First of all, we’d like to give a big shout-out to [Afrotechmods]! After a long hiatus, he has returned to YouTube with an awesome new video all about op-amp characteristics, looking at the relatively awful LM741 in particular. His particular way of explaining things has definitely helped many electronics newbies to learn new concepts quickly!
Operational amplifiers have been around for a long time. The uA741, now commonly known as the LM741, was indeed an incredible piece of technology when it was released. It was extremely popular through the 1970s and onward as it saved designers the chore of designing a discrete amplifier. Simply add a few external components, and you have a well-behaved amplifier.
But as the years went on, many new and greatly improved op-amps have been developed, but either because of nostalgia or reticence, many in the field (especially, it seems, professors teaching electronics) have continued to use the LM741 in examples and projects. This is despite its many shortcomings:
- Large input offset voltage
- Large input offset current
- Low gain-bandwidth product
- Miserable slew rate of only 0.5V/uS
And that’s not even the full list. Newer designs have vastly improved all of these parameters, often by orders of magnitude, yet the LM741 still appears in articles aimed at those new to electronics, even in 2025. There are literal drop-in replacements for the LM741, such as the TLC081 (not to be confused with the similarly named FET-input TL081), which has 32 times the slew rate, 10 times the gain-bandwidth product, and an input offset voltage almost 2 orders of magnitude better!
So, check out the full video below, learn about op-amp parameters, and start checking out modern op-amps!
Along similar lines from earlier in the year: https://hackaday.com/2025/01/06/rethinking-your-jellybean-op-amps/
That’s like asking why people learn to drive in a Ford Fiesta because the Bugatti Chiron is so much better.
Same principles apply for opamps today as they did when the 741 was shiny new and exciting so why not use a part that’s dirt cheap, incredibly well known, likely to outlive the faculty and costs pennies to teach the basics?
It’s to do with the style of instruction. If you say to people “use this part” without explaining why or what it really does, people come to understand that you have to use the part. There are now better AND cheaper parts, but people don’t know they exist or that they can actually use them.
With hobbyists it used to be a common habit to list alternative parts if one was not available, such as listing different transistors, or different models of an op-amp, but when the designs get repeated through the broken telephone of people copying people and turning things into “instructables” for beginners, that information gets dropped out.
Sure, but it’s incumbent on students to learn and have the wit to realise there’s a whole big world out there. I’d also suggest any tutor who’s not explaining that these are just principles and, at the very least, that there are much newer and better parts now is clueless and not fit to be teaching at any level past kindergarten
And why not use a 741 if it’s good enough for the job at hand?
Parts like the 741, 555, 78xx etc etc are still around because they’re still useful.
The 741 has been obsolete for years.
It sucks because better parts have existed for a long time, an obvious reason is is needs a dual rail supply. Who has that these days?
PC PSUs dropped -5v last century and -12v is so rarely used I’m surprised it’s still there.
RS-232 uses it. And I’ve had lots of fun building 741 modems for my radio hobby (HamComm modem aka HamCom modem).
(Built the last one less than 5 years ago, used it on a DOS laptop.)
In this application, the 741 does the work of a comparator or so-called data slicer.
Being able to go rail-to-rail is a good thing here.
What I’m talking about (I’m not affiliated):
https://www.discriminator.nl/2-level/index-en.html
https://www.sm0vpo.com/use/hc-00.htm
https://www.qsl.net/iz7ath/web/03_digital/03_hamc/hc_eng.htm
https://www.youtube.com/watch?v=6rYcxBbmWnM
Sure, something like an aged TL071 can work like a drop-in replacement, but there’s no guarantee it always works as intended or as good as the LM741 here.
Especially “fuzzy modes” such as Hellschreiber or SSTV are hit and miss, quality wise.
It’s best to have both at hand, I suppose.
But that’s just me. Maybe I simply have the wrong kind of hobbies and belong in a correctional camp?
For still using a 741 as such?
is that true? can’t you use ground as one of the rails, so long as your input signal stays in the center between GND and Vcc? the TI LM741 datasheet says “The LM741 can operate with a single or dual power supply voltage.” ??
it seems like it’s just like BJTs, every beginner has to develop some sort of understanding of ‘biasing’ and decide how that’s going to influence their circuits
@Greg A, you’re right in saying it does not need a plus and minus power supply.
