The best sounding audio integrated opamps

If your op amp is set for a heck of a lot of gain and/OR if its circuit is in inverting mode, then decompensated may work well or even better. It IS more dynamic, but in the wrong circuit that would be unpleasant noise. Decompensated shouldn't be used as a buffer--it is a powerful preamp.

If your op amp is a buffer or low gain or low gain non-inverting then you need the compensated version. If a compensated op amp was set to higher gain, then its possible (not guaranteed) for it to be a bit dull or even rear numbingly dull. It may also simply be "laid back" maybe (or maybe not) with possibly lesser dynamics. Its also possible or even probable that the lesser dynamics are less shout.

These factors depend a lot on which circuit supports which opamp. However, it may be possible to identify the type of circuit just by differences in reaction to the opamp. That's unnecessary because its very easy to measure the resistors (located at point blank range) with an ohmmeter.

For instance if between one of the opamps inputs and its output is a short (ohmmeter reads same as leads crossed "0") then you must have a compensated opamp because that circuit is a buffer (doesn't amplify).

And, if you find a resistor between one of its inputs and its output then it is amplifying. Find its partner resistor that attaches to the same input and measure it too. Divide them with the calculator and you have a very close approximation of the gain factor.

A single channel op amp has two inputs, inverting and non-inverting. One of these inputs is used to set the gain factor. The other is used to input the signal from the source and that will tell you what kind of amplifier it is (inverting amplifier uses inverting input to put in signal from the source).
 
The LT1126 is stable in gains of 10 or more... Not for your soundcard(s), leeperry. Maybe you should give a try to the OPA2228, instead...

And the LT1124 looks good. I was tempted to try it a while ago. I don't think it has a chance to challenge two LT1028 though...but it could be a fine dual opamp.



BTW... I just bought myself a new digital interconnect... Supra TRICO-RCA - From £39.50 - HiFiCables.co.uk

And I want to break out that Audioquest King Cobra (analogue interconnect) that I'd bought myself some time ago...for the better times. :spin:
 
ok thanks Daniel for the detailed technical explanations!

can I still try it? or might it kill something?

and Andrea, do you think LT1013ACN8 is worth a shot?

even w/ fresh ears, LT1124 is really AMAZING...it does everything the LM4562 promised, but failed. totally amazing, so many details!
 
Last edited:
I have received & wanna try the THS4631... but not in my DAC (for obvious reasons 🙂 ) but in my 24V "cmoy". Hopefully it's stable... if so, it will sound good because it has a lot of output current. Otherwise, I might try the THS4601 which was a good FET opamp to me, again with a good output current.
 
ok thanks Daniel for the detailed technical explanations!. . .

That was more of an application note. 😉

Next step is HeadWize - Breaking News and pick out a fun opamp project to put together. Perhaps a preamp or buffer type would be helpful?

After that is another opamp project, except that the opamp is larger and can make use of the previous preamp/buffer project. Next up. . . Its the LM1875 power amp. This works quite well with 90+db speakers (like Tang Band W6) and its capable of authentic hi-fi sound, very, very easily. Its also expandable. . .

And, no chip makes as much pretty bass and sharp mids like the LM3886; so, naturally this can be well employed to power a pair of 10" or 12" woofers to add powerful clear bass to the previous LM1875 project.

Well, there's a short description on a high quality multi-amp system that uses nothing more than simple opamps. Your experience with preamps and buffers gives you a head start, because larger scale opamps aren't different except that they're larger. 😉
 
In that case, it would be nice to know what the supporting circuit looks like so that everyone could enjoy the LT1028 as much as you do.

I showed it through an image one page back...

Don't know if those are precisely the component values though (as I said in that post).


Also I've replaced the 220uF power supply electrolytic shown, with a Rubycon ZL 680uF 16V + Wima MKS 330nF + Wima MKS 100nF. All the caps in the signal path are film.
 
I haven't received yet either my AD797BN or my second pair of 1028ACN8...but I got my browndogs 🙂

I'll try replacing Earth by this, to see what the fuss is all about:

An externally hosted image should be here but it was not working when we last tested it.


I know I should be soldering them but I'm already using an extra gold plated socket as riser, and this is already pretty high...not sure an additional soldered socket would fit, and contacts seem OK so far.
 
Last edited:
What? You have placed the chips on the browndog and you intend to be using the thing without any solder? Aargh 😱

And what? You even add sockets as risers? :gnasher:


You should've sourced a pair of LT1028CS8, soldered properly on the proper browndog... That would've been a neat job. Yes, there's always one socket in between... which as I felt seems to compromise the sound of this chip a little, sensitive as it is.
 
Oh and previously I forgot to mention that in my DAC's output stage circuit there are also a couple of SMD sort-of-inductors in series with the output (to avoid series resistors I guess). That guy ("rds" of head-fi, who drew that schematic) has removed them (so he didn't even draw them in the schematic), as I guess he didn't see them as a good thing... I surely would disagree. :no:
 
Last edited:
hehe ok, well I'll be forced to solder sockets on the browndog's then 😀

I'll check if that'd fit....you talk about signal loss due to sockets, but the 2*SOIC8 single>1 DIP8 dual Browndog's are very poorly engineered anyway...if you were to compare to those higher quality chinese adapters, I'd be less dubious 🙂
 
hehe ok, well I'll be forced to solder sockets on the browndog's then 😀

I'll check if that'd fit....you talk about signal loss due to sockets, but the 2*SOIC8 single>1 DIP8 dual Browndog's are very poorly engineered anyway...if you were to compare to those higher quality chinese adapters, I'd be less dubious 🙂
No, the single SOIC adapter is rather poorly engineered [long contorted traces]. The twin SOIC adapter looks fine to me (e.g. short & wider than the rest power traces).

What is no good about sockets is the contacts (made with a few separate metallic elements making contact only out of pressure), plus the plastic (surely worse dielectric than vetronite (Italian word)). In practice, everything 🙂
 
Last edited:
It may work but certainly compromises the sonic quality of the chips. 🙂

The browndogs are already a compromise to begin with... and then add all the sockets... 😱


The extensions of Audio-GD apparently are compatible with the opamps they design. Apparently those opamps don't have a high bandwidth nor they are very fast slewing. And/or they simply want to sell as many as they can...🙄
 
Could be, in your application, with your music and for your taste...

For me the OPA-Earth (At +/-15V and in a rather similar circuit to the one shown some post ago) was good except with too little color saturation and also with a tendency to make (female) vocals sound...compressed, dry, I don't know.

Plus, if I set the output of the SCDAC05 near its max (which is 3V RMS), it began to clearly distort some vocal peaks on a Lucinda Willams record [West]. 🙁

Oh and I had connected the two single OPA-Earths' ground wires where appropriate, as recommended.
 
Last edited: