# Explain how this astable chain(?) circuit works

The circuit below fades on and off a cascading series of leds in an alternating fashion.

I tried reproducing it on a breadboard with 5 leds to try to understand it how it works, and got all sorts of weird results.

• The resistor values don't seem to matter that much
• removing a capacitor doesn't seem to affect the circuit at all. It seems to be using the total or average of all of the capacitors. the capacitor values seem to affect it in some non-linear way that I can't figure out.
• it only works with an odd number of leds. 3, 5, 7, etc.

Does this circuit have a name? What determines the brightness and periods of the leds?

Found here https://boldport.com/shop/ananas

This is called a phase shift oscillator where oscillation occurs using 60 of 90 deg of each cap (RC LPF) to achieve the additional 60x3= 180 deg needed to get positive AC feedback. This is why an odd number of stages is needed.

This design, however, filters each base current so there is not much variation in LED intensity. It would be far brighter to move the cap from the base to the collector.

My improvements Here

Since the load is nonlinear LED, the oscillator is not sinusoidal.

The result is actually 2 of 3 LEDs ON with a moving “black hole to the left” as the cascade moves to the right.

5 stages would share 180 deg to shift at 180/5 deg per stage.

It's basically a form of "Ring Oscillator". Not normally used in discrete form but common in ICs with an odd number of inverters.

Rise times and fall times tend to be slow as there is no positive feedback.

Even numbers of stages won't work because they will latch up stably with alternate stages in opposite states.

• I wonder if this might be better seen as a low-pass phase shift oscillator with 3 RC stages and an inverter (three inverter stages still makes up a single inverter, so to speak.) I get the ring oscillator point, though.
– jonk
May 7, 2019 at 4:06
• @jonk - True. But the amplitude builds until all stages operate in saturation. May 7, 2019 at 11:18
• Good point. And the switch behavior means one side of the cycle is active, so faster oscillation than expected from a phase shift. Point taken.
– jonk
May 7, 2019 at 16:22