What’s an example of where you picked the wrong Op-Amp … and why?

Usually common-mode voltage value : )

Especially with those irritating very-old TL071 etc JFET op-amps. It's easy to forget, because it's nowadays so unexpected to have such a large restriction, especially at the negative end of the rail (4V).

It randomly bites me every few years, because I've got a few of those for random prototyping. I should throw them out, they waste a lot of time out of proportion to their low-costness.

Regarding the '741, yes : ) It was used for a very high-tech purpose, it was a key component:  A Westworld / Futureworld Halloween  

I've been caught out trying to use op-amp as a comparator, although it sometimes works well enough. I have some comparators in my parts bin now.  I'd be interested in seeing a small recommended list of parts to have on hand.  I tend to go to the tables in Horowitz in Hill when trying to narrow a search.

I am currently trying to decode USB-PD CC communication, which is a ~1.6V signal. To sample the signal with a logic analyzer I wanted to amplify it to a high level of 3.3V.
My first choice with a component on hand, was the Microchip MIC6271. It is rail to rail and has a 2 MHz gain bandwidth and a slew rate of 0.9V/usec. PD messages are sent at 300 KHz +-10% with rise and fall time of measured 600 nsec. Using the rise time for calculation, this signal is already beyond the limit of the opamp capabilities. 600 nsec rise time corresponds to 1.667 MHz and a gain of 2 calculates to 3.3 MHz gain-bandwidth already, leaving no headroom to the max GBW of the opamp. Also the slew rate of the signal is ~2.6V/usec measured, exceeding the opamp spec. This adds up to a amplified signal not representing the input signal at all. See image, blue channel 2 is the CC signal; yellow channel 1 is the amplified signal.

So, a different opamp with >10 MHz GBW and >3 V/usec slew rate is on order. I think a good rule of thumb is a GBW 5 times as the edge frequency.
Unfortunately, I have nothing in the parts bin, so this route needs to wait to be explored further.

Edit: The opamp was used in a non-inverting configuration with a gain of 2; feedback and negative to GND resistors had equal values.

Fun fact, slew rate / GBW limits were the driving idea behind this video.

I saw a very similar waveform for a very similar reason. I should have known, but I didn't even consider the limitation until I saw the output.

The other driving factor was mentioned too: Common-Mode. I have a measurement where I completely misinterpreted the datasheet and tried, for more hours than I should admit, to prove it was wrong. My friend looked at my measurements and said: "yeah, your op-amp is limited to X." He was absolutely right. (I have to go back and get the details on that one.)

I'm familiar with their table. However, I haven't checked the status of the 3rd edition. Are the parts still readily available? (Whatever that means in 2022...)

There is no substitute for risking the cross eyes by reading data sheets.

There are loads of op amps and they are all different.

There are a great mass of applications where the choice of op amp is quite flexible, but others are less so.

In the present times of all sorts of parts being on long delivery it's even worse because availablity trumps everything else !

To give an example, I have a current job where the key parameters are:

low voltage noise (< 3nV rtHz @ 5kHz)

work from 2.7V

input common mode must include 0V

rail to rail out

> 50MHz GBW (200 would be nice)

low input voltage offset

less than 3mA supply current

There are not many options !

I can't think of any application where a 741 would be a good choice unless you have a tube of them gathering dust.

The worst things about it, compared with modern op amps, are the limited input common mode range and limited output swing which means that the minimum supply voltage is 6V or +/- 3V (more realistically 10V or +/-5V).

It is cheap.

TI recommend the TLV9301, it's quite nice but not as cheap !

MK

I haven’t ordered op-amps since early in the year but there were unobtainable parts back then, at least from Newark. Horowitz and Hill list several in every category and I was able to find at least something where I was interested. And of course some manufacturers have recommended alternatives on their sites for older parts that are hard to find. 

Other that already mentioned bandwidth and slew rate, the one characteristic that once catched me unprepared was the internal compensation of the opamp. I was in lab at university and I had the opportunity to play with a quite fast opamp, so I went for it. But it wasn't single pole compensated! I did not notice it in the datasheet so my mic preamplifier turned out to be a "nice" oscillator :)

When designing an electronic load here on the community, the first opamp we selected was hard to control near the 0 point. Discussed extensively.