You can, but the output voltage swing doesn’t go anywhere near V- so you’re very limited in single supply operation.
RS-232 is dead as well. And it’s been a long since it’s had a 25v swing.
Most RS-232 is TTL aka 5v these days, you might get 12v out of a PC. All those USB converters +5v, alleged some boost the voltage.
@Bob, RS-232 may be dead for consumer and office settings; but it is still very much alive in industrial settings.
It’s a great piece of history, and good luck building a discrete LM741 that performs in any way recognizably similar to the original: https://shop.evilmadscientist.com/productsmenu/762
The modern equivalents will probably make more student circuits work perfectly simply because they are so much better. However, the use case for the LM741 has now shifted to make its outdated performance a feature for education: students will run into the limitations of the hardware soon enough, and have to internalize just how messy things can get when you translate an ideal circuit into the real world.
I remember a professor telling me they had just helped a group of seniors with their project, where they had tried to get 12V from 5V purely by setting op-amp gain.
This is a terrible analogy. I don’t think there are appropriate car analogies. You could try saying “like telling people to drive a model t ford when modern consumer cars exist”, but then the problem is that the model t ford is a classic and that would be the point of driving it in and of itself.
Driving a Bugatti isn’t easier and better and cheaper than driving a Ford Fiesta.
Even if you ignore all the performance differences between the 741 and any modern jellybean op amp the immediate problem is still that the 741 doesn’t run on voltages that are commonly used nowadays. That alone makes it much much easier to use even the cheapest modern op amp with 5V or 3.3V or a battery or whatever. Which is what most projects are going to use.
I think perhaps you missed the word ‘learn’ and possibly the entire point of the analogy so let me clarify:
A Fiesta has the same basic inputs as a Chiron (brake pedal, accelerator, steering wheel) despite costing a couple of orders of magnitude less.
So you can learn the fundamentals of driving in the Fiesta, then you learn the finesse and skill necessary to drive the Chiron later.
A 741 has the same basic inputs as an esoteric modern opamp, (inverting input, non inverting input and power supply), despite costing a couple of orders of magnitude less.
So you can learn the fundamentals of driving opamps with the 741, then you learn the finesse and skill to drive the esoteric ones later.
Clearer now?
You really don’t need a Chiron to learn to drive, nor do you need more than a 741 to learn the basic functions of an opamp (which are still ‘only’ the analog computing elements, add subtract, multiply, integrate, differentiate etc.)
And, if you wreck a Fiesta or two in the process of learning to drive, no biggie. :-)
thing is, more modern op amps are NOT more expensive. they’re often LESS expensive.
When learning, seeing edge cases is a benefit.
You can use/abuse the modern op-amps more.
Hence a more contrived design will be needed to expose the students to the edge cases where the simple op-amp model fails.
Same reason students should still play with TTL.
TTL is so they understand there is no digital, it’s all analog under the hood.
Crappy op-amps so they understand their mental model is only a first order approximation.
Okay but for your analogy to work the Ford Fiesta would need to be a lot more expensive than the Chiron because the Fiesta is a niche, discontinued car only available used from collectors while the Chiron is dirt cheap and sold on every street corner.
Comparing a Ford Model T to a modern sedan would be a better analogy. A very expensive fossil versus a cheap modern commodity.
You don’t need a 741 to learn to drive. You can use a modern part that costs the same or less and is easier to drive and better in literally every way except in how good a cautionary tale it makes. I still think your analogy sucks.
A 741 is not a Fiesta. An MCP6001 is maybe a Fiesta. A 741 is at best a Trabant.
“Driving a Bugatti isn’t easier and better and cheaper than driving a Ford Fiesta.”
A laudable assertion. If that were true, nobody would buy a Bugatti, they wouldn’t make them, and we wouldn’t be having this conversation.
Damnit. A laughable assertion.
A bugatti is a toy car for rich people – it has nothing to do with being better than a regular Ford Fiesta. If anything, it’s desirable because it’s worse: you can empty the fuel tank in 10 minutes flat out, which is apparently a thing that amuses some people.
