The best sounding op amps for audio work.

[Addicted to music]

On 12/03/2021 at 2:05 PM, Ian F said:

The new breed of TI J-Fet opamps offer superior performance to that old soldier the OPA2134.The OPA827 for example has superior load drive,much lower distortion and miniscule DC offset.Close to zero DC mv at its output would enable removal of the dreaded electrolytic coupling caps which will definitely lower colouration and improve sound quality. 

those OPA827 OPA 1612. OPA 1642  are the best for audio, nothing comes close,  if you’re getting into LM or LME get in there fast from trusted suppliers as they are being wiped out of existence and replaced with OPA....   The result of BB taking over National.

[stereo coffee]

Efforts being made  to reduce offset to negligible level, with op amps is admirable.

 

If we are concerned specifically with op amps as current to voltage stages in  CD players,  one also needs to review data sheets to see how much offset DAC's have at their output.  The opamp with ability to almost cancel its own offset, is unable to correct what comes before it.   The TDA1541 as example has large offset on  one channel.  So offset cancelling likely needs to be done prior to the op amp.

 

 

 

 

 

 

[stereo coffee]

I have been testing a number of opamps, not from their input to output signal AC ability , but rather their DC capability. The circuit to test them see's a window into their ability from what I consider, is a entirely new perspective.   There are vast differences between each device quite easily, then heard.

 

The subjective parameter being first and foremost sought is ability to reveal depth in music. The best device thus far is the opa2107, its ability to pinpoint where instruments position themselves and to suggest at the same time getting exceedingly close to the anticipated sound and that positioning  , is easily the best so far.

[aussievintage]

Nothing new under the sun.   Opamps  have been used in DC circuits since they were invented, (as anyone who has worked with analogue computers will attest).  The ideas predate integrated circuitry, in fact the first ones were made with vacuum tubes. Coping with DC offsets etc was well established as that would effect the calculations.   I have spent many hours designing and , mostly, fixing analogue computers   - terribly prone to stability problems.  

 

Amplifying ac signals was, almost, a misuse of them, and they required heavy feedback to obtain a flat frequency response across the audio range.   Indeed many designs still struggle to compete sound-wise with discrete valve and transistor circuits.   I see this as the reason for this modern op-amp rolling phenomenon.   The sale of discrete component built "op-amp" plugin replacements is probably understandable.   Next best to designing the circuit from the ground up I suppose.  After all,  designs use opamps for convenience and cost reasons, not because they sound best.   

 

 

[catman]

G'day all, an interesting sidenote to this whole thing.  On my DIY P97 line stereo preamp I've replaced the third NE5532 with an OPA2134 and the overall sound quality has improved to my ears anyway.  Interesting!  Regards, Felix.  

[aussievintage]

2 minutes ago, catman said:

G'day all, an interesting sidenote to this whole thing.  On my DIY P97 line stereo preamp I've replaced the third NE5532 with an OPA2134 and the overall sound quality has improved to my ears anyway.  Interesting!  Regards, Felix.  

 

I have often heard that the OP2134 is superior to the NE opamps.  I have only swapped an OP2134 for an NE once in a phono preamp, and thought there was a slight improvement, but without proper testing I can not swear to it.

[catman]

G'day mate, this is an interesting one as I think that the NE5532 is still a very good dual op amp and whilst I have a high regard for the OPA2134 they are actually slightly noisier, being FET input.  Regards, Felix.  

[stereo coffee]

20 minutes ago, aussievintage said:

Nothing new under the sun.   Opamps  have been used in DC circuits since they were invented, (as anyone who has worked with analogue computers will attest).  The ideas predate integrated circuitry, in fact the first ones were made with vacuum tubes. Coping with DC offsets etc was well established as that would effect the calculations.   I have spent many hours designing and , mostly, fixing analogue computers   - terribly prone to stability problems.  

