• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Different opamp can affect sound?

Keltere

Member
Joined
Sep 18, 2018
Messages
44
Likes
13
Hi to all.
I just received my xduoo XD-05 and i saw in another forum people who suggested to replace the opamp to achieve better sound.
I wanted to know if this is just the usual snake oil or if it can really make a difference.

PS: if there is another thread about this topic, apologize, i couldn't find it.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
https://audiosciencereview.com/foru...ng-replacing-op-amps-in-topping-d10-dac.4576/

In this case op-amp is probably driving low impedance headphones directly. The used opamp should be one that is suited for the task in that case.
Not all op-amps can drive headphones that well.
So in this case it could matter but performance may well be worse with some op-amps.
I would not rely on 'subjective reviews' that much in this case.
 
Last edited:

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
Possibly. Not necessarily. Probably not. It is like a lottery when operating parameters are ignored.

Op Amps are cheap. You may prefer one but it may not perform better. It could even be cooking(overheating) and/or causing other problems in the circuitry.

If you don't know what you are doing, why do it?
 
OP
Keltere

Keltere

Member
Joined
Sep 18, 2018
Messages
44
Likes
13
https://audiosciencereview.com/foru...ng-replacing-op-amps-in-topping-d10-dac.4576/

as long as the output is not driving low impedance headphones directly and is suited for the task I would not bother.

Thanks i was thinking that they just didn't make any difference for topping d30 but then i found the nwavguy page link.

Possibly. Not necessarily. Probably not. It is like a lottery when operating parameters are ignored.

Op Amps are cheap. You may prefer one but it may not perform better. It could even be cooking(overheating) and/or causing other problems in the circuitry.

If you don't know what you are doing, why do it?

I've asked this because i don't know.
If there isn't any difference i will not bother to start investigating about it.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
When opamps are driving headphones directly one or more of their limits can easily be reached.
Before those limits are reached they sound the same, assuming they are suited for that task.
When current limits are reached some opamps behave strangely or just clip.
Some even oscillate at that moment.
I did some tests a while ago with some op-amps I had at hand.
 
Last edited:

andreasmaaan

Master Contributor
Forum Donor
Joined
Jun 19, 2018
Messages
6,652
Likes
9,399
There could be audible differences, but only in some cases. It depends on many factors.

I think the two common misconceptions are:
  • Differences are always audible
  • Better spec'd opamps always sound better
In reality, differences are often inaudible, and better spec'd opamps, if they sound/measure different, may actually give worse results.

One of the studies in this thread modelled the distortion profiles of various opamps and then used DBTs to try to determine audibility thresholds. However, keep in mind that rolling opamps in a piece of gear is definitely not going to have the same effect as adding the distortion profile of an opamp as measured directly from its outputs.

Disclaimer: I'm no electronics person at all, so all I'm saying comes as a non-expert who has read up on this stuff a bit, but that's all.
 
Last edited:

egellings

Major Contributor
Joined
Feb 6, 2020
Messages
4,005
Likes
3,245
I put together an op amp based headphone amp. I buffered the op amp with a bipolar output stage. Op amp was a TL072. The amp sounded superb, (i.e. not there), and I could not tell differences in several op amp types I tried. I'd be leery of driving phones directly with an op amp. The little bean would likely struggle, especially with low impedance phones, say 30 ohms or so.
 

John Dyson

Active Member
Joined
Feb 18, 2020
Messages
172
Likes
90
I don't know about the audiability of op-amp distortion, but it is usually much worse than specified in manufacturer spec sheets. Often, the issue of changing/dynamic input capacitance and resistance (impedance) isn't well considered in normal spec sheets. Also, the load on the op-amp has significant effect also -- more than what it seems in typical specs.
Another thing -- about the input impedance, that has a significant effect on many input transistor combinations between the feedback and input nodes. There are some very detailed works by some engineers actually measuring the distortion behavior of op-amps... It is a bit frightening, especially if one believes the 0.001% type figures that are specified so often.

