If you’ve read The Magic Theater, Book 3 of Jack’s Ravenwood: A Seeker’s Memoir, you’ll remember Nick Valentine’s “Six-Way” and the uses to which it was put.
This is a re-creation of my 1993 design, which in turn was an update of Nick’s original one using tech 20 years more recent. I built this originally on Vectorboard but that one’s long gone, probably scavenged for parts. This past month I’ve been noodling with it on a solderless breadboard and gotten the basic functions working again. It still needs work, let alone development into a DIY kit. But in the meantime, for you experimenters out there, here’s what I’ve got:
Sorry, the image came out a little blurry but I hope you can read it. R1 = 100 ohms, R2 = 270K, R3 = 15 ohms, C1 = 22 uF, C2 = .01 uF. Q1 = 2N3904, Q2 = 2N3906. R3 through R9 are all 1K. The LEDs are any generic kind and color you want to use. My original design used red ones, but since they’re all so cheap now it’s up to you. Even T1-3/4 size is optional, they just fit best on the breadboard I’m using.
It all runs off a common 9-volt battery. Turn it on and one of the six LEDs lights up. Press the button “PB” and all six start to glow more dimly. Actually just one is lit at any instant, but they change too fast to see. After a few seconds the light seems to start “jumping” from each to the next, slower and slower until it comes to a stop. Just like rolling a six-sided die, it has one chance in six of stopping on any particular one. It’ll then stay that way until you press PB again and the whole thing restarts, like rolling the die again.
What’s happening? PB charges C1 which then discharges through R2 into a pulse oscillator formed by C2, R3 and the two transistors. U1 is a 4017 decade counter with its first six outputs Q0 through Q5 driving LEDs while the seventh Q6 is connected to its own reset pin. Each pulse makes the output switch from one LED to the next: Q0 to Q1, Q1 to Q2 and so forth. When it gets to Q6 the counter resets to Q0 and the count starts over. And over. And over.
But as all this is happening, R2 is bleeding down the voltage on C1. As it does, the current through R2 drops and so does the frequency of the pulses. After a while it’s slow enough to see the “jumps,” and a little while later it stops completely with just one LED lit.
Nick had a crude analog “memory” to count the number of times the pulse stopped on each output, read using a voltmeter. In my rework I scrapped that, replacing the analog parts with digital ones, but at this late date I can’t find my notes or remember just how I got it to work. I know some 4520 dual up-counters did most of the work, but so far they’re not doing it right. Oh, well. Back to the drawing board…
Watch this space!