A Bugatti Veyron is basically a car that doesn’t drive well on regular roads at normal speeds, isn’t cheap or practical by orders of magnitude, but it can drive really fast in a straight line and go around specially designed closed tracks, almost killing you in the process, if you know how to do it. If you don’t, it will kill you, or failing that, lose you a couple million dollars in repair costs.
It has the same point as jumping off a helicopter, with your feet strapped to a snowboard, over the alps. You’re not doing it to go grocery shopping in the nearest Aldi. You’re doing it because you have far too much money and time at hand, and because you’re fundamentally stupid and someone else is making a dollar out of you.
This is more akin to people driving an original 1970’s Fiesta that has terrible performance, economy, reliablity, and safety purely because they’re familiar with it and don’t want to find out how much better one that’s 40 years newer is at the exact same job.
I think the Russian Lada (specifically, the Fiat 124 derivatives) would be appropriate to reference here. Or perhaps the Volkswagen Bug. I guess it’s been a while since the same basic car chassis was used for decades, though.
The car analogy fails.
Because op-amps haven’t gotten consistently worse for the last 20 years.
If newer op-amp designs had CVTs, needed their intake valves media blasted every year and were designed to bankrupt you the minute the warranty ended…
Your experience is narrow. So is mine. My vehicles haven’t suffered those issues. I would assert that vehicles have gotten better over the last twenty years, but recall, we’re talking about a time scale of more than double that (hence the previous mention of the Beetle) due to the vintage of the uA741 (which FTA predates the LM741).
Having said that, I’d take almost any 2025 model vehicle over almost any 1969 vehicle for reasons of fuel efficiency, power output, driver aids, active and passive safety systems, paint durability, metallurgy, bearing design, etc.
Writing off all of modernity because you’ve chosen, or passively received, a few bad cars strikes me as short-sighted.
TSW:
My experience is informed by regular contact with professional mechanics.
You did 1969, now do 1995-2005.
Even the Honda dealership techs: ‘Nothing newer then 14. They are just shit.’
2 literish turbo 4 in everything not tiny (don’t ask about tiny). 50 MPG required fleet average. No cylinder liners, piston rings running on aluminum casting.
No manual trans anywhere, but Jatco garbage CVTs in everything.
Hello Cuba. Keep your V8s running!
Soon 10W30 oil will be as much as revolutionary act as 2 stoke oil.
They’re not getting my 2 strokes either.
People use it because it’s in the common examples and documented circuits – and people who are beginners to electronics, or don’t care to research the case, don’t know that op-amps are more or less interchangeable in most applications. In other words, it’s an example of “magical” engineering. Repeating the motions without understanding the reasons.
Unfortunately it’s more common these days to teach people the how instead of the why, under the rationale that more people will find it more interesting to do complex stuff with training wheels on than learning the basics – but if you have no interest in learning the basics and wouldn’t try if you had to, then you’re not going to go forwards either without someone else doing the work for you.
It’s wasted effort to teach people who just expect the teacher to provide all the answers. You get more people who can “do” simple projects like program an Arduino to blink an LED, but that’s about it – so the actual outcome and the reason for the popularity of this style of education is to sell the canned solutions that people come to rely on under this style of teaching.
That’s an interesting point; I do wonder how much of this trend is pushed by companies selling ‘diy’ or ‘maker’ type products that appeal to that demographic. Dont get me wrong, I’m glad that I don’t need to go design a single board computer from scratch to use one, but the number of people I’ve needed to help through basic understanding even at a college level where arduino type systems / teaching is being used is…. kinda worrying. So many of the ‘educational’ systems are blackboxed and absteacted away enough that students turn to just looking up answers and trying them untill something works, because the actual mechanisms behind why it works are buried so deep.
You ‘simplify’ to make ‘learning’ easier, but you’re really just mimicking the results of learning while reducing the level of understanding that people actually gain.
I’m personally really glad that I started learning electronics from the bottom up; it takes a bit more work sure but it’s very worth it when you try to do anything other than just copy something else.
It’s sort-of self perpetuating. The more you “help” people, the more they come back for more help, which attracts “helpful” people to cater to that demand, creating a cycle of dependency.
