Hello everyone,
I'd like to share with you an experience replacing one AOP with another.
In a DAC, I have a R2R volume control that ends with an OPA1678 AOP.
This 1678 is an entry-level AOP (less than €1 at Mouser) with a JFET input.
I replaced it with an OPA 1642, a JFET with higher performance on paper.
The result was less good, particularly with less detail.
In my test piece, there was a spectator applauding in the very front row. With the 1678, I could clearly hear it stand out from the other applause. With the 1642, it was barely noticeable over the other applause.
I replaced the OPA1642 with an OPA1656. The result was the same as with the 1642.
So I went back to the original 1678 and rediscovered the spectator in the front row (and all the other details, of course).
My electronics skills aren't developed enough to provide a clear explanation, and I don't have the circuit diagram.
But it proves that a component, even a cheap one, when well implemented, can achieve high performance.
I'd like to share with you an experience replacing one AOP with another.
In a DAC, I have a R2R volume control that ends with an OPA1678 AOP.
This 1678 is an entry-level AOP (less than €1 at Mouser) with a JFET input.
I replaced it with an OPA 1642, a JFET with higher performance on paper.
The result was less good, particularly with less detail.
In my test piece, there was a spectator applauding in the very front row. With the 1678, I could clearly hear it stand out from the other applause. With the 1642, it was barely noticeable over the other applause.
I replaced the OPA1642 with an OPA1656. The result was the same as with the 1642.
So I went back to the original 1678 and rediscovered the spectator in the front row (and all the other details, of course).
My electronics skills aren't developed enough to provide a clear explanation, and I don't have the circuit diagram.
But it proves that a component, even a cheap one, when well implemented, can achieve high performance.
I would look at each specification sheet in turn and decide what I required for the optimum result.
Take a look and read the data sheets plus construction notes.
https://www.ti.com/product/OPA1656
https://www.ti.com/product/OPA1642
https://www.ti.com/product/OPA1678
As you can clearly see, they are all very different.
Usually the ear worm gets involved and without proper scientific test results, the findings become snake oil, just a matter of taste, nothing more.
Take a look and read the data sheets plus construction notes.
https://www.ti.com/product/OPA1656
https://www.ti.com/product/OPA1642
https://www.ti.com/product/OPA1678
As you can clearly see, they are all very different.
Usually the ear worm gets involved and without proper scientific test results, the findings become snake oil, just a matter of taste, nothing more.
A usual problem when replacing old op:s with modern faster types is self oscillation.
That normally shows as loss of detail in music and increased noise.
So if you can hear more noise from the faster op:s there is probably a self oscillation.
Normally they are a few dB silenter.
But OPA1678 is a good OP. 40 or maybe 20 years ago it would have been one of the best. In a good circuit the distortion will be impossible to hear.
So it can be some snake oil too.
That normally shows as loss of detail in music and increased noise.
So if you can hear more noise from the faster op:s there is probably a self oscillation.
Normally they are a few dB silenter.
But OPA1678 is a good OP. 40 or maybe 20 years ago it would have been one of the best. In a good circuit the distortion will be impossible to hear.
So it can be some snake oil too.
It likely does not prove anything. Possibly the voltage regulator/decoupling scheme is not optimal in that device too. OPA1642 is a darn fine pretty universal opamp.
Last edited:
Maybe you can try the AD8599, but you'll need a double soic to double DIP adaptator. It has less noise than the 1678. But imho the op1678 is already a good op amp... Don't be fooled by their price : both are very good op amp. One is Jfet, the other a bipolar. They certainly chose the Jfet as a part of the vocing or the need of low current input noise. In this scenrio of low current noise, the champion to beat is the opa828 (soic that exist in single as double channel). But here a socket migth or migth not a problem due to its speed.
It is hard to advise w/o understanding the exact use of the op amp in your device ! Shematic?
It is hard to advise w/o understanding the exact use of the op amp in your device ! Shematic?
I have various adapters. This opamp is Jfet based, that's why I find it interesting. Noise is not that important for what I need.