John
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
I put together an op amp based headphone amp. I buffered the op amp with a bipolar output stage. Op amp was a TL072. The amp sounded superb, (i.e. not there), and I could not tell differences in several op amp types I tried. I'd be leery of driving phones directly with an op amp. The little bean would likely struggle, especially with low impedance phones, say 30 ohms or so.

The question is HOW you buffered the opamp. Was the output stage class A, AB or B (as often the discrete buffers are). In case of the latter I am quite sure some opamps aren't quite fast enough at their output to make the 0 transition flawless with its overall feedback loop speed.

I would not recommend to drive any headphones directly from a TL07x anyway. It's just not suited for this.
Yes, if you take not suited opamps and experiment with them you might be able to get audible differences, not only in noise levels.
 

Speedskater

Major Contributor
Joined
Mar 5, 2016
Messages
1,610
Likes
1,333
Location
Cleveland, Ohio USA
Different op-amps are better suited for different purposes. Op-amp rollers often use op-amps in circuit situations that they are NOT optimized for. In some of these situations, the op-amps may be ringing or even oscillating at very high frequencies.
An op-amp roller should have a high frequency (30 MHz or higher) O-scope, to check their work.
 

John Dyson

Active Member
Joined
Feb 18, 2020
Messages
172
Likes
90
Different op-amps are better suited for different purposes. Op-amp rollers often use op-amps in circuit situations that they are NOT optimized for. In some of these situations, the op-amps may be ringing or even oscillating at very high frequencies.
An op-amp roller should have a high frequency (30 MHz or higher) O-scope, to check their work.
It is very critical, especially if designing ones own op-amp, to make sure that the input impedance -- and the VARIATIONS in input impedance are considered. For critical AC amplifier situations (audio is one such example), the balance and consistency of the input impedance can cause interesting distortion issues. JFET inputs do not solve all of the problems, but can help in certain cases -- big, fat low noise JFETs often have serious problems with capacitance, and that capacitance is not constant vs. the state of the component. That is one reason why a slightly smaller JFET with slightly worse noise specs might be A LOT better in a general purpose audio situation, and might actually manifest lower in-circuit noise than the bigger, fatter JFET that seems to have lower noise by superficial specs. Op-amp design is truly interesting, and i intend the use of the term 'interesting' in the most positive light -- it is a fun challenge to optimize a discrete op-amp.

If designing an op-amp from scratch, and VERY GOOD op amps can be designed using discrete components, with maybe a few matched pairs, it really benefits from understanding the lower level behavior of the components. The 'beta' model of transistor gain need not apply, but true design requires design down to the diode equation, understanding the capacitance nonlinearities, and understanding the deviations from the ideal. This is one reason why nowadays, it is more difficult to challenge the better high quality, general purpose op-amp chips -- the designers MUST know the physics of the device, but many casual analog EE designs don't require that much detail. (A really good EE design can and does utilize all of the details that the chip designer uses, but it is pretty easy to do a good job without knowing the details -- it is just more hit and miss.) Perhaps very importantly, the limitations of on-chip PNP transistors is less than when the 741/301 or even the TL071 type chips were designed. Also, discretes do give a wider flexibility of the parameters for the constituent transistors -- but with more matching problems. Chip designs even need to consider thermal feedback creating distortion at lower audio frequencies!!!

John
 

cjm2077

Active Member
Joined
Jan 22, 2020
Messages
160
Likes
261
Thanks i was thinking that they just didn't make any difference for topping d30 but then i found the nwavguy page link.



I've asked this because i don't know.
If there isn't any difference i will not bother to start investigating about it.

The original designer picked that op-amp for a reason. Part of that reason may have been cost, or familiarity. But overall the part was picked to meet the goals of the design, and the parts around it were optimized for that op-amp. Can a different op-amp be subbed into a given circuit and perform better? Sure. Maybe the original designer wasn't that great. Maybe an inferior part was chosen because of cost constraints. Maybe they just cut and pasted a circuit in from a different design without putting too much thought into it because it worked before.

