Does Op-amp Rolling Work?

[Wolf11Man]

The Chinese are said to be very skilled people, you know?

 

[amirm]

I have picked countless fake NE5532 OP-amps out of Asian products. If you change one of these for the right part, a TI made NE5532, you hear a huge and not only academic difference. The high end, that was grainy, suddenly get's crystal clear.

??? You never measured this, right?

 

[JSmith]

fake NE5532 OP-amps

Can you post evidence to support that claim? Remembering one manufacturer here posted an invoice from TI... turned out TI have various places where they are produced and don't always appear identical.


JSmith

 

[Ropeburn]

I'm not saying fake 5532 don't exist, but I highly doubt the merits of choosing to use one in a hifi product of any claim of quality (in reputation terms) that uses 2 or 4 and as such it saves mere cents per unit produced. Genuine 5532 are what, 30 cents per piece maybe? Might be even less for a factory that orders thousands.

To my knowledge, not even the cheapest products using dozens of them (think cheap Behringer studio mixers with probably 12-16 per channel) used fake ones, because it's just not worth it. The original product is way too cheap and easily available, for decades now, for this to make sense.

 

[Wolf11Man]

No, it was so ugly to listen to the hyped D-amp that I didn't bother. Never thought I had to prove it to anyone. I got the new amp because of good reviews and was really surprised how bad it sounded. How did I know it sounded bad? Much earlier I had bought an amp board with the same D-chip, that absolutely surprised me in a positive way. I didn't expect anything from 7$ part. So when the I got the complete amp, I was really pi**ed off. Someone gave me the advice to look for the Op-amp, same part, but old, maybe from some gear I already had. I took the 5532 from an old PA amp and the D-amp was suddenly just fine as expected and matched the OP-amp less board. I even used the same supply, from an old HP laptop.
You may take it as an anecdote. The marking on the "fake" TI NE5532 where brownish, scraped into the plastic, while the real one had a clean white marking, that could not be wiped off. Since then I have bought a few dozen NE5532 from Mouser, which have a clearly executed marking on them, no coarse scratching.

I get the point of the NE5532 to be that cheap, this is why I found it so surprising. Only two of them in a 36$ amp shipped from China.
Oh, stop, maybe 2x30 Cent are important to someone selling them to you?

PS pin compatible cheap industry OP-amps are about 0.7 Cent, that would make quite some difference. And why should Behringer make their own products sound worse using cheap fakes?

 

[Bruce Morgen]

People never want to hear that most NE5532's in sub 100 Euro amps are fake.

There is no credible evidence that the popular "most" presumption is accurate -- although I have little doubt that it has occasionally been true at times during the "chip amp" era. It turns out that genuine TI NE5532s are manufactured in two different foundries, neither of them on U.S. soil, and that the electrically identical parts from each foundry have distinctive markings, leading some folks to conclude that one of the two differently marked parts is a fake. Moreover, for a manufacturer there is almost nothing to be gained by replacing an already dirt-cheap "jelly bean" IC like an NE5532 with an inferior counterfeit absent availability issues.

 

[Wolf11Man]

Is there a simple test that can separate a fake NE5532 from a real one? Maybe I still got them somewhere. I got basic electronic tools.

 

[Ropeburn]

It takes a lot of skill to sell something worse than TL072

Infact my Waldorf Pulse synthesizer from 1997 that uses a bunch of TL072 for its oscillators (sound generation out of thin air lol) and various other audio functions says they're perfectly fine. Sound demo of a Swedish synthesizer god:

TL072 may be outdated, but it still sounds excellent. Even more so the 5532.

 

[Bruce Morgen]

TL072 may be outdated, but it still sounds excellent. Even more so the 5532.

Are you saying you tried an NE5532 or NE5532s in a decades-old product that was designed for TL072s and the resulting "sounds" weren't quite as "excellent" as they were before the swap -- and how would such an entirely subjective observation relevant to the discussion at hand?

 

[Ropeburn]

Are you saying you tried an NE5532 or NE5532s in a decades-old product that was designed for TL072s and the resulting "sounds" weren't quite as "excellent" as they were before the swap -- and how would such an entirely subjective observation relevant to the discussion at hand?

A highly loaded question, and not what I said either.

I said "TL072 still sounds excellent" and it does - see example 1.

Of course I didn't try 5532 in the Pulse synthesizer. What for? It sounds excellent as it is, as a musical instrument it's perfectly great and sounds as powerful as it ever did.

Could it probably be improved upon? Certainly. There's a Waldorf Pulse 2 from 2013 that arguably sounds even better (tastes are divided). No idea which opamps it uses, but I doubt it's TL072.

Either way I was talking in absolute terms. If an old electronic instrument (as in sound source) using TL072 still sounds good and very satisfying today, and the 5532 was the direct competitor of it, it just can't be bad at all to our ears today, instead better classified like "transparent".

