As basic as it gets

[lefty]

I'm embarrassed to ask this but here goes...

I'm struggling to understand why I'm getting different readings from my meter on the simplest of circuits. I have a circuit with 1 led, 1 resister and 5vdc. I calculated a resister of 220 ohms to get 22.7mA. But I'm getting different readings based on the color of the led. With a red, I have a 2v drop across the led and 3 volt across the resistor, which was what I was looking for. But with green, I have the opposite. I read that ideally, the LED has no resistance. I know we don't live in an ideal world but my numbers are dramatically different. Why am I having trouble understanding this?

I'm rounding these numbers but the point is, that I'm getting different results, different currents, and different voltages depending on the color of the LED with everything else being the same. I measured the resistor's resistance with my meter and it's dead on. If the resistance is close to 0 on an LED, should I expect the same numbers no matter which color I use?

I've tried this using a 5v pin from Arduino as well as a 5v source from an analog/digital trainer PAD-234-A someone gave me, which I'm pretty excited about.

Sorry for such an elementary inquiry.

Thank you

[ataradov]

Diodes are not linear devices, so they don't follow Ohm's laws. They do not have "resistance", but they do drop voltage. This voltage drop is different depending on the color and design of the LED. It is specified in the datasheet for the LED.

[Andy Chee]

This is due to the different chemical composition used to create the diode junction in "light emitting diode".

Wikipedia has a quick list of chemical composition and expected voltage drops.

https://en.wikipedia.org/wiki/Light-emitting_diode_physics#Materials

[ejeffrey]

LEDs are a type of diode and when forward biased they have an approximately constant voltage across them regardless of current.  They are said to have "low resistance" not because they have no voltage drop but because the voltage doesn't change with load current.

LED color is directly tied to the LED voltage, shorter wavelength LEDs have higher voltage drop.

So what you are seeing is totally expected, and your measurements are correct.  That's just how LEDs work.

[watchmaker]

I'm embarrassed to ask this but here goes...

I'm struggling to understand why I'm getting different readings from my meter on the simplest of circuits. I have a circuit with 1 led, 1 resister and 5vdc. I calculated a resister of 220 ohms to get 22.7mA. But I'm getting different readings based on the color of the led. With a red, I have a 2v drop across the led and 3 volt across the resistor, which was what I was looking for. But with green, I have the opposite. I read that ideally, the LED has no resistance. I know we don't live in an ideal world but my numbers are dramatically different. Why am I having trouble understanding this?

I'm rounding these numbers but the point is, that I'm getting different results, different currents, and different voltages depending on the color of the LED with everything else being the same. I measured the resistor's resistance with my meter and it's dead on. If the resistance is close to 0 on an LED, should I expect the same numbers no matter which color I use?

I've tried this using a 5v pin from Arduino as well as a 5v source from an analog/digital trainer PAD-234-A someone gave me, which I'm pretty excited about.

Sorry for such an elementary inquiry.

Thank you

I have asked much "dumber" questions.  Welcome!

I am on the same journey, but maybe a different path.

Can I suggest that now you have been provided enough information to investigate the results that are discrepant from your expectations?  This can lead to understanding the physical reasons behind the behavior of color LEDs, but diodes and semiconductors generally.

I recently did something similar with a FET, diode and LED.  I metered the gate and source currents, as well as voltages. I proved to myself that the gate of a FET really does work by the creation of a voltage field while passing no easily measurable current.

I think you found the right place.

[RoGeorge]

See the chart from this link https://electronics.stackexchange.com/questions/453467/two-different-led-types-diode-test-on-two-different-multimeters-shows-just-1

That is how the voltage drop on a LED is, by the current through the LED.  Each curve is for a different LED of a different color.

Voltage drop on a LED is correlated with the photon's wavelength.  A higher frequency photon (to the blue and further, ultra-violet range) carries more energy than a lower frequency photon (to the red then infra-red side of the light spectrum).  That how the physics of photons is.  The higher the frequency, the higher the energy a wave will carry, thus the more voltage drop in the LED.

[JohanH]

Good explanations above. Now that you have the background, you can easily calculate the resistor for different types of LEDs.

https://www.digikey.fi/en/resources/conversion-calculators/conversion-calculator-led-series-resistor

Due to the aforementioned non-linearity of LEDs, it's not possible to calculate exactly, so adjustment might be necessary if you use the LED for instance in a constant current source type of circuit that requires a specific current. But once set, it will be relatively stable.

[tooki]

See the chart from this link https://electronics.stackexchange.com/questions/453467/two-different-led-types-diode-test-on-two-different-multimeters-shows-just-1

That is how the voltage drop on a LED is, by the current through the LED.  Each curve is for a different LED of a different color.

