YES! I'm no expert, but I found this simple fact by just swapping NE5532, LM4562 ,TL072,& LF353 in one simple circuit with the same psu. NONE of them sounded the same & each had it's own sound signature! One sounded bassy (TL072) LFxxx dull, NE5532 quite musical & LM4562 very transparent but NOT musical, what ever that means to you!
Remember, every chip may have production variation, different "Silicon" composition & employ different type caps, resistors diodes etc! Even the part count often vary from chip to chip. I still haven't found 2 chips which have exactly the same part count, slew rate, gain band with, in/out or impedance/resistance etc. If they were, then they would be called just generic clones!
Last edited:
Its a box with an input (a Red Book CD) and an output in just the same way that an opamp is a collection of circuit "building blocks".
Simple answer, yes there is a big difference between opamp "sound" although I would hesitate to say any two opamps of different type numbers are remotely similar. You need to look at such things as the spectrum of the distortion they produce, not just single numbers such as THD which is all the figures lumped together.
I even detect noticeable differences among NE5532s made by Texas, Fairchild,NJM & ON. The old Fairchild sounds the sweetest! However all these have different input configurations & different transistor/part count!
So these are actually not exactly original Fairchild NE5532s clones, but different design versions which go under the same name!
So these are actually not exactly original Fairchild NE5532s clones, but different design versions which go under the same name!
Last edited:
Currently going through the process of production component selection for a product, resistors made a big difference enough for stores to complian about the sound being different from what was demoed to them months before. To avoid the situation occurring again, it was necessary to go through many different component changes. Still trying to determine which solder to use. Then is the effect of PCB and plating. All these have lead me to believe anything can make the sound different, and the best selection may not be what one thinks it should be.
This is probably going to cause contraversy. If you compare even the same opamp made by different manufacturers youll find audible differences. Compare the good old 5534 made in the 80s by signetics to the modern versions by say Onsemi. I recall D Self showed that these parts can differ in specs like THD from the old to new and even between different manufacturers.
Are you aware of how opamps are manufactured ?? The processes of yesteryear are not the same as today. For instance study the way that capacitors inside these chips are formed and how it is done today. The forming process of these capacitors can also differ between manufacturers. These manufacturing differences can make an audible difference although the measurements could be very similar.
The OPA2132 has tighter input offset voltage and drift specifications since it was intended for precision applications. The OPA2134 has looser specifications in these areas because they are typically not crucial in audio applications.
I generally prefer the OPA1642 to these as it has lower noise and better performance in most applications.
hard to have a worse reputation than their type predecessor OP27 - just printing "pro audio" on the datasheet doesn't guarantee golden ears endorsement
the BB opa227 part still has poor slew rate, too bad Groner didn't test them so we could see common mode linearity, I would also suspect R rectification at inputs
the past decade+ shift to more fully complementary processes giving good pnp really should be the new baseline - and SiGe can make some difference in input current noise if you want to optimize BJT input op amps
the BB opa227 part still has poor slew rate, too bad Groner didn't test them so we could see common mode linearity, I would also suspect R rectification at inputs
the past decade+ shift to more fully complementary processes giving good pnp really should be the new baseline - and SiGe can make some difference in input current noise if you want to optimize BJT input op amps
That's a different animal all together since it uses BJT input devices. The OPA227 is a great op amp if the somewhat low slew rate doesn't affect your application. Also, some circuits will benefit from using the decompensated version (OPA228) if the noise gain of the amplifier is 5 or greater.
In terms of BJT input op amps, in DIP packages, it's tough to beat the LM4562. The more I play with that part in the lab the more respect I gain for it.
However, I highly recommend any DIYer to learn to solder surface mount devices, at least SOIC packages. It's easy, fun, and opens up a great number of modern op amp possibilities like the OPA1612 and OPA1642.
It's true, I laugh my *** off, when many you see them people trying to make a damn comparison without even changing/messing/worrying about, Impedance/Feedback/coupling/RC setup rofl, with that said I still hear that non-linearity I think is usual with BJT, 4558 is just ugly, NE5532 is much better, but I seem to like BJT toward the very front end/early in the circuitry, JFETs everywhere/anywhere else just to me start noticing much better Linearity/Dynamic responses. So play around and use both haha, of course widely subjective I don't know/how others hear stuff.
I wanna pick up and try that opa2134 or like TLE2072 see what that's about.
I wanna pick up and try that opa2134 or like TLE2072 see what that's about.
Last edited:
I don't know if its just me, but its like a real Middy/Non-linear Characteristic with BJT, some better than others I guess. Like I said had some old 4558 DDs and I couldn't hardly stand it...messed around with NE5532 and was like wow that's a definite damn improvement, kinda startled me how much more clear/detailed really. souped up more clear/detailed 4558 or something haha
slew rate requirement has to be normalized to output Vswing - you need 10x as much slew rate for a 10Vrms output vs a 1 Vrms Vout swing
so it depends on application whether a 2.3 V/us slew rate is adequate margin over the signal slew rate to avoid increased distortion
and the increasing distortion with frequency is plotted in the datasheet for 3.5 Vrms - the 227 THD is 40x worse at 20kHz than 1 kHz
so it depends on application whether a 2.3 V/us slew rate is adequate margin over the signal slew rate to avoid increased distortion
and the increasing distortion with frequency is plotted in the datasheet for 3.5 Vrms - the 227 THD is 40x worse at 20kHz than 1 kHz
Last edited:
1.9 V/us would just cover a 20kHz sine at 15 volts pk/pk swing... max opamp levels in other words.
If you know the slew rate of the opamp and the maximum output level you require then seeing if the opamp can cope becomes,
f = Sr/(2pi * vpk)
So for an opamp of 1.9 volts slew rate and needing 15 volts peak swing we get 20160Hz which is Dougs 20kHz 'opamp level' figure.
A squarewave test would look something like this give or take. So the next question... is it enough ? and the answer is yes assuming a source such as standard CD.
