Choosing of best sounding OP AMPs for the lowest possible THD+N -really the best Way?

[Tfive]

This is just my opinion: look for regulators that maintain good psrr, low noise and importantly low output impedance over a wide frequency range, for I/v opsmps up into the MHz range. Regulator output impedance at the vcc/vee pins of the opamp that is, so regulator and all the bypass caps with their series inductance (PCB traces) form a system.

 

[bimo]

For line level audio signal, it should be use class A output. I designed discrete op-amp with LTSpice and a friend made the implementation and measure it.
THD at 4,3Vrm is 0.003382%, 1kHz, load 1K Ohm gain 4x, and slew rate almost 12V/uS. Then some people compared it with other discrete op-amp (discrete elliot and Exotique M1 SE) and they prefer my op-amp. I share my design in local audio community.

 

[chermann]

hi
just my 2 cents..
the simple cheap NE5532 is well done here:

Omicron headamp..PSU..etc

 

[brjoon1021]

You guys know your stuff.
Bottom line: OPA1611 is pretty good and should be a no-excitement (bad excitement) upgrade for the 5532 ?

I found that two guys did this very thing and are pretty enthusiastic about the results.

Others recommended the LME49710 because it is known to be good and is also bipolar like the 5534 it would replace.

 

[brjoon1021]

Where should I get the OPA1611 from - to be sure to get real ones, not knock off ? I called some local electronic stores here in California, US and they did not have this item. One offered to order a "non-original" replacement part.
Help please... where can I order the real thing ? I need 2 OPA 1611

 

[lolo6990]

Hi
I recently searched the same. Most online electronic store have them in SOIC format, then you buy an adapter if you need the other format
https://www.digikey.fr/fr/products/detail/texas-instruments/OPA1611AID/2126268
https://www.digikey.fr/fr/products/detail/aries-electronics/LCQT-SOIC8-8/4754588
Otherwise, DIP format, from Audiophonics (but Europe and i don't know about shipping) https://www.audiophonics.fr/en/opa/...2490.html?search_query=OPA1611&fast_search=fs
Hope that helps

 

[Mark Tillotson]

Bottom line: OPA1611 is pretty good and should be a no-excitement (bad excitement) upgrade for the 5532 ?

The devil is always in the details - opa1611/12 have more a lot more current noise than the NE5532/34A, and the decoupling requirements might be more stringent...
The 1611/12 have bias current cancellation circuitry so the advertized (poor) current noise is only attainable if both inputs see the same impedance, which is almost never a realistic situation.

So if you were expecting the OPA1611 to be an unexciting replacement for the NE5534A in an MM phono preamp you'd be rudely surprized at the higher noise... You might even see oscillations if the decoupling wasn't adequate (the NE5532/34 only require one cap from V+ to V-, other opamps require decoupling to ground as well to prevent instability/oscillation)

 

[dotneck335]

For phono applications, the old NE5534AP is tough to beat, especially since they are still only slightly over a buck each. The OPA2228 has among the lowest noise figures for phono use (~70db S/N on a Shure V15), whilst still being pretty fast (11nV/uSec) and very low THD (-126db) and not too pricey (~ $6 a pop).

 

[Tfive]

To put things into perspective:

	input current noise density		input voltage noise density
OPA1611	2.5pA @ 30Hz	0.6pA @1kHz	7uV @ 30Hz	5.5nV @1kHz
NE5534	1.5pA @ 30Hz	0.4pA @1kHz	4uV @ 30Hz	3.5nV @ 1kHz
OPA1611	3.0pA @ 10hz	1.7pA @1kHz	2uV @ 10hz	1.1nV @ 1kHz

# values are from TI datasheets. ## note 30Hz values for 5534 and 10Hz values for 1611

I guess the input current noise is dominant for phono applications because of the high input impedance?
Might the OPA1611 still be better for low impedance applications due to it's lower input voltage density?

