• 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!

Op Amp Rolling

MaxwellsEq

Major Contributor
Forum Donor
Joined
Aug 18, 2020
Messages
1,831
Likes
2,793
Running op amps on what is effectively +/-6v will/may compromise the performance and introduce audible defects.
The critical thing in your sentence is: "will/may". You are half correct: it may introduce defects, but you can't be certain that it will.

And this is the issue with op-amp rolling. Personally I would not do it because in general circuits are designed to support the active component and are optimised for frequency response, noise or distortion. Change the active component and who knows what they are optimised for. Your point about rail voltages is valid - unless comprehensive before/after measurements are taken nobody can be certain that the op-amp will work optimally at these rail voltages.
 

egellings

Major Contributor
Joined
Feb 6, 2020
Messages
4,168
Likes
3,407
Rolling a OP amp from a circuit that was designed and calculated for the OP amp that is in use from the factory is like shooting with a shotgun and hoping you hit something far away. I don't recommend OP amp rolling but if you want to go for it there are a ton of skills that can be gleaned from the operations. It's nice to see somebody going at it from a objective approach although we don't know what calculations the designer(s) performed while making this circuitry and any OP amp that would be swapped would be a guess at best.
An uneducated op amp swap could also turn your amplifier into an oscillator.
 

antcollinet

Master Contributor
Forum Donor
Joined
Sep 4, 2021
Messages
7,984
Likes
13,555
Location
UK/Cheshire
Lower supply voltage tends to increase sonic differences IMO
We don't do opinions here. If you don't have specific design details (component spec versus circuit design), or measurements, then "opinions are like assholes - everybody has one"

Sure - if you have a circuit that is designed to operate close to rails and you swap the op amp for one that can't do this, you are going to run into problems. This one of the many reasons why we suggest op amp rolling is both pointless, and may well result in problems.

But your blanket statement can't be taken at face value. like everything, the devil is in the detail.
 

007Shortz

Active Member
Joined
Apr 2, 2021
Messages
102
Likes
163
Location
Bremen
I was curious, bought an Aiyima A07 and replaced the opamps:

- opa2604: Noise in the silence (maybe the Jfet input). Sound, if you want to call it that, unbearable, crumbly. Either I was sold counterfeits at the time (around 2000) or the circuit is not suitable for use with the opa2604. This is what Amir meant by: “... for those of you who have nothing better to do than mess up the design.”

- LM4562: I couldn't hear any difference at all from the standard NE5532.

I've tried op-amp rolling now and ticked it off as a waste of time.
 

MaxwellsEq

Major Contributor
Forum Donor
Joined
Aug 18, 2020
Messages
1,831
Likes
2,793
I changed the op-amps to MUSE02 in my Fosi BT20A Pro. The amp became much better sounding from that. There is one minor issue, I get a weird oscillating sound when the amp first powers on sometimes, for a few seconds.
Which means that the amp is no longer stable when using MUSE02. I would not run an amplifier that did this.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,890
Likes
9,686
Location
Europe
I changed the op-amps to MUSE02 in my Fosi BT20A Pro. The amp became much better sounding from that. There is one minor issue, I get a weird oscillating sound when the amp first powers on sometimes, for a few seconds.
This sounds like a recipe for desaster. What type was the original opamp?
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,890
Likes
9,686
Location
Europe

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,890
Likes
9,686
Location
Europe
NE5532. It's been working fine, do I need to be concerned?
The NE5532 is fine and, if the circuit is designed for it, hard to beat in its performance. I expect both OPAs mentioned to rather decrease S/N (noise is audible) and maybe reduce distortion, but down from a level which is already inaudible.

EDIT: You need to be concerned about the MUSE02. It's instable at power on and then becomes stable. This means it's working very close to the border of instability and any minor change (temperature, supply voltage, load, ...) might make it unstable again. Even if it seems to be stable it could still oscillate at much higher frequencies. For safety reasons put the NE5523 back in and forget about it.
 
