when the LED current is low, wavelength is different, and efficiency is higher. what causes this small shift?

Since you know the topics maybe you can answer it?

Mod edit: created from
http://forum.allaboutcircuits.com/t...ry-high-frequencies-and-lower-current.124615/
Please don't hijack another's thread.

 

when the LED current is low, wavelength is different, and efficiency is higher. what causes this small shift?

LED efficiency increases with current until current crowding starts to decrease efficiency.

The change in wavelength is likely due to temperature. Emitted wavelength can vary by 0.1-0.2nm/°C.

 

LED efficiency increases with current until current crowding starts to decrease efficiency.

The change in wavelength is likely due to temperature. Emitted wavelength can vary by 0.1-0.2nm/°C.

So red 660nm leds can be 648nm when hot? I've made an assembly with lead accumulator precharged bridge from desktop 12v Booster and 2 smaller downconverters.

Today I've reduced the red chip voltage to very low value to run on battery. But on main 12v supply boosted to 21 or 22v. The brightness and heat increase is very abrupt after 22v but then the current limit kicks in at some stage. It can't exceed except 22,2v in a test

 

I remember that post. I think it is really rude to ask about a circuit anomaly, and not state that the anomaly is in a simulation. Imagine that. That means none of the posted answers were relative.
He doesn't even recognize the problem. When his simulation fails, he thinks it's a physical anomaly.
Software can only do instructed math.

As for your question, there are gobs of information on diode dynamics. You can study about your particular one, or the theory in general.
The physical dynamics of your diode will be based on the standard model. Which is an instructed mathematical simulation.

"Just when I thought I was out...they pull me back in."

dl324 has a reasonable explanation.

On the classical side, I don't think the emitted light comes from a particle. I think it comes from the bonding field, who's frequency is determined by the bonding length. When a valance electron leaves a vacancy in an atom, the vacant area still has the bonding field from the other particles. When an electron fills that vacancy, the bonding field must be rippled thru temporarily, to insert the new electron into the field. Some of the old bonding field energy or some of the new bonding field energy or both, is released on electron insertion. The inter atomic and molecular bonding lengths would explain the infrared.

In any case, temperature can change the length of the bonding field.

An external electric, magnetic, gravity field or acceleration can effect that length. An external appropriate "light" frequency, can pop an electron right out of there, by disturbing the bonding frequency. Using frequency to dismantle an atom or matter, would be much more efficient than smashing them. And you don't lose any parts. All we need is a fast switching technique. Perhaps beating multiple lasers to get started. There's a difference between heating an atom and shaking it.

Modern science says the electron emits the light. But I have not heard of a modern dynamic of emission.
In which we would need to know, to understand the frequency change.

 

So red 660nm leds can be 648nm when hot?

The wavelength increases with temperature. It's more interesting to experiment with green or yellow and push them towards orange.

 

I understand they can change light frequency with sound now on these chips.

Maybe that will be incorporated in LEDs. Add a tone terminal or possibly modulate the supply to change color.