Quick one today… I didn’t even bother taking pictures of the end result since it wasn’t much to look at.

I was looking around the Cypress (semiconductors, not the plant or the country Cyprus) web site and saw a little circuit:

The cool thing is I’m starting to understand this!  :-D

First off you have a level-shifter in the form of a capacitor C1. The right side of the capacitor is being held at half-rail by the high-impedance connection (I used a 1MΩ resistor since that’s what I had in stock) to a voltage divider formed by R2 and R3. (I used 150Ω again… stock)

Here’s where I was proud of myself: I looked at the first op amp (top blue triangle) and thought “voltage buffer.” I looked it up and I was right! Then I looked at the bottom op amp and thought “hmmm… op amp as comparator with a bit of hysteresis.” The + side is connected to a voltage divider (R4, R5) and there’s minimal feedback from R6. (I used 1KΩ for R4 and R5 and 10KΩ for R6)

Again, it worked out.

I built up the circuit (skipping the first stage op amp hypothesizing that it’s not needed since the op amp already presents a high-impedance load on it’s input) and… Whoa! It worked! At first I used a quad op-amp that was rated to 1MHz and sure enough, I was able to push perhaps 50KHz though it before it was slew-rate limited in making something that looked like a square wave. I replaced it with another 160MHz rated one and I was able to get around 10MHz before my shoddy breadboarding’s poor impedance matching was starting to wreak havoc with the signal integrity.

This isn’t rocket science to me any more. Not just that but I was able to just whip it together with parts I had in stock and measure it and drive it with instruments I had on my bench already.

So freaking neat!  :-D

It’s one thing to read about this, but it really does seem a lot more tangible when you’re looking at it and, quite literally, poking at it.