Last edited:

MaxwellsEq

Major Contributor
Forum Donor
Joined
Aug 18, 2020
Messages
1,831
Likes
2,793
Is it because I got the MUSE02 op-amps from Digikey? They are not from Fosi. I find it hard to believe a reputable manufacturer would recommend something that is unsafe, but I guess it's possible. Anyway, I will continue to use them because they sound better, and work fine. Is there some sort of malfunction I should watch out for?
Fire. Speakers destroyed. These are the things I'd keep an eye out for. But then I wouldn't need to because I know that (thanks to working on mixing consoles full of them) the 5532 is perfectly adequate for audio and so would not have swapped them in the first place.
 

mikemag

Member
Joined
Dec 17, 2020
Messages
5
Likes
9
Location
Blaine, WA by BC Canada
I was curious, bought an Aiyima A07 and replaced the opamps:

- opa2604: Noise in the silence (maybe the Jfet input). Sound, if you want to call it that, unbearable, crumbly. Either I was sold counterfeits at the time (around 2000) or the circuit is not suitable for use with the opa2604. This is what Amir meant by: “... for those of you who have nothing better to do than mess up the design.”

- LM4562: I couldn't hear any difference at all from the standard NE5532.

I've tried op-amp rolling now and ticked it off as a waste of time.
If you have an op amp inferior to the 5532 then its not a waste of time
you did not notice an improvement because the 5532 is a great op amp
 

OldHvyMec

Senior Member
Forum Donor
Joined
Oct 5, 2022
Messages
389
Likes
316
If you roll in a high bandwidth opamp, they can be very useful for inducing oscillation. This can destroy a lot of other circuitry, which allows the roller much opportunity to learn something about the circuits he's now repairing. If everything goes well, it can even cause failure of other parts of the electronics chain, which allows the roller to help the economy by needing to buy new gear, thus promoting the audio industry.
Ya know it's strange I've used the exact extensions 7 different times in 4 different products and NEVER had a single issue other than being able to roll the
OpAmp and give it a thumbs up or a thumbs down. Some people may be able to hold the cables in some strange way to cause what you claim but never
in the 7 time or the other 10+ people I know that have used the very same cabling, that recommended the extensions to me. I've read all kinds
of claims of shorter cables, the cables not being twisted in the build, soldering instead of plugs, no change in sound from rolling, toroidal transformers
being better, the potential for noise from a switching power supply and don't worry about the input or output typed of plugs "they make no difference."

Maybe I'll have some 24" extensions made up and test your claim on actual amps working in an actual stereo system. I'll bet money they work fine the way
I'll use them. I understand there are people that can break an anvil, personally I never have. :)

The only thing I've ever experienced personally is Speak-On connectors suck for fidelity and anything other than copper, silver or copper/silver clad
or gold over any of them, is a crap shoot. I did and do use Speak-On with Behringer NU12K BUT they are verified copper by me.

I have 9 pairs of class D and one triple rail, triple P/S with the option to roll OpAmps. The only ones that are having any problem are the ones that
don't have the option. ICE ASP1000s. Even the little 39.00 specials from PE with a voltage regulator upgrade have worked perfect for close to 10
years with Sparco, Sonic Image, and Wiess. One did fail, I contacted Andrew at Sparko and he quickly identified the issue and confirmed he had
simular experiences with the Rev A-B Nord buffer boards. Simple PS upgrade using a Sparco regulator.

I read aluminum power cable weaves cause oscillation in the early class D tech. I'm not silly enough to temp fate and try that one. BUT I don't
use aluminum cabling either.

I been trying to blow up class Ds for over 10 years now. The closest I've seen is Wyred4sound SX1000 (no R). Upon inspection one amp in each of two
pairs visually had cap bulge. I had 4 pairs. 2 pairs had NO visual issues and no failures on any of them. I found that both pairs (that had cap heave)
were plugged into a single 15 amp receptacle vs a voltage maintainer like mine of the same vintage. They all were used in bass duty on the same type
of speakers. VMPS RM40s. I think they were browning out and heating up. Just a theory. I also left mine on, in standby for the 5 years I used them. He
turned his off and on after every use.

