizing the current to keep it at the value desired, but this does not undermine the fact that in principle you could apply the correct voltage and get the current desired.

It would be nice if you quoted the complete paragraph:
"In many circuits an LED needs to be operated from a supply much higher in voltage than that which appears across it in operation.
An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow."

The second paragraph follows directly from the first.

[IanB]

It would be nice if you quoted the complete paragraph:
"In many circuits an LED needs to be operated from a supply much higher in voltage than that which appears across it in operation.
An ideal LED would not have the variations referred to above,but would still need a series resistor to limit current flow."

The second paragraph follows directly from the first.

I'm sorry about that, but I'm trying to address Simon's point here:

no I'm sorry but this course is just crap all over. The explanation of using a current limit resistor on an LED was incredibly stupid, apparently you need the resistor to make up for the differences in voltage between LED's even though made in the same batch (manufacturing tolerances), they totally omitted the fact that a diode will not control current but allow as much through as you try and push until it blows and that is why you need the resistor.

The series resistor is to make up the difference between the higher supply voltage and the required operating voltage of the LED.

[Bored@Work]

Wikipedia is not really a trustworth source. Try http://www.ti.com/lit/an/slod006b/slod006b.pdf

Tell that to the asshole that wrote the course

No, I am telling that to you. You want to learn something.

[Simon]

I think have that one but thank you and I can't take a 400+ page aside as part of this course as I have 11 months to complete it. aim here is to learn what there is to learn from it and get the bit of paper so my employer is happy. Then I can worry about going further on my own.

[GreyWoolfe]

When I went to ITT Technical Institute in the late 90's, over the course of the 2 year AS-EET program, we found 3 or 4 errors in the textbooks we were using.  The class teachers had no issue with us finding the mistakes as long as we proved why there was an error and what the correct information was.  Then we were allowed to use our answers.

[IanB]

firstly we are talking about a simple course, what you refer to which I understand is somewhat advanced for this course and seldom used in practice. Take a look at your current source driving an LED, what is being used to sample the current for making adjustments ? oh let me guess a resistor......

Let me illustrate with an example. Here is a test that I just ran, with a voltage source driving an LED. There is no current regulation and no resistor in the circuit. I am simply placing 2 V across the LED and it is quite happily lighting up and passing 7 mA. There is no magic smoke.

So why don't we simply run all red LEDs with a regulated 2 V supply? It's because of variations in LEDs, due to manufacturing tolerances, design variances, ambient temperature, and so on. So how do we allow for that? We use a series resistor to compensate for the differences.

This is not advanced, it is simply what you need to know to design a practical, working circuit.

[vk6zgo]

firstly we are talking about a simple course, what you refer to which I understand is somewhat advanced for this course and seldom used in practice. Take a look at your current source driving an LED, what is being used to sample the current for making adjustments ? oh let me guess a resistor......

So why don't we simply run all red LEDs with a regulated 2 V supply? It's because of variations in LEDs, due to manufacturing tolerances, design variances, ambient temperature, and so on. So how do we allow for that? We use a series resistor to compensate for the differences.

I'm sorry,but that's just plain silly!

We use series resistors as it is a far more economical alternative when the LEDs are incidental to the operation of the circuit,& the operating voltage of that circuit is higher than that which could directly power a LED.
Apart from anything else,when LEDs first made an appearance,a regulated 2v supply was as rare as hen's teeth.

I don't know anyone who selects series resistors so the LED has exactly the voltage required for that individual device.

They have a look at what voltage normally appears across a generic LED of that colour,subtract it from
the supply voltage,& using Ohm's Law,determine what resistor is needed for around 20mA or so through the LED,

Or,if it is a 12v supply, they try 1kOhm,& if that isn't bright enough,go to 560 Ohm,or 470 Ohm.

[Simon]

Ian, well done but your missing the point entirely. Take that LED off and get another one from a different source that looks the same and wee what happens. Tell me honestly would you propose that solution for a design that was going to be made by the 1000's ? what would happen when the exact model of your LED was not in stock of a different batch came in from the manufacturer, can you stake whatever you like on it still working ? what happens with temperature variation, if we are going to have the advanced version of the discussion which the course material was certainly not trying to have lets put all the factors on the table not just a one off experiment where you fine tuned it for that one single individual LED.

I am currently looking at lighting LED's in a battery powered box and i do need to make the most of my power because I will have some AA batteries powering 80mA of LED's for a week, so my plan is to come as close to what you have done without loosing my head: I will use as low a value resistor as i can and run them off a small SMPS with a voltage as close to the LED voltage as i can get, but again this will be a one off and i can afford to make a broad assumption on the parallel supplied LED's being pretty much identical and using just a few ohms of current limit with a fine tuned voltage supply...... but I'd be being more careful if i was getting a manufacturing house down the road to make me a thousand.

[Kremmen]

If you are planning to use an SMPS as part of the solution then may i recommend you check the Supertex HV 9910B. It is a simple buck type _current_ controlled LED driver that avoids the series resistor and its power loss. There is a series current sense resistor that is unavoidable but depending on your exact numbers it could be a near optimal solution. Notably, the chip supports constant-off-time mode as standard thus avoiding certain undesirable buck characteristics when the duty cycle is very low.
I have used and recommended the HV9910B earlier with good success.

[IanB]

Ian, well done but your missing the point entirely. Take that LED off and get another one from a different source that looks the same and wee what happens. Tell me honestly would you propose that solution for a design that was going to be made by the 1000's ? what would happen when the exact model of your LED was not in stock of a different batch came in from the manufacturer, can you stake whatever you like on it still working ? what happens with temperature variation, if we are going to have the advanced version of the discussion which the course material was certainly not trying to have lets put all the factors on the table not just a one off experiment where you fine tuned it for that one single individual LED.

Yes of course, that's why we use a series resistor. But aren't your words above exactly what your course material was saying, which you rejected as wrong back in post #8?

I'm not trying to tell you that you can make a design that runs LEDs off a constant voltage supply. That would be silly. I am trying to tell you that the explanation given in your course material, that variations between LED samples and other factors mean that you need a dropper resistor to control for those variations, is quite a reasonable explanation. It's the same explanation you have given above.

Specifically, my experiment was designed to show that an LED without a dropper resistor can in fact control the current flow and it will not necessarily turn into a short circuit and blow up as you suggested. There is quite a range of voltages I can apply to that LED above and below 2.0 V without it blowing up.

[Simon]

unfortunately i have 4 independently switched LED's so I'll have to use voltage regulation and current limiting resistors but it's only 15mA each so not much sweat

[Kremmen]

unfortunately i have 4 independently switched LED's so I'll have to use voltage regulation and current limiting resistors but it's only 15mA each so not much sweat

Well - you could get creative and connect the LEDs in series and short out  those you want to blank . That would be perfectly OK for a current source but would need a mechanical contact or a slightly complicated transistor/FET switch for each led.

[Simon]

Given the time i have or rather have not I'll stick with what I mostly know, at the rate I'm going it will be batteries and resistors only

[Simon]

It would also mean for my 4x2.6V LED's I will end up needing twice the battery power planned before I account for drained batteries.