I don't see how this would work.
You start with a 60Hz signal, and use it to drive a frequency divider so as to pulse an LED, say at 1Hz.
Now you take 3 of these 60Hz signals, displaced in _time_ by about 5.5mS. You take each of these signals and run it through a circuit as described above. Won't the output be three LEDs, each pulsing 1x per second, displaced in time by 5.5mS?
-Jon
As each phase peaks it turns on the respective LED. So you get three LEDs blinking in the same sequence as each phase.
Here is a very simple crude frequency divider. If you take the source voltage and reduce it, you could use 3 of the following circuits to run 3 LED's. Then you could visually see the the three pase sequence. You would need to reduce the frequency more than what is shown by altering the values of R1 and C1.
Why a pause?
It is my understanding that this would be a crude way but the OP did ask for a DIY circuit.
I believe the phase angles would be roughly maintained. The blinking LED would correspond to the phase its connected to telling you when its peaked in relation to the other 2.
I understand what you are saying and I think the 555 circuit would do what you describe, however the 4017 is actually a decade counter. So by using only 1 of the LEDs as an output you will have effectively created a divide by 10 circuit. I don't believe The 4017 circuit would not have the issues you described with the 555.
The U2 is a 4017 decade counter.
So where are the other phases and how do we get a rotation indication?Pin 14 on the 4017 is the clock or frequency input. Everything that is connected to pin 14 can go away and be replaced by one of the phases.
You would need three of these circuits. One for each phase. Your would have to figure the rotation out yourself by looking at which LED is on and what phase its connected to. I don't think it would pose the same problem that the 555 circuit would.