I gave up on repairing any of these modules for one simple reason it's too easy to swap a board out. To date not a single NC500oem has given me any
problems and there is well over 15 pairs of MB from DIY, and Nord. I have one Peachtree stereo unit that's been on for 3 years (the wife's) and one
3 rail from Nord I've used a woppin' 8 hours since it was gifted to me from the wife almost 5 years ago. Rev Ds.

I have a new Flex 8 on the way, I'll see what all the rage is with a speaker build of GSR Neo 10s and GSR 3.0 tweeters. Lets see what they use for OpAmps,
and subjectively sound like, compared to a First Watt analog active crossover. I care about one thing "How does it sound." After all it doesn't matter what
anyone speculates or measures it still has to pass the most important test of all. "How does it sound in my room with my gear, to me?"

Regards
 

OldHvyMec

Senior Member
Forum Donor
Joined
Oct 5, 2022
Messages
389
Likes
316
That is rather unlikely.
I'd say what is very likely is that you don't know and are speculating and that he does know because he did the swap and listened to the result. The most important
test there is, personal hands on experience and listening with your own set of ears. Who cares what is sounds like if you don't like the results.
Who cares how it measures, unless you HEAR a problem and are looking for a way to change it. The last test in every amp build is listening to the results.
If I'm wrong at least I've been constantly wrong for 50+ years.

It's like hooking up an engine analyzer for fun, WHY? It sounds good, it runs good, it smells good, the exhaust looks good and it gets good fuel milage,
why would I check, unless there was a reason. Maybe I worked with a different types of engineer for the last 50+ years. They seldom look for problems
UNLESS there is a complaint. That was worked out in the R&D department and we are way past second guessing work that is already done. Now it's down
to paying attention for changes from normal operations.

For the record OpAmps change the sound and it would be difficult not to hear a difference unless you depend on instruments to listen with. Better, worse
or indifferent they change what many people hear and if you can't sell them measurements they understand, it's better to let THEM make up there own
mind WITHOUT uncalled for criticism like "its rather unlikely." It's likely he bought a product, heard a difference and THEN confirmed a difference with
measurements as he suggested.

I HEAR my chickens cackling, they must be hungry, the "low food sensor didn't go off," they must be fine, right? Tell them that! lol!

Regards, I think!
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,890
Likes
9,686
Location
Europe
The response from Fosi was:

"Yes, you can use MUSE02 op-amps (from DigiKey) in the Fosi Audio BT20A PRO with 48V 5A PSU.
The changed op amp needs to meet DIP8 dual op-amps. And BT20A PRO supports 48V supply, here is a power table for your reference."
What did they write about the strange noise at power up?
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,625
Likes
25,552
Location
Alfred, NY
I'd say what is very likely is that you don't know and are speculating and that he does know because he did the swap and listened to the result.
Not even close. If he "listened" without basic controls, his impressions are worthless. Anyone with basic experience in design using opamps knows how incredibly unlikely these claims are.

I don't have to drop his particular baseball off a roof to know that it falls downward.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,625
Likes
25,552
Location
Alfred, NY
Ya know it's strange I've used the exact extensions 7 different times in 4 different products and NEVER had a single issue other than being able to roll the
OpAmp and give it a thumbs up or a thumbs down. Some people may be able to hold the cables in some strange way to cause what you claim but never
in the 7 time or the other 10+ people I know that have used the very same cabling, that recommended the extensions to me. I've read all kinds
of claims of shorter cables, the cables not being twisted in the build, soldering instead of plugs, no change in sound from rolling, toroidal transformers
being better, the potential for noise from a switching power supply and don't worry about the input or output typed of plugs "they make no difference."

