In https://audioxpress.com/article/mod...-part-2-a-few-more-bjt-types-and-jfet-devices it is stated that the OPA1678 family is inferior to NE5532.
But is it really for DIY?
Had a go with a 2,5kHz 4th order buffered LP and HP filter, sallen key on EASYEDA. 1-2k feedback resistors. All melf and NP0, about 0.1$ a piece.
Really good measuring results with a ADCiso. Really low spureous, but of course some, about -110 dB, thermal and voltage noise.
"Cheap" 1mV noise +-15volt power supply cirquit.
The cirquit showed really at it best at 1 volt. Just noise left.
To me it looks like the drive capability and UHF attuantion that TI advertice really are functional?
Saying that because of the flat noisefloor and really low spurea frequency peaks.
Sallen key will be circuit c or d, dependent on frequency. Unity gain negative feedback + positive feedback filter
But is it really for DIY?
Had a go with a 2,5kHz 4th order buffered LP and HP filter, sallen key on EASYEDA. 1-2k feedback resistors. All melf and NP0, about 0.1$ a piece.
Really good measuring results with a ADCiso. Really low spureous, but of course some, about -110 dB, thermal and voltage noise.
"Cheap" 1mV noise +-15volt power supply cirquit.
The cirquit showed really at it best at 1 volt. Just noise left.
To me it looks like the drive capability and UHF attuantion that TI advertice really are functional?
Saying that because of the flat noisefloor and really low spurea frequency peaks.
Sallen key will be circuit c or d, dependent on frequency. Unity gain negative feedback + positive feedback filter
I think the EMIRR is worth having, but you can just as easily add filtering externally. Also, modern opamps actually do beat the NE5532 on quite a few performance variables. If the OPA1678 doesn't suit your fancy, how about the OPA1642, OPA1656, OPA1612, or LM4562?
That said, the NE5532 offers exceptionally good performance for a 1979 opamp and in reel quantities it's certainly affordable.
Tom
That said, the NE5532 offers exceptionally good performance for a 1979 opamp and in reel quantities it's certainly affordable.
Tom
Good to hear that I’m not way off with my choice. In my application i think the 1612 could do 5 to 10 dB better on noise, but also 5 to 10 times the price at EASYEDA.
And gain is 0dB so -110dB is very good, at least for audio.
To me the OPA1679 seems like a forgiving workhorse for a 4 opamp package.
Puts little demand on the layout skills of the hobbyist creator.
And gain is 0dB so -110dB is very good, at least for audio.
To me the OPA1679 seems like a forgiving workhorse for a 4 opamp package.
Puts little demand on the layout skills of the hobbyist creator.
Is is so strange that Doug Self does not khow the best modern audio-oriented for DIY opamp is OPA2192 
Offset 5 uV, noise 5 nV/sqrtHz, THD 0.00008% See here https://www.patreon.com/posts/riaa-pre-2192-v-128347692
Offset 5 uV, noise 5 nV/sqrtHz, THD 0.00008% See here https://www.patreon.com/posts/riaa-pre-2192-v-128347692
Douglas told me that he doesn't believe that CMOS opamps are any good. The datasheets and practical circuits say otherwise, as witnessed by, among several, the OPA1656.
Jan
Jan
Whether believe or not is for priests, not for circuit engineers. 1656 has 500 uV offset, 2192 has 5 uV, this makes it possible to get rid of separating electrolytes, i.e. excess noise and distortion.
That does have good specs but I would be hard pressed to hear the difference between it and NJM2068 that cost 1.21 in DIP8 package.
If anyone owns an APx555, he can repeat all of Douglas Self's measurements with the new op amps and provide hopefully better results. I'd be happy to be positively surprised. I can only recommend the measurements by D. Self in "EE Times" May 25, June 22, July 27, 2011. Eye-opening.
What's curious to me is the assertion that, "The OPA1678 is unusual among op-amps because it shows classic crossover distortion." I don't see how the test results support this -- am I missing something? Or is this based on general knowledge about the device that I'm not clued in to?
The OPAx192 can't even drive a 2k load without reduced performance, let along 500R like the NE5532. Its distortion rises markedly as the load resistance drops from 10k to 2k according to the plots. Its GBP is a bit skimpy at 10MHz leading to distortion starting to rise above about 500Hz too. Its clearly designed for DC precision with the extremely low offset voltage, but its not a great match for standard audio circuitry compared to many others. For instance the OPA1642 is significantly cheaper and can drive 2k and has somewhat better distortion v. freq graph too. The OPA1656 is even better.Is is so strange that Doug Self does not khow the best modern audio-oriented for DIY opamp is OPA2192
In general any opamp with 1mA quiescent current draw will be compromized for audio performance, 2 to 4mA seemingly is needed if you look at lots of datasheets, that's what the best audio opamps take.
Also I think most of what Doug has written about opamps is at least 20 years old, so yes the more modern opamps are under-represented.