The Venerable 741

I can't think of one. The design is 56 years old. It has its own Wikipedia page.

There are faster, cheaper, lower noise, lower bias current/offset RRIO amps around these days.

I remember the day when, still a kid in college, I decided to replace LM709s with LM741s in a control system. The 741s were more expensive (the cost of a pretty good lunch) but didn't need external compensation parts or front-end-zener protection, and current limited.

The early 741s had bad popcorn noise, but I'd expect that to be better now.

My default gumdrop amp is OPA197 now, in SOT23. It makes a good comparator too. There are cheaper parts if you can tolerate low supply voltages.

They will work fine as compatators but a TL081 won't.

LM358s are still useful, though. I use them in things like bias loops, where their limited speed and fairly poor input accuracy don’t matter.

In our licensing conversations, pointing out that we’re saving more on the BOM than the royalty costs is a pretty persuasive argument, entirely aside from the improved performance.

Cheers

Phil Hobbs

AFAICT it was just the original 324 that had the shared current source problem. I’ve never run into it in any 358 or 324A version.

The problem is pretty unsubtle, though. Pulling both (PNP) inputs of any section high enough causes its tail source to saturate, so its beta tanks and it sucks all the available drive, turning off the input stages of the other 3 sections.

There are all sorts of Chinese parts on LCSC that are even cheaper and apparently better, but the 358 is a known quantity, which matters a lot.

Cheers

Phil Hobbs

I've used them in power supply regulators exposed to radiation. Being old designs and all-NPN, they're pretty rad-hard.

Jeroen Belleman

The TS27L4 is a good low-priced quad for high voltage (16V) applications, I still use the TL084 sometimes too

For low-voltage/low power the OPA4322 has impressive performance for the price:

For audio the NE5532 is still a go-to.

The LT1723 for voltage feedback and LT1396 for CF are my high-speed jellies

Blimey, I just checked: It isn't all-NPN! Fortunately for me, it still kept working under irradiation...

Jeroen Belleman

The early integrated PNPs had betas in the single digits, so a bit of radiation poisoning was no big deal.

Have to go back to Jim Thompson's MC1530 era to get that:

Non, au contraire, in bipolar transistors, irradiation tends to reduce beta at low current. If it's already low to begin with, a bit of irradiation will soon make it drop to useless values, and the chip stops working.

My typical example used to be the LM317 as compared to the LM337. The latter has (had?) a large lateral PNP which did not survive for long (<50Gy). The LM317 just kept going (>1kGy).

However, chip design changes that do not affect operation under normal circumstances can radically change radiation resistance.

Jeroen Belleman

How could you loose anything that size?

He probably designed that without Spice.

The first IC opamp I tried was made by GE. All NPN, with an internal zener in the level-shift part. It was amazingly noisy.

I designed some little baby-board opamps, all TO-92 transistors and folded-over axial parts. By selecting some resistors we got sub 1 mV offset and under 1 uV/degC, but it was tedious.

I still have some Philbrick opamps that I found at a flea market.

I visited a guy in repair facility near the local airport. He said "Let me show you my repair shop." It had a somewhat disassembled 747 inside with room to spare.

They have a setup for testing jet engines at full power too.

We provided him some electronics for testing APUs, which are prtetty dinky gadgets.

In the life of an airplane, more is spent on maintanance than the plane cost. So the maintanance is an interesting market.

The "natural" CMRR and PSRR of a diff pair stinks without active loads, at least on one of the rails; Thompson had to use tricks to get around that to keep the output sitting at zero when the inputs were equal

Are those the radioactive ones?

No, just 12AX7 dual triodes. The schematic can be found online.

The old Foothill Flea Market was great. Some old coot would die and leave behind a garage of cool stuff and his kids would take it to the flea market and practically give it away just to get rid of it.

The krytrons are radioactive.

As recently as 2006 I designed 741s into a product, mainly on their availability and price, I think they were being dumped on the market at the time. Now they are very unattractive.

My favourite op amp of that era though is the LM301 which has the very useful feature of input common mode range up to positive supply rail.

uA709 was my first but 741 blew them away.

They are, aren't they? Amazing things to watch close up spinning into life and revving up to god only knows what sort of RPM. I have a friend who's seriously into radio-controlled model aircraft. They're about 5 feet long and are powered by actual jet engines! Still can't get my head around that.

Back when I was interested in model aircraft as a kid there were model jet engines, but they were pulse jets, of the sort used to power the V1 missiles at the end of WW2.

Even more interesting to people who fly around in planes that have been around for a while.