LEDs, internal resistors, large Vf

[saturn1bguy]

I would like to have a good answer when one of the kids asks me why. Up to now I have made do with saying that LEDs are greedy and without the limit resistor they may burn out. More of a what will happen then a how or why.

Looking at a typical LED datasheet, e.g. the one Audioguru posted, shows that Vf is advertised at a particular If. Looking at the curve, If runs away pretty quickly as Vf is increased.

 

[Sceadwian]

LEDs with a built in resistor is good for quick and simple lights but it might be better to teach a student the reason for a current limit resistor on an LED in the first place, or they're going to be confused the first time they plug a common LED into 5 volts and it goes up in smoke. Ohms law does not strictly apply to a diode as it's a semi conductor device and VERY non linear, the exact forward voltage and how fast the current increases after that point is going to depend on manufacturing variations and the ambient temperature, mind you this means that when the diode starts conducting it's going to heat up and it's forwards voltage drop is going to change.

 

[3v0]

Looking at a typical LED datasheet, e.g. the one Audioguru posted, shows that Vf is advertised at a particular If. Looking at the curve, If runs away pretty quickly as Vf is increased.

Does it follow ohms law I=V/R which would have I increasing directly proportional to V. Or does R change as V and I increase to make it non linear. Maybe R is temperature depedant?

I do not see where audioguru posted a link. He looked at the datasheet I linked to datasheet which has no graphs.

Do you agree or disagree with Audioguru

If the forward voltage is exactly 3.0V and you applied a 3.0V supply without a current-limiting resistor then the current could be a few mA or a few A.

 

[audioguru]

An LED is a diode, not a incandescent resistor.
If you allow for the temperature change affecting its resistance then an incandecent light bulb is a resistor.

You cannot apply Ohm's Law to a diode especially when you don't know its actual forward voltage.

An LED without current-limiting either burns out or does not light.
If you find one that has a forward voltage exactly the same as the battery voltage then its forward voltage reduces as it heats which might cause thermal runaway. Its blink as it burns out might be so fast that you don't see it.

 

[audioguru]

Does it follow ohms law I=V/R which would have I increasing directly proportional to V. Or does R change as V and I increase to make it non linear. Maybe R is temperature depedant?

An LED is a diode, not an incandescent resistor. Its current increases a lot when its supply voltage is increased a little.

I do not see where audioguru posted a link. He looked at the datasheet I linked to datasheet which has no graphs.

Your datasheet is from Taiwan and has no graphs.
My datasheet is from Fairchild and has many graphs:

 

[saturn1bguy]

Does it follow ohms law I=V/R which would have I increasing directly proportional to V. Or does R change as V and I increase to make it non linear. Maybe R is temperature depedant?

There is no linear relationship between voltage and current in a PN junction. An LED is still a diode: look at the forward conduction curve for a rectifier. We all say that Si diodes turn on at 600-700mV. That's the Vf. Once achieved, current rises logarithmically, i.e. dramatically! Similar for an LED, except at a lower Vf (a function of materials used). And just like any rectifier has a datasheet specified maximum forward current, so does an LED. We all use a 1N4001 for currents less than 1A because we know what its good for. We can't allow more current through it or else it fails. An LED too. After Vf is achieved, If rises dramically, soon to exceed Ifmax. Thus it must be limited.

I do not see where audioguru posted a link.

On the previous page (page one of two of this thread), he posted a JPEG. (I'm pretty sure it was him?).

Do you agree or disagree with Audioguru

We have both indicated that one must limit the current in an LED.

 

[3v0]

One has to balance what one could teach with what one needs to teach. The goal is to get them hooked, so to speak. Blinking lights and moving motors is what hooks the kids. The sooner we can get there the better.

At first I wanted to have them suffer through every gritty detail. I have since changed my mind. There is little value in using a resistor and a LED providing they understand what is going on. Class is too short as it is.

Last year I had the kids build an h-bridge from transistors. I provided a PCB which they populated using parts I supplied. They learned how it works and learned to read the schematic. It took way to long to get to the other end.

This year I am going to concentrate on programing the first half of the year. We will start with the Simulator and Junebug as a target and expand on to solderless breadboards. Teach them to use the Junebug LA tool. Driving a LCD with a shift register, SPI bus, that sort of thing.

I want to see them programing with the simulator and Junebug at the end of the first 1/4. The second 1/4 we will program robots that I will provide. The 3 and 4 1/4 will be building robots of their own.

LEDs with a built in resistor is good for quick and simple lights but it might be better to teach a student the reason for a current limit resistor on an LED in the first place, or they're going to be confused the first time they plug a common LED into 5 volts and it goes up in smoke. Ohms law does not strictly apply to a diode as it's a semi conductor device and VERY non linear, the exact forward voltage and how fast the current increases after that point is going to depend on manufacturing variations and the ambient temperature, mind you this means that when the diode starts conducting it's going to heat up and it's forwards voltage drop is going to change.

 

[3v0]

@audioguru and saturn1bguy

The curves mean a lot. I get it. Do not apply ohms law to semiconductors.

One more question. I am trying to understand how we rectify current with voltages many times that shown in the Vf If curves.

Suppose I build a bridge rectifier from 1N4001's and hook it to 12 VAC. I then regulate the DC down and hook it to a 200mA load. For kicks lets say the regulator eats another 50mA for a total of 300mA. The diodes will be operating past the Vf max on the graphs pass only the current required by the load If the load goes too high the diodes die.

Is that the way it works?

I feel like a kid asking why the sky is blue.

 

[audioguru]

I am trying to understand how we rectify current with voltages many times that shown in the Vf If curves.

When the AC voltage is near the peak voltage then the rectifier conducts with a current of up to 30A. At 30A its forward voltage is about 2.0V and the rectified output voltage becomes 2.0V less than the peak voltage of the AC.
The anode voltage rises with the AC signal rising and the cathode voltage follows it at up to 2V less.

Suppose I build a bridge rectifier from 1N4001's and hook it to 12 VAC. I then regulate the DC down and hook it to a 200mA load. For kicks lets say the regulator eats another 50mA for a total of 300mA. The diodes will be operating past the Vf max on the graphs pass only the current required by the load If the load goes too high the diodes die.

Is that the way it works?

No.
The forward biased diodes follow the AC voltage. The max forward voltage of a 1N4001 is 2.0V at 30A when it quickly charges a big filter capacitor. In your circuit the diode might not have more peak current than only 3A then its max forward voltage is only 1.0V.

 

[saturn1bguy]

One more question. I am trying to understand how we rectify current with voltages many times that shown in the Vf If curves.

Suppose I build a bridge rectifier from 1N4001's and hook it to 12 VAC. I then regulate the DC down and hook it to a 200mA load. For kicks lets say the regulator eats another 50mA for a total of 300mA. The diodes will be operating past the Vf max on the graphs pass only the current required by the load If the load goes too high the diodes die.

I'm not positive what you're asking, but:

There is no Vfmax, only Ifmax and Vrrm, which for a 1N4001 is 1A and 50V, respectively, as I recall. The diodes blocks below Vf, and do what you want when they're above Vf. As long as each diode doesn't draw more than an amp, or have to block more than 50Vp (35VAC), you're okay. (They don't graph Vf more than they do because If, the max allowed current, has already been shown. Meaning it's of no use to plot Vf more than shown because the diode has likely already died from too much If.)

(Plus, how'd you get to 300mA? I count 250mA, but that's neither here nor there.)

 

[saturn1bguy]

I'm done responding, as Audioguru beats me every time! (Although my delta-t is getting smaller...)

Sheesh