Determining LED values

[Stoneww]

I'm just curious if there is a way to get the LED threshold voltage, or the current rating from any other way than checking the manufacture's datasheet.

 

[Terry01]

You can get the forward voltage with most multimeters but the current "rating" you need the data sheet. You could work the rating out yourself with a power supply and a few resistors but it would cost you your led when you found out the "max" current the LED would accept.

I think....I'm new so looking for someone to give the correct answer and I'll learn too

 

[Stoneww]

You can get the forward voltage with most multimeters but the current "rating" you need the data sheet. You could work the rating out yourself with a power supply and a few resistors but it would cost you your led when you found out the "max" current the LED would accept.

I think....I'm new so looking for someone to give the correct answer and I'll learn too

Thanks man really helpful sadly I don't think my multimeter has the forward voltage setting because it only cost me like £6

 

[Terry01]

It's the diode setting you need. -->|- kinda like that. The guys here will have the proper diode symbol. Like I say I'm still learning too so we will both learn when someone gives us the correct answer.

 

[Audioguru]

The datasheet for an LED will show its range of forward voltages. If it is white, blue or bright green then it will have a voltage range from about 2.8V to about 3.5V and each one will be different even if they have the same part number. The voltage also changes when the temperature changes so you cannot power an LED from a voltage source, it needs a current source which can be a resistor in series.

If you must parallel LEDs then do what Chinese flashlight manufacturers do: buy thousands of LEDs and hire a kid to test them all and sort them into groups that have the same forward voltage. But each parallel group must be fed a current, not a voltage.

 

[kellys_eye]

You can do it 'rule of thumb' using just the colour, from around 1.2V for red to 3.6V for white.



But I tend to break them down into the Red/Yellows (using 2V as a figure), the Green/Blues (2.5V) and Violet/White (3.5V) and series resistor calculations will be near enough for most purposes.

 

[Audioguru]

A white LED is a blue LED with a yellowish phosphor on top. A modern bright green LED uses the same technology and has the same forward voltage as a blue one.

 

[(*steve*)]

The simplest method is:

get 4 AA batteries in a holder that has a 9V style battery clip

Then get a 9V battery clip with leads and solder a 330 ohm resistor to one lead (apply some heatsink to insulate the join leaving a small party of the free end of the resistor exposed if you wish).

Set your multimeter to a volts range and hold the ends of the wires from this battery and resistor combo onto the probes of your meter (the meter should read about 6 V or a little higher.

While holding the wires against the probes, probe the leads of the LED. When the LED lights, read the voltage on the multimeter.

There voltage from the battery is low enough that connecting it to the LED the wrong way around should not cause damage.

 

[Terry01]

If you must parallel LEDs then do what Chinese flashlight manufacturers do: buy thousands of LEDs and hire a kid to test them all and sort them into groups that have the same forward voltage. But each parallel group must be fed a current, not a voltage.

Sorry for sounding thick but why the current and not volts? I've thought about it and 0 volts won't create any current and 0 current won't allow even a single mV through. Don't you need both?

I'm not questioning your answer,I know you'll be correct. I want to learn so could you explain that part for me please? Thanks for taking the time to share your knowledge.

 

[Terry01]

You can do it 'rule of thumb' using just the colour, from around 1.2V for red to 3.6V for white.

View attachment 36879

But I tend to break them down into the Red/Yellows (using 2V as a figure), the Green/Blues (2.5V) and Violet/White (3.5V) and series resistor calculations will be near enough for most purposes.

Thanks for that. Easy to study and remember or check back.

 

[(*steve*)]

Audioguru is right. If there is no series resistor then you shouldn't apply a voltage, no matter what the arrangement.

However, if LEDs are in parallel strings, then each string should have a series resistor. With this it matters little whether you apply a voltage or a current, but in this configuration it is more typical to apply a voltage.

 

[Terry01]

Cool thanks Steve.

