What is wrong with op-amps?

[Gordy]

After using mainly tubes I am now looking for high performance low gain (about x4) from solid state, and hence looking at op-amps.

There seems to be some bad sentiment towards them for some reason.

So what is actually wrong with op-amps?


(If you say ‘they don’t sound good’ please give a technical reason why you think that might be).

[bob91343]

As far as I know, intelligent application of op amps results in superior performance. Of course there are hundreds of types so you need to select judiciously, although for mundane audio work many will do.

If you drive from a high impedance source you need to stick with low bias current units. If you want wideband operation you need a high speed unit. Not all op amps will be stable at a given gain.

But over all, they are terrific devices and won't let you down if you treat them right. Much, much easier to apply than tubes.

[Mooly]

Hi Gordy... try something like the OPA604 or OPA2604 if you have been used to valves. OpAmps do have their own sound, the fact that we seem able to pick up on these tiny differences always amazes me, but we do. The 604 has the "right" kind of distortion spectrum to sound pleasing I find.

Edit... whats wrong with opamps ? Nothing, it's the way they are implemented that's often wrong

[teemuk]

Originally Posted by Gordy

So what is actually wrong with op-amps?

Nothing. There are just awful alot of myths and erroneous beliefs that are reigning in the scene of audio electronics. The belief that discrete circuits for some mystical reason are better than integrated circuits is one of them.

[knutn]

There seems to be some bad sentiment towards them for some reason.

I don't use OPAMP's of two reasons:
1. They have a very low open loop bandwidth. The implication is that the distortion is rising 6dB/octave, resulting in a complete alteration of the overtone spectrum of musical instruments (You may argue that the distortion level is too low to make this effect audible, I am not sure that this argument holds)
2. I never know if they work in class A, normally the output stage work in class A/B; in some of them even the input stage is working in class B. Just look at the power consumption. Power amplifiers in class B is bad enough, small signal amplifiers in class B is magnitudes worse.

[analog_sa]

Originally Posted by Gordy

(If you say ‘they don’t sound good’ please give a technical reason why you think that might be).

Have you missed all the numerous polemics about NFB, solid state vs tube transfer functions, open loop bandwidth, etc, or do you just want all these thousands of opinions and claims repeated yet once again?

At the end it all boils down to: do you know how well can you actually hear? Or would you rather take someone else's word for it?

IME it's easy to build a poor sounding tube circuit and quite possible to build a surprisingly good sounding opamp circuit. When maximum effort is made, tube circuits do sound better than opamps to my ears (or else i would not use tubes at all) but i doubt there is a simple single reason for that.

[abraxalito]

They sound fine when properly implemented - but unlike tubes they're really susceptible to RF and this I think is responsible for much of their 'poor sound' reputation. People give them poor power supplies and decouple them in such a way (as shown in many textbooks and app notes) as to couple noise into them then complain about the 'solid state glare'.

[tiefbassuebertr]

Originally Posted by Gordy

After using mainly tubes I am now looking for high performance low gain (about x4) from solid state, and hence looking at op-amps.
There seems to be some bad sentiment towards them for some reason.
So what is actually wrong with op-amps?
(If you say ‘they don’t sound good’ please give a technical reason why you think that might be).

Each OP-Amp type has it's own THD/IM spectrum, strongly independend from the layout and load conditions so as the quality of power supply (mostly the badest solution with 7815/7915 or LM317/337 in use and hence the bad reputation by some music lover's)
For you an good solution could be this discrete and very easy topology:
BF862 Preamp
If you want also deal with various OP-Amps for the aim of compare, prefer the types AD797, OPA604, OPA134, NE5534 oder LME49710. Avoid in all cases dual or quad OP-Amps (no perfect ground management possible) and use for each op-amp a complete independend power supply and independend power supply transformer. This means for a line stage two power supplies and if there is RIAA head amp integrated, six power supplies.
A good power supply solution is that one from the RIAA head amp NAD model PP-2, schema to find anywhere here on the forum.

Originally Posted by knutn

I don't use OPAMP's of two reasons:
1. They have a very low open loop bandwidth. The implication is that the distortion is rising 6dB/octave, resulting in a complete alteration of the overtone spectrum of musical instruments (You may argue that the distortion level is too low to make this effect audible, I am not sure that this argument holds)
2. I never know if they work in class A, normally the output stage work in class A/B; in some of them even the input stage is working in class B. Just look at the power consumption. Power amplifiers in class B is bad enough, small signal amplifiers in class B is magnitudes worse.

I agree, especially if there are high parasitic capacitances by large and long signal leads to the power amp. The AD797 and the NE5534 are op amps with highest idle current in the normal DIP-8/SO-8 outline - as I know.

[juma]

Originally Posted by Gordy

... So what is actually wrong with op-amps? ...

Most of all - they are easy to use ("anybody can do it").
So, if you are a circuit designer or audio equipment manufacturer you can not charge the same ammount of money or gain the same respect for op-amps based circuit as for "proper, discrete circuit that separates men from the boys !"

[AndrewT]

Maybe, it's down to voltage and current overheads when higher level signals are passing through along with transients. These transients could be in the music or interference.

I am a great believer in ensuring that a power amp can deliver adequate current.
I see the opamp as powering what comes after it.
I cannot see why the same current capability rules should not be applied to opamps.
Just look at what an opamp feeds and what parasitics it might have to drive.
Then see what current overhead capability is left to spare.

That's part of the reason I started looking at discrete opamps. One can choose the level of ClassA output and the level of ClassAB output.

Have a read of Walt Jung's papers on composite opamps. He recognises the problem of high current demands on the output stage affecting the sensitive/delicate devices in the front end. So separate them.