Can op amps now be as good as discrete circuitry?

[AnalogSteph]

Like anything it's down to application. Given the choice between a jfet opamp and a couple of pairs of zetex low noise, low rbb transistors for a mc input stage I'm taking the discrete option every time.

But that's mostly because opamps don't have the super-low-Vnoise, super-low source impedance terrain covered particularly well (if al all) and those that do aren't cheap. Even then you generally wouldn't go all discrete though and merely use the super low rbb' transistors in a discrete input stage. This is one of those niche applications that do benefit from going discrete, particularly because at the same time other requirements like input voltage swing are modest so nobody cares about common-mode issues.

On the other end, opamp options quickly thin out once high voltages and/or currents are required. You can just about find some parts suitable for a higher-power headphone amplifier (although people have resorted to using speaker amplifier ICs, too), but at a certain point you're just better off going discrete, at least partially where it makes sense.

 

[jan.didden]

Well... this opamp:
View attachment 390609
indeed and quite measurably, performs much better than the one shown below.
View attachment 390610

But... maybe this one is even better ... in certain aspects ?
View attachment 390611

You did all the measurements? Care to share them?

Jan

 

[dfuller]

In most respects, IC opamps are better than discrete circuits. Discrete circuits can do some things better however they tend to be at large signal levels.

 

[TurtlePaul]

i think one thing worth explicitly mentioning is that opamps beimg integrated circuits means they build many transistors into one package. Don’t let the simplified schematic fool you - opamps have dozens of transistors etched onto that single piece of silicon. Descrete designs advertize the ‘simplicity’ because it would be cost prohibitive to duplicate even a simple opamp transistor for transistor.

 

[DonH56]

Discrete circuits are custom built by Burson for specific applications rather than an op-amp's jack-of-all-trades-master-of-none specifications. The art of circuit design yields superior sound in every way in comparison to simple plug-a-chip engineering.10 Jul 2024

Speaking as an analog IC designer that statement is ludicrous. And there are a myriad of application-specific ICs for audio, instrumentation, video, almost anything you could think of.

As for mixing devices, many opamps do that now, and for many years. BiCMOS and CBiCMOS processes exist, and years ago a pin driver design I did for a mixed-signal tester combined JFETs, NPNs, PNPs, and CMOS devices on one IC die, along with all sorts of diodes and other passive elements, including nice metal-film resistors laser-trimmed to better than 0.1% matching. The last IC design I did was in a HBT CBiCMOS process with all the above plus SiGe HBTs.

Discrete design can yield excellent performance, as can ICs, so hopefully the designer can pick the one that works best. Custom ICs are very expensive, however, limiting access for smaller or lower-volume companies.

 

[Doodski]

I assume that the OP (no pun intended ) is asking whether integrated circuit operational amplifiers can be as good as operational amplifiers built using discrete circuitry?
Vacuum tube op amps FTW!

View attachment 390563

Discrete operational amplifers with vacuum tube active components were the heart and soul of analog computers for many years.
I actually had one of these big Heathkits for several years -- although I never really learned how to use it for anything useful.
In some respects, I wish I had kept it -- but it was very large and very heavy.

Nice old gear. What is the approximate year of sale would this be? $945 might be a large sum depending on the time frame.

 

[LTig]

Like anything it's down to application. Given the choice between a jfet opamp and a couple of pairs of zetex low noise, low rbb transistors for a mc input stage I'm taking the discrete option every time.

Absolutely, given this limited choice. Luckily there are opamps like the venerable AD797 which are perfectly suitable for this task.

 

[solderdude]

You did all the measurements? Care to share them?

Jan

The middle one is the uA741... not very high performance... measurements are out there, the upper one is a 'discrete' one and the ones have seen perform better than uA741.
The 3rd one is just an example of a MOSFET op-amp and am sure will most will perform better than 741 in a lot of areas.

Note: I did not say anything about their sound.
Probably a lot of early recordings may have passed 1 or more uA741's (or TL07x) back in the days which are liked by op-amp haters without them realizing it.