Whereas, if you have a teacher who maintains standards and demands proof of knowledge, well, the students avoid that teacher. This automatically sets the trend.
teach a person to fish, etc?
maybe that is why none of my students ever want to talk to me hahhaaha.
seriously though, your point is well made. I’m probably a lot more demanding in my teaching style, and definitely more bottom–>up. I’ve seen “the kids” that had the misfortune of being taught a results driven mentality and they crumble under stress when something doesn’t go well, and can’t figure out meaningful solutions under stress because they are basically technicians that can do a narrow set of things well, but if anything is abnormal they don’t understand underpinnings and it all goes to pot in a hurry.
.
I guess it is working I’ve had many students tell me the same, and my best students are all voracious learners. But I’m fortunate to have a mostly excellent group of motivated and self-selected graduate students.
anyway thanks for your thoughtful notes on the matter
cheers
I’d say not so much “magical” engineering as “lazy and conservative” – if you can copy & paste a very well established and proven circuit circuit using a device that has decades of use and is demonstrably “good enough” and “cheap enough” it’s a non-zero risk to rock that boat and pick an alternative device.
A lot of projects won’t need any more performance than the 741 gives, and people designing more serious stuff likely already know enough to pick better devices.
The difference is that the conservative or lazy engineer knows that there are other options, whereas the magical engineer doesn’t because they don’t actually know what they’re doing.
Giving very specific solutions to very specific tasks isn’t educating. You don’t understand what’s happening but it works if you do it exactly, so what you’ve actually learned is a magic trick.
TBF, there’s also people like me who use the 741 (and other antiquated semiconductors) for audio circuits because those non-ideal characteristics are desired. Those unique sounds from clipping, nonlinear gain, and low slew rate are just what I’m looking for. It’s why I pull from my stockpile of 1970s transistors and diodes instead of using newer components for many things.
Using the half-baked analogy from above, some people drive Ford Fiestas because we like them, even if we know there’s better cars available.
But it’s trivial to make a modern op-amp worse in a controlled manner. What you’re doing is basically designing a filter and a limiter around the op-amp, providing non-linear feedback, etc. which allows you to tune the effect and replicate it, so you’re not relying on some old IC having the exact same defect for that “unique sound”.
This just goes back to the point of magical engineering: not knowing what is actually happening and not being able to replicate it by design.
If you want a car analogy, go cut the belt to your power steering pump and come back to say you like it better like it was in the 1970’s.
He’d say that driving new cars don’t give him the same workout he got with the old one :-)
Modern cars don’t have power steering belts.
Less crappy ones have electric motors providing variable power, cheap or older ones use a serpentine belt.
Like I said upthread, car analogy fails.
Newer cars are worse, government mandates for suck, plus better planned obsolescence..
I would happily drive a Ford Fiesta, but there’s no way I’m using a 741 unless I need those non-ideal characteristics to replicate a classic circuit.
The analogy annoys me because I like small, cheap, reliable, fuel efficient cars. I admire the engineering in making things like that. And this is also why I wouldn’t use a 741 in a new circuit. Some people seem to arrive at the opposite conclusion.
When it doesn’t matter that something is old and an old option works just as well or better I tend to like the old “built to last” tool and the classic aesthetic. Like an old woodworking hand tool or something. For me this does not extend to electronics. Old electronics don’t work just as well as new ones :)
I am s college professor who stocks and uses the 741. We use it for the very reasons you claim it is not a good choice.
It is hard to demonstrate how to measure slew rate when the slew rate is high. Give us a relatively slow change in voltage over change in time, and the student can capture it on the scope and understand it.
And the gain bandwidth product of the 741 allows you to see the effect in our lab exercises The same effect happens at much higher frequency on modern op amps, but would be harder to observe.
Thank you :-)
I would suggest, rather than using an ill-behaving part to demonstrate non-ideal behavior, you use a modern op-amp in a circuit that deliberately introduces the non-ideal behavior such as the Miller effect, or limited slew rate. You just put it in a “black box” and say “This is a bad op-amp. Let’s see how bad it is.” You can then switch the effects on and off for comparison.
This way you’re explicitly showing what is happening, why it is happening or the most likely reasons for it, and how you can model it happening in your circuit design when using ideal circuit elements in a simulator to test whether your design is sound and robust to implement in the real world. With the 741 the students may be observing the effect, but aren’t really learning anything other than “use a better op-amp”.