 

Amplifying ac signals was, almost, a misuse of them, and they required heavy feedback to obtain a flat frequency response across the audio range.   Indeed many designs still struggle to compete sound-wise with discrete valve and transistor circuits.   I see this as the reason for this modern op-amp rolling phenomenon.   The sale of discrete component built "op-amp" plugin replacements is probably understandable.   Next best to designing the circuit from the ground up I suppose.  After all,  designs use opamps for convenience and cost reasons, not because they sound best.   

 

 

Yes DC circuits is where they afford good application. Walt Jung as example proposing many circuits where op amps assist regulation. The circuit though I am using, and finding marked differences with each device, is quite a bit different.  

 

[aussievintage]

6 minutes ago, catman said:

G'day mate, this is an interesting one as I think that the NE5532 is still a very good dual op amp and whilst I have a high regard for the OPA2134 they are actually slightly noisier, being FET input.  Regards, Felix.  

 

Of course it is a good opamp, or it would not be used so often.  It may still be a compromise, but I  suspect the difference in sound is very small - not enough the warrant the extra cost of the 2134 in a large production run,  and where the majority of purchasers would not even realise the difference.

[stereo coffee]

1 hour ago, aussievintage said:

 

Of course it is a good opamp, or it would not be used so often.  It may still be a compromise, but I  suspect the difference in sound is very small - not enough the warrant the extra cost of the 2134 in a large production run,  and where the majority of purchasers would not even realise the difference.

Yes they would, and explains the manufacturer Texas Instruments own effort  https://www.ti.com/amplifier-circuit/op-amps/audio/soundplus.html

 

It is good some manufacturers care,   and approach to do better.

[stereo coffee]

In my tests so far - all subjective of course, the opa2134 is second, it loses out with depth to the opa2107.  IMO depth portrayal is the key to calling one opamp as better than the other. Without it or less of it, is just not the same. an example being  the reproduction of  natural sound of instruments like piano. 

[aussievintage]

30 minutes ago, stereo coffee said:

IMO depth portrayal is the key to calling one opamp as better than the other. Without it or less of it, is just not the same. an example being  the reproduction of  natural sound of instruments like piano. 

 

It's an interesting point of view, but I feel there is much more to it.  Depth portrayal is too vague and subjective a term for my liking anyway.   Nevertheless, if you are happy rating your opamps just this way, why not.

[roger777]

I recently replaced the LM833 op amp in my phono pre amp with an LM4562. The new op amp presents none of the issues I found with the LM833, which clearly presented audible signal degradation for which the only solution I found was to lower the closed loop gain by about 7dB by changing the ground resistor in the feedback circuit from 390 ohm to 1K ohm. The circuit design of this pre amp is straight out of the data sheets of both ICs, with some additional RF filtering applied on the input.  With the LM4562 installed I was able to reinstate the recommended RIAA equalization (that is restoring the recommended overall gain) and the sound quality and relatively high level of detail indicate the distortion is quite low.

The LM833 was released way more than 20 years ago, and if memory serves, OPA2134 was released around the year 2000, and LM4562 around 2006. I should have done my research a bit better because because I overlooked OPA1612 which is purported to have superior performance compared to LM4562. However, I won't swap to an OPA1612 because I am sufficiently happy with the LM4562 performance.

I also swapped an LM4562 into the tone control circuitry at the same time as replacing a failed bass potentiometer. I almost always bypass the tone controls (as I assume would many discerning listeners) and the bass control carbon potentiometer probably failed because the circuit had the op amp DC bias current passed through it. Anyhow, it sounds okay using an LM4562, that is, for a tone control circuit.

Edited April 20, 2021 by roger777
Clarifing that in referring to Restoring RIAA equlisation, I meant restoring the overall gain from the RIAA feedback circuit.

[catman]

G'day mate, yes the LM4562 is a very fine dual op amp!  Regards, Felix.  

[stereo coffee]

8 hours ago, aussievintage said:

 

It's an interesting point of view, but I feel there is much more to it.  Depth portrayal is too vague and subjective a term for my liking anyway.   Nevertheless, if you are happy rating your opamps just this way, why not.

