Class D and Op Amp upgrades.

[P.M.B.66]

Class D the Op Amps are critical and swapping Op Amps to tune the Amp to your liking is maybe the best and simple solution with power upgrades.

What is your Amp and what did you replace the stock Op Amps with?

I have two different amps that I upgraded. One is Aiyima DAC A3 that is a Headphone Amp and came with 4 swappable Op Amps TL072x2+Lm4562+NE5532. Pretty mixed lot. I swapped all 4 with Opa1612 Op Amps. Made a noticeable positive difference from average to good.

Other is two Aiyima A07 Premiums that are Bridged and Bi Amping High and Low separate. For the Amp that is driving the High I have opa1612 Op Amps that have good treble and overall quality. For the Amp that is driving the Low I have opa2134 that have nice warm Bass performance. Together work well. Stock were ne5532 Op Amps that are OK.

 

[knotscott]

I added OPA1612s to my Aiyima A07 Premium, along with a 48V 5A PSU.

 

[P.M.B.66]

I added OPA1612s to my Aiyima A07 Premium, along with a 48V 5A PSU.

Nice. But did you realize how big the difference is between power Supplies? 48V 5A = 240W while 48V 10A = 480W. opa1612 is good but 5A might not be enough to drive the Op Amp to it potential. Low can be a challenge and it seems the High gets more clarity and less distortion. That's my understanding about it. I have 2 x 48V 10A. Of course it is also about what speakers you are driving.

 

[knotscott]

Nice. But did you realize how big the difference is between power Supplies? 48V 5A = 240W while 48V 10A = 480W. opa1612 is good but 5A might not be enough to drive the Op Amp to it potential. Low can be a challenge and it seems the High gets more clarity and less distortion. That's my understanding about it. I have 2 x 48V 10A. Of course it is also about what speakers you are driving.

Yes. 48V/10A is a better option performance-wise. When I bought my 48V PSU, the listing was confusing, as it listed both options.... thought I was getting the 10A, but apparently I had selected the 5A. I don't typically push the volume much, but 10A would have been better....still an upgrade from the stock PSU. Something I could pursue at some point, but it's only being used in my shop system at the moment.

 

[P.M.B.66]

Yes. 48V/10A is a better option performance-wise. When I bought my 48V PSU, the listing was confusing, as it listed both options.... thought I was getting the 10A, but apparently I had selected the 5A. I don't typically push the volume much, but 10A would have been better....still an upgrade from the stock PSU. Something I could pursue at some point, but it's only being used in my shop system at the moment.

I come from 80's gear where your main focus was Wattage and brand name. I have a main system close to top Pioneer components but to be honest are not really used anymore. It was a shock to the system learning the differences of the Class D and how it works and what matters. For example the Aiyima A07 Premium the power supply 48V 10A is more expensive that the amp itself but vital and then we talk about the Op Amps that change sound to your liking and these can be double the price to your amp. I'm talking about Sparkos of course. Bit too expensive for me. Very good but the opa1612 gets you more than halve there with $5 a pop. Also it seems that different brands work a bit different and options for Op Amps changes. Also I think it matters what Op Amp and power supply you have. Some Op Amps seem to demand more power. Not as simple just swapping and expecting a change you also need to research if the power is suitable and that the Op Amp that it replaces matches the stock at least close. I think that's why I opened a new tread for Op Amps because there was some old ones and outdated. So much has changed in last few years and in attitudes and mindsets of audiophile with the Class D. Changing Op Amps has become a norm with upgraded power supplies.

 

[Hammett]

Changing op amps makes a VERY big difference on the 3255 amps, a silly video by a silly website notwithstanding.

I've had good luck with OPA 2132, which is warm and full and the various OPAs (1612, 1642, 1692, but most of all 1656).

When I upgraded the Sparkos 3602 in my Geshelli J2 with a 2590 it left me with a spare 3602. Rather than let it sit in a drawer, I bought a second and put it in a number of 3255s.

The most notable improvement came in the original Aiyima A07, but less so in the Premium or in the Douk 100. It is good enough in the original A07 that I am running it as my main amp.

For the most part, I would recommend sticking with the OPA 1656. It is a really nice op amp and sounds much better than the 5532 these come with--greater clarity and separation of instruments. This one is a no brainer.

I have not found any difference whatsoever between power supplies with more or less amperage, and I have some serious doubts about the traces on a 3255 board taking 10 amps worth of power.

 

[P.M.B.66]

Changing op amps makes a VERY big difference on the 3255 amps, a silly video by a silly website notwithstanding.

I've had good luck with OPA 2132, which is warm and full and the various OPAs (1612, 1642, 1692, but most of all 1656).

When I upgraded the Sparkos 3602 in my Geshelli J2 with a 2590 it left me with a spare 3602. Rather than let it sit in a drawer, I bought a second and put it in a number of 3255s.

