If you want a TL072 lookalike that goes full rail-to-rail, look at TL972. Also, (your remarks about using it as an FET driver brought this to mind): don´t use an opamp as a comparator. Even the ones where it won't damage the device (many opamps have a low limit on differential input voltage) aren't really built for it. They often recover really slowly from clipping. For a comparator, in addition to the classic comparators (331, 393, 339, plus low-voltage versions), there's the CMOS TLV3701/3702/3704, (or LMV3701/2/4). For higher performance look at TS391 (pretty much a better 311), TS3021 (full rail-to-rail inputs, 8 ns rise/fall time), LT1016 (ultrafast, low input offset . Somewhere in my notes, I've got a reference to a datasheet for a comparator that can tolerate either input going beyond the positive rail as long as it doesn't exceed 36V above the negative rail, and promises no phase inversion as long as it's just one input out of bounds, but I can't find it right now. With almost all comparators, you have to be careful never to drive either input pin more than about 200 mV below the negative supply. The construction of the chip often has a parasitic SCR that will latch up if you do. Even a Schottky clamp is often not good enough, particularly with a low-impedance source. Sometimes you just have to resort to a clamping diode with a voltage divider after it (and put up with the noise of the resistors. Life is full of tradeoffs.)

Thank you for this video! I had no idea this existed. I am working on a small project that needs an op amp and have had to use a buck boost with a +/- output and it seemed like a lot of hoop jumping for a buffer / voltage follower capable of going all the way to ground. Just ordered some! Thanks again!

Very informative 👍 Great video!

Thanks to you,i will switch MCP602 for TLC272.

Useful info. Thanks

At 2:46 It's not starting to "oscillate" - it just looks that way because the extra edges produce false triggering on the scope. It's known as phase inversion and occurs because one the transistors in the signal path has become biased in such a way that its collector-base is now forward conducting. That results in an inverting stage becoming a simple pass-through and it inverts the phase of the output compared with normal operation.

Not bad for such an old design, 1987 according to the datasheet. And it (TLC272) seems to work all the way down to 3V as well. I still have a few, lets see what I can do with them.