Voltage drop on a LED is correlated with the photon's wavelength.  A higher frequency photon (to the blue and further, ultra-violet range) carries more energy than a lower frequency photon (to the red then infra-red side of the light spectrum).  That how the physics of photons is.  The higher the frequency, the higher the energy a wave will carry, thus the more voltage drop in the LED.

While that’s a cool chart in theory, I think they got some stuff seriously wrong, since yellow should be between orange and green, and white isn’t an LED color as such (every white LED is a blue LED with phosphors; UV-based white supposedly exists, but despite serious searching, I have found zero evidence of such a device ever being manufactured). The white and blue LED curves should be identical. (And indeed, for blue and white LEDs from the same series of LEDs, they usually are.)

Of course, in real life, the wavelength does not correlate precisely 1:1 to the Vf. High-efficiency LEDs generally have a higher Vf than low-efficiency ones of the same color.

[sparkydog]

yellow should be between orange and green, and white isn’t an LED color as such

...but (as you note!) "white" LEDs certainly exist. Probably the chart labels correspond to the visible color of each LED type rather than emitted wavelength. That being the case, the "yellow" on the chart might be a green and red LED in a single package (or a green-blue LED with phosphor?), which would make that curve much less surprising. In fact, that would also suggest that the chart is describing "white" LEDs using a wavelength somewhere in between an LED that is visibly blue and a UV LED. (Whether that reflects typical "white" LEDs, I couldn't say.)

In any case, such charts should be taken with a grain of salt, since no two LEDs (even of the same model!) are identical.

[Psi]

There's multiple chemistries that can be used to make a Yellow LED. we don't really know which one he tested.
From memory, both gallium phosphide and gallium nitride can make yellow leds, and also a while(blue) LED with yellow phosphor

[lefty]

Thank you to everyone for the responses. This was throwing me. If I'm to understand correctly, my choice of resister like this with LED's is more about controlling current than voltage?

When I was targeting 2v for the light based on my voltage of 5v, I figured 150 ohms. So when went with 220 ohms, I saw a little under 2 with the red light. But when I went with the green with 220 ohm resistor, I got confused when I saw almost 3v across the light.

Since reading the responses and trying different resister values, it makes more sense now.

[pickle9000]

Lefty, you explained the situation well. Remember "How can you know what you don't know?".

[JohanH]

You are right that you should set up LEDs based on current. You don't need to check the voltage, unless you use the LED in a specific circuit where you tune the circuit's behavior based on LED forward voltage (e.g. a constant current transistor bias circuit).

Also, you don't have to run LEDs at full current (e.g. 22.7mA) if you don't use them for illumination. If you use a LED as an indicator or similar, most run perfectly fine on 5 mA or even less and the light isn't then as eye-piercing (blue ones can be annoying). In battery powered applications you save a lot of current also.

[EPAIII]

Here is the step by step, basic way of calculating a series resistor for an LED:

1. On the data sheet for THAT LED, look up the forward Voltage for the current level you want to operate it at (Iled). This can be approximate, but make it as close as you can. And don't be surprised if individual LEDs differ so you can get different results from one to the next. Remember we are using basic units so this current is in Amperes, not milli-Amps. And Voltages are in Volts and resistances are in Ohms.

2. Subtract that forward Voltage from your supply Voltage. Vsupply - Vled = Vresistor. That is the Voltage that the resistor must drop at your desired LED current.

3. Now use Ohm's Law to find the resistor value. R = Vresistor / Iled. Use the closest standard value resistor.

4. Calculate the power level for the resistor: P = Vled x Iled. Common practice is to multiply by a safety factor: 2X is one suggested value but this can be fudged. Just don't use one with a lower power rating or it will go up in smoke.

[tooki]

yellow should be between orange and green, and white isn’t an LED color as such

...but (as you note!) "white" LEDs certainly exist. Probably the chart labels correspond to the visible color of each LED type rather than emitted wavelength. That being the case, the "yellow" on the chart might be a green and red LED in a single package (or a green-blue LED with phosphor?), which would make that curve much less surprising. In fact, that would also suggest that the chart is describing "white" LEDs using a wavelength somewhere in between an LED that is visibly blue and a UV LED. (Whether that reflects typical "white" LEDs, I couldn't say.)

In any case, such charts should be taken with a grain of salt, since no two LEDs (even of the same model!) are identical.

The chart isn’t plotting wavelength at all. It’s voltage vs. current.

My point is that because white (and other phosphor-converted) LEDs always use a blue (not blue-green, by the way; deep blue is optimal) LED light source, it shouldn’t have a separate curve at all. And that generally speaking (remember, that chart is labeled as “typical curves”, not “curves of exotic exceptions”), a yellow LED’s Vf should be somewhere between that of lime green and orange. So that chart is… questionable.