 

[Mark Tillotson]

Absolutely - around 500 ohms to 2k is its best low noise performance. But its a common trap to forget the importance of current noise or of source impedance and fixate on voltage noise only.

 

[dotneck335]

Yeah, the crossover point where current noise begins to dominate over voltage noise is at about 3KΩ source impedance.

 

[Tfive]

please note that first row of data is the NE5534 (w/o SA). I goofed up there. notified mods team to correct my goofup.

 

[brjoon1021]

On this thread there have been the cNautionary posts especially brimming with technical and then the two guys who said, "yep, I did it it sounded great". So those of you who are concerned about the technicalities would you suspect that going from the 55XX to a 1611 is a fairly decent fit (Not causing the maladies you said could happen) because as swapping op amps could go this is a pretty straight and easy move ?
I'm hoping to do this soon if I can find a good source for two 1611s. It bothers me a little bit though that my PCB revision is not same as the two gentlemen who made the change so I don't know if that should be much of a consideration or not.

 

[Cal Weldon]

SA removed from the original table as requested.

 

[Tfive]

To put things into perspective:

	input current noise density		input voltage noise density
NE5534	2.5pA @ 30Hz	0.6pA @1kHz	7uV @ 30Hz	5.5nV @1kHz
NE5534SA	1.5pA @ 30Hz	0.4pA @1kHz	4uV @ 30Hz	3.5nV @ 1kHz
OPA1611	3.0pA @ 10hz	1.7pA @1kHz	2uV @ 10hz	1.1nV @ 1kHz

I probably was unclear, so I will quote my own message here. Table should look like this. >Probably I should have done this in the first place? Mods please comment or edit to your likings. Thanks!

 

[dotneck335]

The OPA1611 is STILL going to be about 2 db NOISIER than an OPA2228 in a phono application although a bit faster (27 vs 11), if you think that speed improvement is audible from a phono record. Also, the 2228 is still available in a PDIP package, which the 1611 is NOT.

 

[Tfive]

I don't disagree, especially in phono applications. Just wanted to point out that the labeling of my table was wrong

 

[Thx1326]

I am also a hobbyist and our local audio club is in the midst of rebuilding several (6) vintage Kenwood Basic C2 preamplifiers. We only want to do this once and are trying to reach a consensus on what upgrades to use for the 4 opamps in the unit. We have been considering both the OPA1642 and now the OPA1656. There are 4 sections within the preamp with the following original opamps:

Flat Amp Section: NJM2041D (original)

Phono Preamp Section: NJM4560D (original)
Tone Control Board: NJM4560D (original)
Filter Board: NJM4560D (original)

The general past thoughts have always been to favor JFET (OPA1642) over CMOS (OPA1656 opamps. The 1642 has been described as having "silky smooth highs" while the 1656 is getting lots of praise but I cannot find any real comments on its audio qualities.

Would either of these 2 opamps be better suited over the other in any of the above sections and in that case which one for each section?

I would appreciate any comments you may have and/or any other recommendations.

I am assuming that either of these two newer opamps would give improved performance over the original NJM's?

Thanks in advance.

 

[bucks bunny]

One more time the can of worms concerning op-amp rolling is opened.
There are tons of opinions from different humans around.
At the very end they are all anecdotes with unproven results.
You better do your own comparisons based on technical data (my choice) or listening tests.

 

[Drbulj]

Hi, might be attached pdf might contribute.
Personally I don't think commonly blamed cheap opams found in commercial pro gear are that bad at all for usual applications.
Then, for extremely high gain (Mc phono or mic preamp) fully balanced approach or Ina chips are probably better way to go... Than again, what's the source impedance? What's the load the thing needs to drive? There is no answer what's the best without specifying exactly what the thing is supposed to do.... Most of recorded music we listen to passed through several NJM's and NE's (I'm so lame, this was said 1000's of times before 🥴), so why bother another 1 or 2?