Back in the late 70s, when I was in high school, I used to build these and give them to girls I had a liking for. I even managed to get a date or two from some of the recipients. Heh, the life of a high school science geek. Well, anyway...
The first few specimens I built used three 555 IC oscillators slaved to each other. After those first few I'd made of 555 chips I started thinking about how to make an equivalent circuit out of fewer, cheaper, more clever parts. Like 4xxx series CMOS inverter gates. After all, almost anyone can build a three-oscillator circuit using off-the-shelf oscillator IC chips (or a microcontroller, for the more sophisticated).
Remove these ads by
Signing UpStep 1: Theory (such as it is) - waveforms
Why CMOS? High circuit impedance and very low current requirements lead to long battery life and, as we'll see below, lower parts cost and better "character" in the finished toy. Why CMOS 4xxx series? Because they can be run from a wide voltage supply: from 3 to 18 Volts -- just the thing for a toy using a 9V battery as a supply.
Aside from its extremely low current requirements digital CMOS provides a few other advantages in this analog application. First, the extremely high impedance of CMOS circuits leads to increased "noise susceptibility" -- the opposite of noise immunity -- which is something that we want in this application. Nose susceptibility adds character to the Tribble. The extremely high impedance and low internal leakage of CMOS also allows us to use small capacitor values and high resistor values in our RC networks. Resistors (of a given precision and thermal dissipation capacity) generally cost the same regardless of their Ohmic value. The same is not true of capacitors; the higher capacitances tend to cost more, and be of larger physical size, than the lower ones.
But perhaps the main attraction of using CMOS gates is the satisfaction of knowing that you are using digital logic blocks to do an analog task. A corruption of logic ... so to speak.
Some of you may have seen oscillator circuits that use simple inverters, and you could certainly build our three oscillators from a single hex inverter IC, but then you would still need a way of multiplying (AND-ing) the outputs together. Worse, since the inverter circuits require two gates per oscillator they are not the most clever, parsimonious way to build this particular toy. Therefore, the IC that I've selected for this version of the Tribble is the 4093B quad 2-input NAND Schmitt Trigger. It can provide all three oscillators as well as do the multiply/mixing of the signals.
(Actually, I would have used a hex Schmitt Trigger inverter, the 40106B, if I could have found one at the store. As I'll show below an oscillator using a Schmitt Trigger needs only a single inverter. Perhaps I'll do that one another day because it is even more interesting than the current design as an example of Micky Mouse logic and applied bullshit.)
Figure 1 shows the voltage vs. time graph of the three signals we want to mix (and the mixed result). Red traces are the RC voltages. Blue traces are the outputs from the "inverters" (actually NAND Schmitt Trigers) of the stages.
Figure 2 shows the three signals multiplied (AND-ed) together.
Note that the frequency ratios in the figures are not to scale.

















































Visit Our Store »
Go Pro Today »




Tuning is managed by means of capacitance alone, or a combination of capacitance and resistance? Seems they should be interdependent values and some alteration might be achieved by means of preset variable resistors.
I have a slightly different project in mind and something that makes wibbly-wobbly noises at different tones and frequencies would be fun.
However, it is best to go with small C and big R for a given T because the cost of a resistor is basically independent of its resistance value while the cost of capacitors generally is higher for higher values of capacitance. Also, it is better to use low precision resistors (20%, if they are even still made) for this project because they are both cheaper and have more "character".
None of the electronics in this project are intended to be precision in any way. After all, we don't want Tribble clones, do we? (Yeah, yeah, they are born pregnant, I know. But they weren't all identical in the show.)
Another more fanciful related toy is the Snailbot (original by Evil Mad Scientist).
I've also made a more complicated Tribble using a Qprox touch sensor chip and a wire antenna sewn to the inside of the fur covering. The touch sensor works, I believe, similarly to your variable capacitor (my understanding is that it's a capacitance sensor), and I used the same tiny vibrator motor which in this case is triggered when the Tribble is petted. I don't have an instructable on that as I only made one so far and it was kind of sloppy; I took it apart to see what I'd done so I could make another, and that's where THAT project is at just now. I do plan to write it up when I make the next one.
I will probably have to make yours, too, just to round out my Tribble-making powers (I will post a picture). The instructions look super clear and detailed! It will also help improve my electronics knowledge which is still pretty small.
Heeheehee. Great job on this. It's how electronic instructables should be done.