I'd got as far as you need the resistor in line as most Dc V PSUs are even a tiny bit unstable (a little Ac V too) and even a small change in voltage can cause a huge change in the current (proportional to a LEDs) so the resistor keeps the volts steady and so more steady current. We're dealing with small V and A with LEDs so even a small change in the current is huge enough to pop the LED. Magic smoke?

Is this correct? I just like to check with the guys who know best.

 

[(*steve*)]

You can assume that a very small change in voltage across a LED will result in a very large (and not easily calculable) change in current.

For example, if a string of LEDs has 20mA flowing through it, and the voltage across them is 10V, an increase in this voltage of 0.5V would increase the current, but it's not easy to say by how much. It might be to 40mA, 100mA, or maybe higher. If the LED has a maximum current of 30mA, the smoke might escape.

Having a resistor in series will cause the change in current to be proportional to the voltage across the series resistor. Thus, a change in voltage of x volts will result in a change in current of x/r amps.

As an example, if a 100Ω resistor has 2 volts across it (with say 10V across the series LEDs), and the input voltage rises by 0.5V, the current will increase from 20mA to 25mA. If the LEDs are rated for a max of 30mA, they'll still be fine.

The key issue is that the voltage across the resistor should be large compared to fluctuations in the supply voltage. This will allow for a relatively constant current even allowing for small changes in the supply voltage or the voltage across the LEDs (which can vary with temperature).

 

[kellys_eye]

Unless I'm looking to match LED brilliance I tend not to bother calculating values of resistance for LEDs but instead opt for 1kΩ every time, every colour and for every voltage from 5V up to 24V.

At 5V this means passing 3mA (ish) in a red LED and 2mA in a white LED (and all others, some intermediate value in between). Doesn't sound like a lot but it WILL light them enough for most applications. If brilliance becomes a real issue then I slap in a 470Ω - usually fixes it!

At 24V this means passing 22mA in a red LED and 20mA in most other colours. Within the usual current limitation of most standard LEDs too.

Saves poncing about calculating and fitting 'ideal' resistors. I'm 'lazy' that way....

Of course, this only applies to the majority panel-sized LEDs/indicators and not your multi-die arrangements.

 

[(*steve*)]

For a lot of LEDs I use, a couple of mA is plenty. Like you, I don't bother much with calculations most of the time because I'm not going for maximum brightness. A couple of days ago I wanted to have the same brightness for two LEDs, one powered from 24V, the other from 5V -- The rough calculation was LED is about 2V, so about 7x the resistance for the one operating at 24V.

I ended up using 5k1 and 22k because those were the SMD resistors I happened to pull out (3k3 ones were hiding). Looked OK, close enough. And yeah even lower current than I would normally use, but these are 0805 LEDs and they are a pretty bright point source even at low currents.

 

[kellys_eye]

50Ω/volt should do.... (rule of thumb for more accuracy if it's an issue)

250Ω (nearest E12 being 220Ω) @ 5V
600Ω (nearest E12 being 560Ω) @ 12vdc
1200Ω (spot on E12) @ 24V

 

[Audioguru]

Sorry for sounding thick but why the current and not volts? I've thought about it and 0 volts won't create any current and 0 current won't allow even a single mV through. Don't you need both?

You must never apply a voltage to an LED because it is a diode, not a resistive light bulb. A small increase in the voltage will cause a very high increase in the current which will burn out the LED immediately. Besides each LED has a different voltage and the required voltage changes when the temperature changes. If you feed an LED a current then it works fine.

No current will flow if the voltage is too low. The LED needs a certain voltage plus the current-setting device (maybe a series resistor?) also needs an additional voltage.

 

[Audioguru]

Audioguru is wrong. If there is no series resistor then you shouldn't apply a voltage, no matter what the arrangement.

I did not say that a resistor is not needed. Of course a voltage should never be applied to an LED.

 

[(*steve*)]

I did not say that a resistor is not needed. Of course a voltage should never be applied to an LED.

Sorry about that. I read something completely different to what you wrote. I'll go back and edit my post.