OK, so the uA741 is 42 years old now. For its time it may have been a great opamp; the requirements weren't as high as today, and there was far less competition. But I was wondering what's the 741's appeal today.
- it's slow. GBW 1MHz, slew rate < 0.5 V/us
- it's not low power, nor low voltage
- it doesn't have low bias current FET inputs
- it doesn't have rail-to-rail inputs or outputs
- it's not low noise
- many more modern opamps have comparable price
Why is the 741 still used today?
It's an ideal op amp to learn the basics on due to its non-ideal nature. The first thing we learn is infinite input impedance, infinite gain, as well as a few other silly things. The 741 obeys none of these idealities, forcing students to learn the hard way how to cope. They see bandwidth limitations without using expensive oscillators or function generators; they see early saturation, nowhere near the rails, allowing the use of cheap multimeters. Many textbooks use the 741 as an example due to its ubiquitous availability and simple verification of non-idealities.
Today, we can buy op-amps with mV offset and noise, 100s MHz bandwidth, nA leakage, etc.. One of the most time consuming part of a design is looking for parts, especially for the inexperienced. Academics aren't experienced design engineers, and will use the parts they know, as they have better things to do than look for parts (like write that grant application, right? :). This outdated part therefor gets introduced into new designs from copying legacy modular designs, and familiarity from instruction.
Many old designs are still around. Plus, some positives are
1) It is readily available from multiple sources (ST, TI, National) which (having multiple sources) can be a big issue for certain industries.
2) It has been around for a while, is well understand, reliable, and will most likely will continual to be available for a while, again, very important for long life applications.
3) It has a large voltage range, many newer op-amps don't.
4) Output short circuit protection.
5) It's slow. This is good for many applications. Why is faster always better? Having an overly fast op-amp just increases susceptibility to noise.
6) Many people know it and use it, there's something to be said about not having to evaluate, test, etc., a new chip, as well as not having to stock a new part.
7) It doesn't have FET inputs. There are pros and cons to such inputs. Certain designs may be better with them.
Is it really used per se, or simply kept around for legacy designs? Changing an op-amp in an application where those specific parameters are important (i.e. feedback loops) can be problematic at best (or outright dangerous at worst) - best to let sleeping dogs lie in these situations, sometimes...
I can speak to the fact that in both of my career stops thus far, there haven't been any new designs that I've seen or touched using the 741. For me, the LM358/LM324 is the 'go-to' part where things like input offset voltage or rail-to-rail capabilities aren't critical. It's well understood, it 'works', and it's cheap.
In many cases I've seen in online electronics forums, the 741 is designed in by beginners who just don't know of any other opamps. They may have read about it in a textbook or seen it in another old design and assumed it was a reasonable choice. Once they learn that LM324, etc. is readily available, cheap and easier to use, they'll normally switch.
The other big reason already stated is legacy designs. Something works, is still in production, and there's no compelling reason to change it. If you've been selling a product for 30 years and you won't run out of parts, and you won't make any more money by changing to a newer opamp, they why change?
My guess: people tend to stick with what they know. If you've learned the errata and gotchas of the 741 and it works for your application you'll use it rather than learn something new for no reason. Also, my guess is that many applications don't require terribly high performance, so the 741 works just fine.
