Hi again,

Oh so one LED had blown out then? I guess we have to assume something is going wrong then.

I have another experiment to add to the mix now. This is with a 5mm red LED and 50k resistor.
In the forward direction the LED lights up with 30vdc applied to the string, not sure of the brightness as i was talking to someone over the phone and they were doing the experiment so i didnt see the LED myself. Next the LED was reversed, then 10vdc, 20vdc, and 30vdc was applied and a few in between voltages like 15v and 25v. It was clear that in reverse, the LED voltage was 'zenering' at around 14.5v and changed only a little like 14.7v with 30vdc applied. So clearly this LED "zenered" with reverse voltage.
Now on reversing the LED again so it was back to the normal forward biasing, he reports (over the phone) that the LED looks dimmer than before, but it's hard to be sure, although he seems to think it is much dimmer. On changing the resistor to 220 ohms and applying 5vdc however, the report was that it looks like it has full brightness. So it is questionable if the LED was damaged or not and it seems like maybe it was not, but the LED will be tested again in a week or so. What is not questionable though is the zener behavior in reverse for this particular LED.

So the conclusion must be that not all LEDs are going to behave the same way in reverse based on this experiment and your experiment.
It is still too soon to tell if the LED used in this experiment was damaged.

One other thing is also clear though. That is that the manufacturers usually specify a maximum reverse voltage of 5v, and that means if we go over that voltage spec then we go out on a limb by assuming that our knowledge of the LED behavior in reverse is better than the manufacturers knowledge of the LED in reverse. The way i see it is they must have a reason for specifying it so low.

So now when i go back to the original problem about the LED in the relay, i have to think that the best logical conclusion is that even if it does work i would still be taking a chance that it will blow out sooner than most LED's at low current would blow out, and i am not comfortable with that.
Adding a diode across the LED is not an option unless the resistor value is increased or the power rating increased. Adding a diode in series would still take the chance that there could be some reverse conduction. Adding one in series and one in parallel would be best, but only one in parallel would only require raising the value of the resistor and then the brightness could be affected but possibly it wont be too bad.

Also, now that i think about it i think i only have 3mm LEDs not any 2mm LEDs, and this one is 2mm

 

so LEDs can be used as zener diodes, capacity diode, light sensor, solar cell

 

so LEDs can be used as zener diodes, capacity diode, light sensor, solar cell

Hello there,

Well i would not go so far as to say that and LED can act like a zener diode in reverse, but in forward bias it would be ok but the stability would have to be investigated because i think the temperature variation effect on the voltage could be a little high.
In reverse it is not supposed to be used like that, that's why i often use quotes around the word "zenering". The data sheets usually recommend 5v max in reverse.

I've also used glass diodes as solar cells, but very limited current output.