Maybe I'll have some 24" extensions made up and test your claim on actual amps working in an actual stereo system. I'll bet money they work fine the way
I'll use them. I understand there are people that can break an anvil, personally I never have. :)

The only thing I've ever experienced personally is Speak-On connectors suck for fidelity and anything other than copper, silver or copper/silver clad
or gold over any of them, is a crap shoot. I did and do use Speak-On with Behringer NU12K BUT they are verified copper by me.

I have 9 pairs of class D and one triple rail, triple P/S with the option to roll OpAmps. The only ones that are having any problem are the ones that
don't have the option. ICE ASP1000s. Even the little 39.00 specials from PE with a voltage regulator upgrade have worked perfect for close to 10
years with Sparco, Sonic Image, and Wiess. One did fail, I contacted Andrew at Sparko and he quickly identified the issue and confirmed he had
simular experiences with the Rev A-B Nord buffer boards. Simple PS upgrade using a Sparco regulator.

I read aluminum power cable weaves cause oscillation in the early class D tech. I'm not silly enough to temp fate and try that one. BUT I don't
use aluminum cabling either.

I been trying to blow up class Ds for over 10 years now. The closest I've seen is Wyred4sound SX1000 (no R). Upon inspection one amp in each of two
pairs visually had cap bulge. I had 4 pairs. 2 pairs had NO visual issues and no failures on any of them. I found that both pairs (that had cap heave)
were plugged into a single 15 amp receptacle vs a voltage maintainer like mine of the same vintage. They all were used in bass duty on the same type
of speakers. VMPS RM40s. I think they were browning out and heating up. Just a theory. I also left mine on, in standby for the 5 years I used them. He
turned his off and on after every use.

I gave up on repairing any of these modules for one simple reason it's too easy to swap a board out. To date not a single NC500oem has given me any
problems and there is well over 15 pairs of MB from DIY, and Nord. I have one Peachtree stereo unit that's been on for 3 years (the wife's) and one
3 rail from Nord I've used a woppin' 8 hours since it was gifted to me from the wife almost 5 years ago. Rev Ds.

I have a new Flex 8 on the way, I'll see what all the rage is with a speaker build of GSR Neo 10s and GSR 3.0 tweeters. Lets see what they use for OpAmps,
and subjectively sound like, compared to a First Watt analog active crossover. I care about one thing "How does it sound." After all it doesn't matter what
anyone speculates or measures it still has to pass the most important test of all. "How does it sound in my room with my gear, to me?"

Regards
Now there's a delightful Gish Gallop of misinformation and superstition. Well done! All we need now is the wife in the kitchen.
 

olieb

Senior Member
Joined
Jul 1, 2023
Messages
374
Likes
610
Well, the MUSE02 op-amps definitely sounded better.
I admire this level of confidence. I do not know how to reach it.
How do you know? You listened to the amp, switched off, opened it, changed the opamps, closed, switched on again and then listened?
And you "know" that the sound was different then?
After all the measurements say that the differences cannot be that big. And acoustical memory will wane after a few seconds.
 

olieb

Senior Member
Joined
Jul 1, 2023
Messages
374
Likes
610
When I put in the new op-amps, I immediately noticed a difference that felt like an improvement to my ears.
But you are aware of the phenomenon called "bias", aren't you? There is more than a good chance that this "difference" is not in the sound at all but only in your "hearing", aka the feeling connected to the situation.
 
Last edited:

rsc1

Member
Joined
Mar 7, 2024
Messages
57
Likes
59
Location
Jupiter, FL
I don't know. I used the amp for a while with the stock op-amps and got an idea of what it sounded like. When I put in the new op-amps, I immediately noticed a difference that felt like an improvement to my ears. I don't know if it measures better, but it sounded better subjectively speaking.
:facepalm:

Make sure you also try the orange volume knob, that increases the sound quality by another 5%.

Hope that sarcasm is coming through, not meaning to offend anyone. To each their own - if you like messing with this stuff by all means. Just don't expect us to believe your claims of "improved sound" without any factual evidence.
 
Top Bottom