Discrete circuits are custom built by Burson for specific applications rather than an op-amp's jack-of-all-trades-master-of-none specifications. The art of circuit design yields superior sound in every way in comparison to simple plug-a-chip engineering.10 Jul 2024

Some discrete op-amps perform/measure better than some regular op-amps but there are plenty of op-amps that perform better than (purpose built) discrete ones so it is a nonsensical statement for advertising purposes.

 

[KSTR]

Like anything it's down to application. Given the choice between a jfet opamp and a couple of pairs of zetex low noise, low rbb transistors for a mc input stage I'm taking the discrete option every time.

Ics are easier to get right, mostly outperform discrete in most applications, but sometimes....

The clever solution is to combine things in a composite amplifier. Discrete input stage for low noise (notably if JFET), wrapped in the feedback of an opamp for gain and precision.

 

[jan.didden]

The middle one is the uA741... not very high performance... measurements are out there, the upper one is a 'discrete' one and the ones have seen perform better than uA741.
The 3rd one is just an example of a MOSFET op-amp and am sure will most will perform better than 741 in a lot of areas.

Note: I did not say anything about their sound.
Probably a lot of early recordings may have passed 1 or more uA741's (or TL07x) back in the days which are liked by op-amp haters without them realizing it.



Some discrete op-amps perform/measure better than some regular op-amps but there are plenty of op-amps that perform better than (purpose built) discrete ones so it is a nonsensical statement for advertising purposes.

So no real data. OK.

 

[mhardy6647]

Nice old gear. What is the approximate year of sale would this be? $945 might be a large sum depending on the time frame.

late 1950s -- yes, pricey.

 

[dfuller]

Some discrete op-amps perform/measure better than some regular op-amps but there are plenty of op-amps that perform better than (purpose built) discrete ones so it is a nonsensical statement for advertising purposes.

The only discrete design that can match high performance ICs is the 990, AFAIK, and that meets and occasionally exceeds the OPA134, which is arguably the best line level audio IC op-amp on the market.

 

[TurtlePaul]

The only discrete design that can match high performance ICs is the 990, AFAIK, and that meets and occasionally exceeds the OPA134, which is arguably the best line level audio IC op-amp on the market.

Isn’t the 1656 about 7 dB better?

 

[dfuller]

Isn’t the 1656 about 7 dB better?

Than the 134? Yep, appears to be the case. Point stands about the 990 being the only DOA I know of that isn't miles behind in terms of specs.

 

[TurtlePaul]

Than the 134? Yep, appears to be the case. Point stands about the 990 being the only DOA I know of that isn't miles behind in terms of specs.

Just pointing out that there is still a $3.12 opamp (<$3 for quantity 10+) which is better than a $60 discrete opamp.

 

[Bruce Morgen]

I assume that the OP (no pun intended ) is asking whether integrated circuit operational amplifiers can be as good as operational amplifiers built using discrete circuitry?
Vacuum tube op amps FTW!

 

[solderdude]

Just pointing out that there is still a $3.12 opamp (<$3 for quantity 10+) which is better than a $60 discrete opamp.

Often it is not the op-amp that is the most important aspect but rather how you use it (the type of circuit and for which particular task, the input and output requirements of the circuit it is used in).
With high performance/BW op-amps even the PCB design layout matters.

 

[mkii]

The question correlates to what exactly? In terms of audio/hifi applications, OPs are easy to use and implement but can't achieve sonical performance against properly designed discrete designs. In low/mid level segments, most designers go with OP but on higler level only discrete and tubes
Finaly, one can't change resistors nor caps that are already in OP and most of them are influencig a sound in a bad way (for sake using cheapest poossible materials and size).
Try to get Cello Palette and compare to any modern OP exotic design pre amps, and listen

 

[somebodyelse]

OPs are easy to use and implement but can't achieve sonical performance against properly designed discrete designs.

Assumes facts not in evidence.

 

[SIY]

In terms of audio/hifi applications, OPs are easy to use and implement but can't achieve sonical performance against properly designed discrete designs. In low/mid level segments,

Evidence of this remarkable claim?