I was looking for an adjustable oscillator which I could use as a frequency source for my Tesla coil. I know my Tesla coil happens to resonate at about 2mhz because I had previously built a slayer-exciter circuit and measured its frequency.
My first go-to choice for a simple oscillator is a 555 timer, however, on constructing the 555 oscillator I found that the 2mhz was a little too fast for the poor little chip. I believe you can get faster CMOS 555 timers that will handle this, but I didn’t have any in stock.
As an alternative I considered using an astable multivibrator:
The multi-vibrator seemed like it would do the job, and it would definitively be be fast. Unfortunately, to adjust its frequency one needs to adjust the value of R1 and R2 simultaneously, this is a bit of a pain as I don’t have double-ganged potentiometers handy. To add insult to injury, the above circuit doesn’t produce sharp edges without adding a few extra diodes; this all seems like quite a lot of work to produce an adjustable square wave.
Schmitt Inverter To The Rescue
While browsing the intertubes I came across an incredibly simple oscillator
This looks great because it can produce square-waves with only 3 components and you can adjust the frequency by varying a single resistor. You can find specific details of how this oscillator works here:
You can also watch a video about schmitt triggers here:
I built the circuit using the Texas Instruments SN74AC14N — I happened to have spares from a previous circuit. The SN74AC14N is interesting in that it has a very short switching time of 1.5-11 nanoseconds, so we’d expect to produce a square wave with pretty sharp edges, even at my desired 2mhz.
The above video demonstrates winding the oscillator through a fairly large frequency range by adjusting R1. I’m pretty happy with this. The oscillator produces clean square waves over a very large frequency range. I managed to make this oscillator work up to around 4 or 5mhz (as above) and down low enough to blink a LED, about 0.3Hz. You also get six inverters in one of these chips, so you could make six oscillators if you desired, or use some for something else. The SN74AC14N is also cheap chip (60 cents on RS components).
I ran into an issue whereby at power-on the oscillator wouldn’t start reliably. Adding a largeish resistor (50k) in series with the capacitor should help the oscillator start by pulling the inverters input to ground, and thus sending the output high to kick-start the oscillator…However, in my case, no dice
A solution which solved the problem entirely was to run the oscillator on 3.3V rather than 5V, which happened to be fine for my application.