For example:
https://www.ti.com/lit/an/sboa218b/sboa218b.pdf
Using external circuitry to introduce the behaviour changes the lab from “measure the performance and characteristicscof the device”, defeating the purpose.
If the purpose is to demonstrate the effects of e.g. slew rate, then I don’t see why it would defeat it. More so, it would allow you to vary the slew rate and give a better understanding of what the practical consequences are.
Why, when it’s so simple to demonstrate those effects without the extra circuitry?
Because then you can answer questions like, when does effect stop or start being a concern? How good or bad is an op-amp that has this slew rate or that GBW?
You can’t really tell just by observing one single case, since there’s no comparison.
this reminds me of an artticle here a few months ago that had a “simple circuit” that “clearly doesen’t oscillate” to illustrate the punchline that “you don’t really understand your oscilloscope”. since i truly don’t understand my oscilloscope, articles like that always catch my eye.
i spent some time with it before i understood that the argument that the circuit “clearly doesn’t oscillate” comes down to a hand-waving argument that something like “V2 = V1-Vgs” of some BJT transistor, therefore V2 < V1 it doesn’t oscillate QED. but Vgs is arbitrarily small for small currents and the power supply voltage was chosen so that Vgs was effectively less than 1% of V1. so it came down to 0.99 < 1.0 therefore it doesn’t ring so long as you can disregard negligibly tiny sources of noise and overshoot.
that really upset me because i had been misled by the text to believe it was an innocent and clear circuit, when in fact it was intentionally designed just to highlight this behavior. there really is a frustrating limitation of my oscilloscope here, but the circuit simultaneously proved it existed but also gave me the impression it’s irrelevant because i would have to contrive something in order to demonstrate it.
in education, it is so much more valuable to run into a limit genuinely, rather than through contrivance. let the limit be a problem you have to solve instead of a textbook exercise you don’t see the point of.
That sounds like a circuit that was designed to (not) oscillate, but not when you load it up by connecting the oscilloscope probe to it – which is a valid point and a good thing to remember anyways, since you can easily disturb weak signals with it.
it was clearly designed to oscillate, because it was designed for one purpose and only one purpose: demonstrating surprising aliasing behavior in oscilloscopes. i wouldn’t generally make a just-shy-of-unity gain amplifier just to feed high-impedance inputs for any other reason. that’s what frustrated me about the past article i’m referencing.
but the present comment i’m replying to is where you said explicitly that you ought to use contrived evil circuits to illustrate op amp weaknesses instead of letting the weaknesses arise naturally. i think there is a place for contrived examples, and kudos to you for disclosing that you’re recommending an explicitly-contrived example. but for education, there’s nothing like actually genuinely meeting the limitation while trying to do something desirable for its own sake.
Exactly this. I was told about the 741’s shortcomings as a student over 20 years ago. We were also told that this was the very reason they were instructing us to try both the 741 and a modern OP amp in our lab. The 741 is great for teaching just because it sucks! Just be sure to tell students that though.
Thank you for this. For one, everything you need to know about BJTs, you can find in the 741. And, it’s pretty clear that because it relies on them that you can, as you say, see the effect whereas with FET and other newer op amps it would be much harder to observe.
There’s an awful lot of neophiles reading hackaday.
“a personality type characterized by a strong affinity for novelty”?
Precisely. People who want to use new stuff because it’s new and not because it’s any better for their application.
Most of the arguments here are incoherent and boil down to ‘but it’s really old and there’s lots of newer ones’
The argument isn’t that there’s newer ones.
The argument is that the newer ones are more efficient, reliable, and cheaper.
If you’re sticking to LM741 because you think it is good enough and there’s no need for improved components, you’re going to struggle with miniaturization, low power, high precision amplification, and more.
But sure, stick with your expensive, power gulping, noisy, bulky components. The neophiles are the ones designing the circuitry for your hearing aids, grandpa. Good luck getting a 741 into one of those.
“The neophiles are the ones designing the circuitry for your hearing aids, grandpa. Good luck getting a 741 into one of those.”
Aha. Ever tried to repair one of these modern hearing aids or get replacement batteries at the super market?
With the models from late 20th century you can do all this.
There were those external box models with a headpone that fitted in a shirt pocket
or those more modern, skin-toned ones that wrapped around the outer ear.