 

[wwenze]

Is there a simple test that can separate a fake NE5532 from a real one? Maybe I still got them somewhere. I got basic electronic tools.

Idle current

 

[MAB]

Regarding NE5532 vs. TL072, here is a preamp with two opamps in non-inverting configuration, Elliott Sound Products Project-88.
The first opamp is non-inverting with 6dB of gain. The second is also non-inverting. Total gain is 12dB, and can be configured as needed. I had a leftover board and built it with opamp sockets. I also used two different feedback resistance values on the first amplifier stage (the different values can be seen in the photo below).

It's built in a used box I had, with a spare PCB and some leftover parts from a previous project, with two 10-turn pots as L and R volume controls.

I don't think anybody is going to hear the difference between the preamp with two NE5532 or two TL072:

Also note, the gain didn't change one bit as I swapped the opamps. The FR didn't change either.

In fact, the biggest difference is the Left vs. Right channel Distortion and Noise results. Which is due to the different feedback resistors on the first amplifier stage. Here is the schematic to illustrate:

The TL072 is less happy with the low resistance feedback loop than the NE5532 at high gain. Otherwise, the lower resistance feedback has an advantage at lower output.

Here is THD+N for the two opamp configurations across preamp output voltage:

Further nitpicking...

The low resistance feedback loop used on the Right channel (red trace) gives better performance using the NE5532 but starts to give up that performance above 2V. Distortion rises and causes worse overall performance above 3V.

The TL072 on the Right channel (orange trace) barely outperforms the higher impedance network, gives up performance above 2V. It enjoys the same distortion benefits as the NE5532 at low output, but suffers higher noise across the output range.

These preamp output voltages are all unrealistically high volume in most systems, and if any of this was an issue lowering the source output, increasing the gain of the second amplifier stage and the input opamp would run in a more optimal range.

I'm pretty sure none of this is even remotely audible.

 

[boxerfan88]

^^ nice!
Thanks for testing...

 

[Sokel]

Regarding NE5532 vs. TL072, here is a preamp with two opamps in non-inverting configuration, Elliott Sound Products Project-88.
The first opamp is non-inverting with 6dB of gain. The second is also non-inverting. Total gain is 12dB, and can be configured as needed. I had a leftover board and built it with opamp sockets. I also used two different feedback resistance values on the first amplifier stage (the different values can be seen in the photo below).
View attachment 444543
It's built in a used box I had, with a spare PCB and some leftover parts from a previous project, with two 10-turn pots as L and R volume controls.
View attachment 444544

I don't think anybody is going to hear the difference between the preamp with two NE5532 or two TL072:
View attachment 444545
Also note, the gain didn't change one bit as I swapped the opamps. The FR didn't change either.
View attachment 444559
In fact, the biggest difference is the Left vs. Right channel Distortion and Noise results. Which is due to the different feedback resistors on the first amplifier stage. Here is the schematic to illustrate:
View attachment 444546
The TL072 is less happy with the low resistance feedback loop than the NE5532 at high gain. Otherwise, the lower resistance feedback has an advantage at lower output.

Here is THD+N for the two opamp configurations across preamp output voltage:
View attachment 444548
Further nitpicking...

The low resistance feedback loop used on the Right channel (red trace) gives better performance using the NE5532 but starts to give up that performance above 2V. Distortion rises and causes worse overall performance above 3V.

The TL072 on the Right channel (orange trace) barely outperforms the higher impedance network, gives up performance above 2V. It enjoys the same distortion benefits as the NE5532 at low output, but suffers higher noise across the output range.

These preamp output voltages are all unrealistically high volume in most systems, and if any of this was an issue lowering the source output, increasing the gain of the second amplifier stage and the input opamp would run in a more optimal range.

I'm pretty sure none of this is even remotely audible.

Yep.
This little cheap thing is full of TL072/074 :

Note levels (inevitable, it's an x-over), application and my silly measuring rig and there's no more to ask.

 

[Cbdb2]

TL072 are FET input where the NE5532s are BJT input. So youll see the TL072s where high input impedance is needed like guitar inputs.

 

[DaneelOlivaw]

I did this easy test before i swapped the 5532 to a 49720 in a fosi bt20a pro and according to the test it was an original 5532 with diodes. (But I didn't heard a difference)

P.S.: You find the explanation of the special architecture of the 5532 from about 3:09 and how to test these diodes with a simple multimeter. In the next minutes an original and a fake sample is beeing tested.

 

[T&T]

Have anyone tried opa2891...

 

[antcollinet]

Come on people. Stop it with the drive by youtube links:

Notification about “External YouTube Video Link Dropping”

Forum Wide Notification: We are strongly discouraging random YouTube and similar “Video links” being posted without some detailed type of description as to what you want to highlight, or what you think is important to our community. Another words, what in the video do you want to...

www.audiosciencereview.com