The most notable improvement came in the original Aiyima A07, but less so in the Premium or in the Douk 100. It is good enough in the original A07 that I am running it as my main amp.

For the most part, I would recommend sticking with the OPA 1656. It is a really nice op amp and sounds much better than the 5532 these come with--greater clarity and separation of instruments. This one is a no brainer.

I have not found any difference whatsoever between power supplies with more or less amperage, and I have some serious doubts about the traces on a 3255 board taking 10 amps worth of power.

Aiyima A07 Premium is a 48V 10A amp. But that said for example LME49720 Op Amp the 10A is too much and 5A better. Op Amps and power goes hand in hand and needs to be matched. I find that opa1612 is not giving it best with 5A but with 10A it seems to cover more of the range. The distortion at the edge of the range at Treble seem to correct itself. Usually in my knowledge it should be the low that needs more power but here it seems to be upside down. That said tweeters there can be 50% loss in power as heat while woofers it can be 7%. I wonder if that has something to do with it. Aiyima A07 Premium has higher rated capacitors and components. It is not bad with ne5532. Maybe the upgraded hardware gets more from the ne5532 making the gap smaller because I've heard some comparisons to much more expensive Class D amps and the Premium do sound better to my ear.

 

[P.M.B.66]

Op Amps are a subject that many think is nonsense. For A/AB Amps it would hold true as the Op Amp does a different thing. But Class D the Op Amps have a direct and critical effect to the sound.
Following is a reference from the net that gives you some reasons.
-----------------------------------------------------------------------------------------------------------------------------------------
Op-amps with the same specified resistor values can sound different because their underlying electrical characteristics—such as gain-bandwidth product, slew rate, noise, and input impedance—vary significantly, causing them to interact differently with the surrounding circuit. While ideal circuit theory suggests only the ratio of feedback resistors matters, in practice, the internal design, component matching, and thermal behavior of the op-amp dictate its actual performance.

Here is a detailed breakdown of why this occurs:
1. Electrical Performance Variances
Slew Rate and Bandwidth: Different op-amps have varying speeds at which they can change their output voltage (slew rate) and how much gain they provide at different frequencies (GBW - Gain Bandwidth Product).
Input Impedance and Capacitance: The input stage of an op-amp can introduce frequency-dependent loading, which changes the behavior of the input filter in the circuit.
Common-Mode Rejection Ratio (CMRR): Variations in how well an op-amp rejects noise from both inputs (common-mode noise) can alter the sound, particularly at higher frequencies where CMRR falls off.
Output Loading: Op-amps may have difficulty driving low-impedance loads, resulting in distortion that differs between models.

2. Circuit-Level Interactions
Stability and Oscillation: An op-amp with different internal compensation might become unstable in a circuit designed for a different type, leading to inaudible, parasitic, or audible oscillation.
Negative Feedback Loop: While negative feedback linearizes behaviour, different op-amps respond differently to this, particularly in circuits that are not conservatively designed.
Power Supply Noise: Op-amps have different Power Supply Rejection Ratios (PSRR), meaning some are more susceptible to noise from the power source.

3. Non-Ideal Characteristics
Distortion Profiles: Different op-amps produce different types of distortion (harmonic distortion).
Manufacturing Differences: Even when bearing the same part number (e.g., 4580), different manufacturers may produce parts with slightly different specifications or internal structures.

Summary of Key Factors

Feature Impact on Sound
Slew Rate Faster = cleaner; Slower = warmer/softer
Input Impedance Higher = less loading of the source
GBW Higher = more feedback, less distortion
Output Loading Ability to drive low impedance without distorting

While some engineers argue that in well-designed circuits, the differences are minimal, audiophiles and musicians often find that replacing "stock" op-amps with different models changes the tonal quality, soundstage, and, in some cases, stability.

 

[kyle]

While I agree with the general idea of the sound of opamps differing, the implementation is a major part of the effect. Class D amplifiers with Post-Filter Feedback get less benefit from rolling opamps due to the feedback flattening the frequency response more compared to traditional Class D amps. They can still alter the sound but it is far more subtle and is largely system dependant. Sources and speakers with less impedance variations will have less effect from changing opamps in these amp types.
One place where I have found a large difference in sound quality with different opamps is phono stages. The combination of capacitance, inductance and resistance of the cartridge and cables makes tuning more important and variable.

 

[P.M.B.66]

Phono needing more amplification naturally any changes on the chain will make more difference.

In general when discussing Class D and Op Amps in a same sentence you think affordable Chinese Class D amps that have created a specialized Op Amp industry (Sparkos for example) made for these Amps.

Also with Op Amps you need to talk about power. Power is the number one issue with any Op Amp.

 

[fredgarvin]

Amp builders using Purifi, Hypex, IceEdge modules etc. often use ops amps in their custom input buffers to tailor a sound rather than using the on board input.