My grandma had a modern hearing aid, it was a pita.
She could barely use it, it was too tiny to held in her old shaky hands.
It’s amplification factor was way to high, damaging her ear even more in long run.
Next time, please think before you speak. Thank you!
No need to learn to fly in a cessna. Start with a U-2.
Additionally, when dealing with op-amps, intentionally using a low-slew part makes the circuit significantly easier to neutralize. I’m not saying that a 741 is appropriate for general use outside of educational contexts, but it’s generally a good idea in many circumstances to avoid using parts with specs that vastly exceed the requirements. The cessna requires much less care to fly than the u-2, and the 741 may suck but its low slew rate makes rf feedback/oscillation much less of a risk, and easier to avoid.
At least it isn’t a 709.
I once bought a 709 – round metallic case – a long long time ago.
I never used it, it’s still part of the inventory !
The 709 is externally compensated. In some situations, the Gain-Bandwidth product and some other specs are greatly superior to the 741.
In the late 60s while in high school I was mad about electronic. I had read about these devices called operational amplifiers but had no access to them so I built a discrete device copy of the Fairchild uA702. I found it was somewhat capable of being used as described in the various application notes.
In late 1969 I saw an advertisement in a electronics magazine promoting a new Fairchild op-amp, the uA741. This promised various features said to make its application nearly foolproof and it had internal frequency compensation. Most importantly its one off price had been reduced by half from the original $50 so I decided I had to have one, even though that was still a lot money.
Unfortunately after purchasing buyers remorse set in. I could not bring myself to use it in case I let out the magic smoke. So I continued with discrete circuitry.
I subsequently went on to university to do electronics engineering and my interest moved away from analog to micros.
Recently while sorting through various components gathered over 60 years of working with electronics I came across a TO-5 8 pin metal can uA741, still in its plastic shipping carrier. It was my original device.
Surely I could use it now. Unfortunately I now read that “the LM741 (aka uA741) sucks. This chip will ruin your circuit”. What a disappointment! Or perhaps this statement is a little overblown?
With one cap and a couple resistors the uA741 could be a gain of 1000 amp. And +-15V or +-18V matched up well to a lot of instrumentation in the 1970’s. At the low frequencies of seismology, two of them in series gave a gain of 1 million which was great for seismometers before modern neodymium magnets.
By 1983(?) the LM308 and LM358 were available with orders of magnitude better drift and offset, and speed respectively. Where you could integrate for a fraction of a second with the 741 before the offset drove it to the rails. With the LM308, one can integrate for seconds and with lower power and higher impedance. Compared to the 741, it behaves like an ideal op-amp. I never understood why the text books did not switch to the LM108/308 and the excellent LM324 quad.
This. I don’t remember ever using a 741, but I thought the ‘358 and ‘324 were the same thing without the compensation pins brought out. Looking up the datasheets though, I find that no, the ‘358 and ‘324 were indeed a lot better. I’ve used them a lot in commercial designs, where the input and output had to span a 10V range which the ones recommended in the video cannot. They also have much lower supply current than the ones recommended in the video, and I’ve used them for timing even minutes, not just seconds, even though they don’t have FET inputs. When I didn’t need the speed, but wanted lower power and lower input current, I’ve used the LP324 rather than the LM324.
Funny thing, today I was organizing some old drawers of parts and I ran across a plastic snap-box with conductive foam. In it are some LM308, LM224 and LM3900 Norton OpAmps. And with the box was the very cool Linear “Quadzilla” booklet (National Linear Quad Application Book) with piles of OpAmp circuits for the 324 and 3900. Also in the box, LM385 voltage reference, 78L05 regulator, DS0026 MOS Driver, LM6361 50Mz OpAmp, AD790 Fast Comparator.
Time to do something interesting! (I can not find Quadzilla online. Maybe I’ll scan mine.)
Learning how op-amps work with a 741 is like going to see an old doctor : you hear the worst and later you only get good surprises !
Some of the limitations of the 741 were due to the processes available in the day. Compared to today the accuracy of photolithography was miserable. That and less pure chemicals meant inferior offset voltage. If the 741 design was adapted to a modern process but left otherwise unchanged, offset voltage would be much improved.
