For usual home audio line levels and impedances? Nope, no advantage one way or the other, except marketing.
Reading the thread title, I thought you meant LM1036 or the like.
Please note that what you wanted to know relates to operational amplifiers (op-amps). You can use those to build a preamp and that is commonly being done, but not every preamp necessarily makes use of them, not even when you take their current-feedback relatives into account.
Discrete OPs have a theoretical dissipation advantage, so they can push higher currents while staying in Class A, for example. That doesn't necessarily translate to great performance though, in fact a few of them are downright bad (Samuel Groner's set of measurements includes some, too). Those from Sparkos Labs appear to be some of the better offerings, though reading the datasheet between the lines reveals potential issues with common-mode and input impedance distortion and possibly current noise as well - not like I'm surprised given the equivalent schematic. These pimped with a cascode frontend with common-mode bootstrap could be quite something.
Please note that what you wanted to know relates to operational amplifiers (op-amps). You can use those to build a preamp and that is commonly being done, but not every preamp necessarily makes use of them, not even when you take their current-feedback relatives into account.
Discrete OPs have a theoretical dissipation advantage, so they can push higher currents while staying in Class A, for example. That doesn't necessarily translate to great performance though, in fact a few of them are downright bad (Samuel Groner's set of measurements includes some, too). Those from Sparkos Labs appear to be some of the better offerings, though reading the datasheet between the lines reveals potential issues with common-mode and input impedance distortion and possibly current noise as well - not like I'm surprised given the equivalent schematic. These pimped with a cascode frontend with common-mode bootstrap could be quite something.
Hi,
I agree that it doesn´t make much difference replacing one (elaborated) circuit structure with basically the same, just simpler one and of which most specimen reside on a much less refined level.
The Q should imho rather be, if a different circuit structure could be advantageous or better suited to our hearing than a OPAmp.
It is not at all intuitive to design a circuit that does anything but linear amplification in first place and then have to use brute force to make it an (t)error-amplifier. 😉
Best examples imho are Buffers, where a simple single transistor or tube functions in its natural way.
Why add hundreds of active parts to achieve DC-gains in the millions, just to knock this down to 1 again?
Yes, the data sheets impress with more 0s than necessary, but our hearing is rather crappy in its quantitative performance, and a 0 more or less doesn´t talk better sonic quality.
jauu
Calvin
I agree that it doesn´t make much difference replacing one (elaborated) circuit structure with basically the same, just simpler one and of which most specimen reside on a much less refined level.
The Q should imho rather be, if a different circuit structure could be advantageous or better suited to our hearing than a OPAmp.
It is not at all intuitive to design a circuit that does anything but linear amplification in first place and then have to use brute force to make it an (t)error-amplifier. 😉
Best examples imho are Buffers, where a simple single transistor or tube functions in its natural way.
Why add hundreds of active parts to achieve DC-gains in the millions, just to knock this down to 1 again?
Yes, the data sheets impress with more 0s than necessary, but our hearing is rather crappy in its quantitative performance, and a 0 more or less doesn´t talk better sonic quality.
jauu
Calvin
The hidden issue is interference rejection. High end discrete designs use FETs or tubes which are better for that than bipolar junction transistors either discrete or integrated. In real pro gear advantage is taken of IC's bandwidth and the input stages are actually low pass filters set well above the audio band.
Now IC's have good power supply rejection ratios that drop as frequency increases so power supplies should be well filtered not just for ripple but actually up to tens of megahertz.
Cover those bases and for 10% if the cost of high end units you can build your own.
Now IC's have good power supply rejection ratios that drop as frequency increases so power supplies should be well filtered not just for ripple but actually up to tens of megahertz.
Cover those bases and for 10% if the cost of high end units you can build your own.
- Status
- Not open for further replies.