But I wouldn't trust just anybody on the internet to select a new op-amp for the circuit. They probably don't know the circuit as well as the original designer, and don't know the parameters the original designer was using to select parts. They probably aren't doing the testing and measurement the original designer was. If it's just a guy with a soldering iron, some parts they ordered online, and set of headphones for "measurement", they're probably doing more harm than good to the circuit.

Basically, if the design was crap, don't buy it and think a new op-amp somebody on a forum suggests is going to make it better. If you like the design and it tests well, trust that the original designer knew what he was doing and let it be.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
The original designer picked that op-amp for a reason. Part of that reason may have been cost, or familiarity. But overall the part was picked to meet the goals of the design, and the parts around it were optimized for that op-amp. Can a different op-amp be subbed into a given circuit and perform better? Sure. Maybe the original designer wasn't that great. Maybe an inferior part was chosen because of cost constraints. Maybe they just cut and pasted a circuit in from a different design without putting too much thought into it because it worked before.

But I wouldn't trust just anybody on the internet to select a new op-amp for the circuit. They probably don't know the circuit as well as the original designer, and don't know the parameters the original designer was using to select parts. They probably aren't doing the testing and measurement the original designer was. If it's just a guy with a soldering iron, some parts they ordered online, and set of headphones for "measurement", they're probably doing more harm than good to the circuit.

Basically, if the design was crap, don't buy it and think a new op-amp somebody on a forum suggests is going to make it better. If you like the design and it tests well, trust that the original designer knew what he was doing and let it be.
One exception I can think of is replacing a part with a newer variant designed to be compatible while having lower noise or distortion. The designer wouldn't have used the improved part simply because it didn't yet exist.
 

cjm2077

Active Member
Joined
Jan 22, 2020
Messages
160
Likes
261
One exception I can think of is replacing a part with a newer variant designed to be compatible while having lower noise or distortion. The designer wouldn't have used the improved part simply because it didn't yet exist.

That would be a good reason to upgrade a part. But I'd still want to have a pretty well trusted, experienced designer make the decision that the part is the same and will behave identically in a circuit except for improved specs. And I'd still want to measure before and after. Personally, I would trust myself, but I've got 20 years in analog and power electronics design. I've had to look at the part swaps all the time, sometimes just when parts become unavailable. But I still measure before and after to make sure nothing changes for the worse. I can't say the same for most of the people swapping op-amps out there on audio forums.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
That would be a good reason to upgrade a part. But I'd still want to have a pretty well trusted, experienced designer make the decision that the part is the same and will behave identically in a circuit except for improved specs. And I'd still want to measure before and after. Personally, I would trust myself, but I've got 20 years in analog and power electronics design. I've had to look at the part swaps all the time, sometimes just when parts become unavailable. But I still measure before and after to make sure nothing changes for the worse. I can't say the same for most of the people swapping op-amps out there on audio forums.
Oh yes, the expected outcome should always be verified through measurements. No disagreement there (or in general).
 

Speedskater

Major Contributor
Joined
Mar 5, 2016
Messages
1,610
Likes
1,333
Location
Cleveland, Ohio USA
One exception I can think of is replacing a part with a newer variant designed to be compatible while having lower noise or distortion. The designer wouldn't have used the improved part simply because it didn't yet exist.
Been there and done that at work and we got burnt! It was a product that we had been making for years. Then they updated the simple 741 op-amp. The product would work for a while then the 741 would fail, due to oscillation.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
Been there and done that at work and we got burnt! It was a product that we had been making for years. Then they updated the simple 741 op-amp. The product would work for a while then the 741 would fail, due to oscillation.
Was the change in behaviour documented? Were you doing something unusual?
 

JeffS7444

Major Contributor
Forum Donor
Joined
Jul 21, 2019
Messages
2,347
Likes
3,508
Thanks i was thinking that they just didn't make any difference for topping d30 but then i found the nwavguy page link.

Yes, in the case of Nwavguy's modification to the Behringer UCA202, but that is a specific instance where the opamp is driving the headphones, and the new opamp has much higher output and much lower output impedance. The modified UCA202 sounds about the same as the O2 or any other decent headphone amplifier.
 
Top Bottom