How has no one mentioned the greatest long play dad jock of all time.
Afrotechmods has not put out a video in 7 years. He released in on April (4) 1st.
741.
It was a work of art damn it.
Not a peep here about semiconductor manufacturer’s pricing of op-amps, it’s ridiculous. Op-amp price-point per penny depending on the performance. It keeps the old dogs alive and well.
China even did their own die-shrink and optimization of the LM358, The People’s op-amp lol.
That I have to pay 10x the price of a 741 for better performance, not relevant to the silicon or fab costs, just marketing wank. It seems really silly leading to 1,000’s different op-amps out there.
What? There are a huge number of much better performing op-amps that are much less expensive than the 741… I don’t know what you are talking about having to pay 10x for better performance. It’s like paying 0.1x for 100x the performance.
Can you give some part numbers please, of these better performing yet cheaper than the LM741. I don’t see them.
The LM358 is no darling, either.
https://www.youtube.com/watch?v=VgodYtiD_F0
I learned long ago, from experience, to put that resistor to the negative rail to get rid of the crossover distortion problem. If for example, I had a 30K capacitively coupled load, a 10K resistor to the negative rail (always ground in my case), for a 3:1 ratio, always took care of it.
continuing to still use the lm741 is a no-brainer. the tlc081, for example, is 2x-3x as expensive, and a quick survey shows it’s available at mouser but not at jameco. i trust and respect the jameco concept. i’m confident that if i got out of my hammock and into the basement, i could find an lm741 sitting in a drawer, just like i could find an lm393, 7805, lm317, 555, 2n2222, 2n2907, 1N4001, 1N4148, red and green indicator LED, disc capacitors up to 100nF, electrolytic 10uF and 50uF, 1/8W 5% resistors between 1ohm and 1Mohm, and a toggle switch probably the wrong shape and size for my project. and almost all of them came out of a 10 or 100 count bag or blister pack that said JIMPAK on it.
i don’t do a whole lot of analog circuit design, but even so i know there’s better versions of all of those things. for example, i’ve wanted a low dropout version of a 7805 often enough that i’m certain i’ve got a few of those sitting around even though i also stock 7805s.
i’ve met the limits of a lot of them. for example, one day i had a eureka moment and i could see a voltage-inverter (produce -5V from a gnd/+5V input) in my head, so i built one…a 555 and a couple BJT transistors and a 50uF capacitor, iirc. it worked! i lit an LED with it. i thought, what would it take to increase the current capacity? or to get a smoother output? and suddenly i found myself imagining ‘power transistors’ and lo, you can go to mouser and get a transistor with almost any ridiculous spec you can imagine. i’m even vaguely aware that you could do better with a coil.
i mean, it’s the same idea as the 100-in-1 electronics kits i used to play with. a basic equipment list that gives me just enough flexibility to try things out, and then be frustrated by the limits.
when you start getting into ‘better’ products, you really want to know what you want. there will be trade offs, and until i’ve actually got a specific circuit in mind, i won’t know what specs are worth trading off for.
I’ve missed [Afrotechmods] and his past references to rare squirrel pr0n.
Seriously, his old “joke” articles were just as informative to me as his later serious ones.
… so what is the next better/modern OpAmp that would perform better and cost about the same (or less)?
(as far as my meager experience went I never progressed beyond obvious/naive circuits on them bread boards – nothing that would be called “useful”).
Above, I told of the improvements the LM358 (dual) and LM324 (quad) op amps brought, and I forgot to mention they’re cheaper too, and cheaper than the ones recommended in the video. We’ve put possibly hundreds of thousands of them into products at work, and we were buying them for something like fifteen cents apiece. In small quantities, hobbyists will probably pay something like 39¢ each.
Thank you, ACK; coincidence had it, whilst sorting out loose parts drawer just this weekend, came across few LM324Ns : – ] Turns out I am already stocked on these, time to put them to use.
During college I got a few LM741 handed to me specifically because it was so awful!
It was the perfect OpAmp to demonstrate how a badly chosen OpAmp can make your life a living hell and how to find a suitable replacement.
If you had a box of 741s sitting in your garage and were building one-off projects would you still not use them?
Well.. if you happen to also have those voltages already in your circuit… Ya, ok, they kinda suck.