As I already said: despite numerous sources claiming that UV-pumped white LEDs exist, I don’t believe they have ever been manufactured. UV LEDs are still much more expensive than blue, and less efficient. Meanwhile the entire industry has optimized its phosphors around 405-420nm (deep blue).
Near-UV LEDs (like the 395nm commonly sold as “UV” online) are also still more expensive and less efficient.

[tooki]

There's multiple chemistries that can be used to make a Yellow LED. we don't really know which one he tested.
From memory, both gallium phosphide and gallium nitride can make yellow leds, and also a while(blue) LED with yellow phosphor

Heh, I have some cheap Chinese “warm white” LEDs that are basically yellow. :p

[MarkT]

While that’s a cool chart in theory, I think they got some stuff seriously wrong, since yellow should be between orange and green

No, it doesn't quite work like that - the voltage drop depends on the bandgap of the bulk semiconductor, but the colour depends on the badgap of the heterojunction active layer, which is typically a different alloy.  Modern high-brightness LEDs can be heterojunction (aka quantum-well) devices as they are much more efficient (100's of times often), due to the structure.

GaN based LEDs will all have the same basic voltage, yet might be green, blue, white, green-blue etc.  Optoelectronic devices have come a long long way in the last few decades.

Also the voltage can vary with doping level as well as the semiconductor alloy used.  I say alloy because most opto devices are exactly that, not pure semiconductor compounds.  In semiconductor lasers the heterojunction layer is also engineered to have a higher refractive index to trap the light like in an optic fibre.

[tooki]

While that’s a cool chart in theory, I think they got some stuff seriously wrong, since yellow should be between orange and green

No, it doesn't quite work like that - the voltage drop depends on the bandgap of the bulk semiconductor, but the colour depends on the badgap of the heterojunction active layer, which is typically a different alloy.  Modern high-brightness LEDs can be heterojunction (aka quantum-well) devices as they are much more efficient (100's of times often), due to the structure.

GaN based LEDs will all have the same basic voltage, yet might be green, blue, white, green-blue etc.  Optoelectronic devices have come a long long way in the last few decades.

Also the voltage can vary with doping level as well as the semiconductor alloy used.  I say alloy because most opto devices are exactly that, not pure semiconductor compounds.  In semiconductor lasers the heterojunction layer is also engineered to have a higher refractive index to trap the light like in an optic fibre.

While I am not disputing any of that (and actually it’s quite interesting), if you look at the source of that chart (http://lednique.com/current-voltage-relationships/iv-curves/ ) and the basic generalizations it makes, then yellow absolutely should be between orange and green.

A non-systematic perusal of yellow LEDs on Digi-Key didn’t turn up any based on GaN. The overwhelming majority of yellow indicator LEDs, even high-efficiency ones, have a Vf of 2.1V, right where it should be, just below the 2.2V of old-school lime green. Yellow power LEDs have even lower Vf at the usual 20mA, so again contrary to the chart.

Anyhow, my point throughout has simply been that the chart is sloppy and doesn’t seem correct for either traditional LEDs nor for modern power LEDs.

[sparkydog]

My point is that because white (and other phosphor-converted) LEDs always use a blue (not blue-green, by the way; deep blue is optimal) LED light source, it shouldn’t have a separate curve at all. And that generally speaking (remember, that chart is labeled as “typical curves”, not “curves of exotic exceptions”), a yellow LED’s Vf should be somewhere between that of lime green and orange. So that chart is… questionable.

As I already said: despite numerous sources claiming that UV-pumped white LEDs exist, I don’t believe they have ever been manufactured. UV LEDs are still much more expensive than blue, and less efficient. Meanwhile the entire industry has optimized its phosphors around 405-420nm (deep blue).

...and that is exactly why blue and "white" are different. I looked up a random blue LED and its dominant wavelength is... 470nm. 470 ≠ 405-420. "Visible blue" is not (necessarily) the same wavelength used to drive "white" LEDs. 420nm is actually pushing the limits of human vision, and is not optimal for something that's supposed to emit visible light. (And it's arguable whether that even counts as "blue", since "blue" is specified as 'predominantly between 450nm and 495nm'. 420nm is actually closer to the edge of ultraviolet than it is to "blue"!)

That said, my own (somewhat cursory) attempts to determine what's used to drive "white" LEDs suggest that the typical frequency is more likely 430-450nm... which is still not ideal for a "blue" LED. What is your source for 405-420nm?

[coromonadalix]

blue is considered  a recent color,   few years ago   it was not possible to create  blue in a single led,  technology evolved, and more colors appeared  over the years   from

the basic red green and the "citrus" yellow

now you have 2 3 4 pins leds  and can now do R G B W  colors   

Example :  led display panels,    normally the "W"hite was composed of the basic R G B  colors / intensity values, but now you can get to almost a pure white from a single led,  and  cold / warm / neutral   white too