Here's one for Bryce lol!

[Badstarr]

Whilst having one of my random looks around on that newfangled "interweb" all the kids are talking about these days, I found this ...
Series circuit - 3 LEDs - YouTube . It is 2:15 am, and I'm rather tired, but I think I witnessed a miracle ! Well ok not really, how the hell did they do that, pretty clever lol! So over to you Bryce, how they do it?

[Ynot.zer0]

It's that devil / black magic surely?!    actually I am curious.  that defies logic.... albeit we don't know what is 'inside' the switches and I'm curious why the need for the 9v battery... oh masterful merlin/bryce please enlighten us with your wisdom 

[Bryce]

Very neat trick. I suspect that the switches have a resistor in parallel inside. LEDs should never be connected in series, because the LED with the lowest internal resistance will always light first and if the current is not enough for all the LEDs, the voltage drop across the first lit LED will stop the others from lighting. Here's a diagram of what I think is really happening.

When all switches are open, the resistance (R1 + R2 + R3 +R4) is too high for any of them to light. If SW1 is turned on R1 is then shorted so the total resistance (R2 + R3 +R4) supplies enough current to light one LED - The one with the lowest trigger voltage. The resistor in SW3 must be slightly lower that the resistor in SW1, because switching that triggered a different LED. The current is slightly higher, so the LED with the second highest trigger voltage will turn on, blocking the other two. SW2 has the lowest value resistor and turns on the LED with the highest trigger voltage. It looks cool, but it's also pretty random and you would need to hand pick the LEDs and match them to the resistors in the switches.

You can test this with two LEDs in series with a fixed resistor (about 330R if you use a 9V battery) and a variable resistor. Start with the variable resistor at its highest resistance and turn it slowly until the first LED lights. If you now turn it further down, the current rises and the second LED will come on at some stage, but it will turn off the first LED. If you then continue turning, you will eventually have enough current to supply both LEDs and they both light.

Bryce.

Edit: Sorry for spoiling the "magic".

[TFM]

Very neat trick. I suspect that the switches have a resistor in parallel inside. LEDs should never be connected in series, because the LED with the lowest internal resistance will always light first and if the current is not enough for all the LEDs, the voltage drop across the first lit LED will stop the others from lighting.

But he can select the single LED to be switched on. How can this be explained?

[Devilmarkus]

Bryce already explained this

[Bryce]

But he can select the single LED to be switched on. How can this be explained?

Because the resistance of the resistors in any two switches produces the required current to light one particular LED. So shorting one switch will result in one LED lighting.

Bryce.

[Devilmarkus]

LED's behave different to normal light bulbs...
They behave like a diode. TV's (CRT) use diodes in serial order to produce very high voltages. (Combined with capacitators)

[Bryce]

That's called cascading and can be pretty dangerous if you don't know what you're doing. I made a small circuit to get 180V out of the 9VAC on a C64 expansion slot once. Really low amps but very high voltage.

Bryce.

[Badstarr]

A great solution Bryce! Have you by any chance had a look at the comments posted on the youtube page for the video? There is one very complicated solution involving radio controlled stuff hidden in the battery! I think i prefer your circuit much easier to impliment and uses the "quirks" of the components to do some unexpected stuff.

[TFM]

Bryce already explained this

No, but you obviously haven't read thoroughly.

Because the resistance of the resistors in any two switches produces the required current to light one particular LED. So shorting one switch will result in one LED lighting.

Bryce.

Well, that would work well with two LEDs, but not with a third one!?! The third LED must jump active at a current to low AND to high for both the other LEDs at the same time. (I could explain it better in German). However, IMO there must be a third level of reguation, independent of resistors.

I don't have to parts here to rebuild such a circuit though.

[Bryce]

@TFM: That is exactly what it does - Three different levels, one for each LED just using resistors. This will work with 3 or 4 or even 5 LEDs, just it will keep getting more difficult to match the resistors to the LEDs and the LEDs are more difficult to hand pick. When the "middle" (in terms of current) LED turns on, in reality both the lower and middle LED turned on, but the middle one pulls all the current and immediately turns off the lower LED. You're in a university aren't you. Show it to any of the guys studying physics and they might be able to explain it better or even build you one

Bryce.

[TFM]

LED's behave different to normal light bulbs...
They behave like a diode. TV's (CRT) use diodes in serial order to produce very high voltages. (Combined with capacitators)

That's basic knowledge Markus, we're not in elementary or in south park *hehehe*.

[TFM]

@TFM: That is exactly what it does - Three different levels, one for each LED just using resistors. This will work with 3 or 4 or even 5 LEDs, just it will keep getting more difficult to match the resistors to the LEDs and the LEDs are more difficult to hand pick. When the "middle" (in terms of current) LED turns on, in reality both the lower and middle LED turned on, but the middle one pulls all the current and immediately turns off the lower LED. You're in a university aren't you. Show it to any of the guys studying physics and they might be able to explain it better or even build you one

Bryce.

Thanks' for the explanation. From own experiments I did only see (decades ago!) that you can "exclude" LED from burning by using too high or low values. But I didn't know that you can actually do this also in more than two (upper + lower) levels. That point is new to me.

BTW: I studied Physics by myself for four semesters, in Germany it part of the basic studies for Biology. So I'm not a novice in these things. Having said this circuit design is not my profession though