Thanks, ...  Yes a lot more to it, namely this is DC via many discrete parts but also a opamp , with no physical connection to a audio signal, directly influencing audio signal, with only properties of resistance.  

 

Have a listen to your selection of opamps and see if you can attain realistic  front to back (f/b)  relationship with any of them , a good CD for this is Anat Fort Birdwatching on ECM      https://www.anatfort.com/album/birdwatching/

 

Once you begin hearing f/b, see if you can rate each op amp in your selection for best to worst.  You should find the very best designs also have ability with f/b , where others less well designed miss out on this altogether. F/b is in music, in various degrees  regardless, it  just takes the best designs to reveal it. 

[gmw76]

Hi guys interesting post. I am currently running a Muffsy with dc supplied dual rail. It is fitted with Genuine OPA2134 first stage and LM4562 second. I would like to try some alternatives as suggested but to pose a thought. I wonder how differing capacitor quality and dielectric affect the sound as well. Obviously op amp change in a given circuit can privide side by side comparison. I find the configuration I have very pleasing but sometimes wonder if there is a little too much sshh sshh tst tst treble sounds on vocals.

thank and interesting in trying some alternative chips.

Mark

[gmw76]

A recent email exchange had someone suggesting ad8066 on a dip convertor and newer OPA1612 as being a good step up from the above combo. I have not tried this as yet.

Enjoying the thread

Mark

Edited April 25, 2021 by gmw76

[Addicted to music]

1 hour ago, gmw76 said:

A recent email exchange had someone suggesting ad8066 on a dip convertor and newer OPA1612 as being a good step up from the above combo. I have not tried this as yet.

Enjoying the thread

Mark

Be very aware that the AD8066 is only a 24V max device that is +/- 12 VDC max!  It’s not unity gain as it specify a +1 gain and that PSRR is only 100db 

The OPA 1612 has a higher PSRR .  Heavily compensated for audio so they will remain stable, not only that it’s also stable at unity gain and operates within +/- 18VDC.

AD or ADI devices are a hit and miss and most like the AD8066 isn’t compensated for audio.   
When looking at devices like this ensure supply voltage compatibility, ensure that it’s unity gain as most gain is around the 2X .  And check whether you’re compatibility in single or dual package.
Both need to be on a SOIC to 8pin adapter as both are SOIC platform.

As previous posted you can’t go wrong with the OPA 1612 or 827.  Most Chinese audio produces such as Topping and SMSL will use OPA 1612, easy to get the measurements they need.   Note that the 1612 is a bipolar input device,  to my ears they are smoother than any FET.  

 

Edited April 25, 2021 by Addicted to music

[gmw76]

Wow thanks for that, a bit to digest there. Voltages noted and not an issue as I am using 317 reg so can tweak. From this though looks like OPA1612 worth a shot. Is this better as the second stage and retaining a JFET device (not as susceptible to curent noise from MM) in 1st stage e.g. the OPA2134?

thanks

Mark

[Addicted to music]

5 hours ago, gmw76 said:

Wow thanks for that, a bit to digest there. Voltages noted and not an issue as I am using 317 reg so can tweak. From this though looks like OPA1612 worth a shot. Is this better as the second stage and retaining a JFET device (not as susceptible to curent noise from MM) in 1st stage e.g. the OPA2134?

thanks

Mark

Unless you have an unusual fault condition with an opamp, I doubt changing an opamp will resolve or minimise your noise issue especially in a MM/MC circuit,  it’s due to the low cartridge voltage where resistor thermal noise come into play.

[gmw76]

Hi yes sorry the query was more around which op amp to use in which stage. Not particularly to address a tiny issue I may have. I was led to believe that for MM cartridges, as they are more susceptible to current noise, that Jfet devices are better in first stage. So would the OPA1612 be better in second stage and retain the 2134 in the first. 
Thanks

Mark

[Stereophilus]

Does anyone have thoughts on or experience with Sparkos opamps?