Is it because of lower noise or higher linearity?
Just like a lot of valve amps, and most vinyl pick-ups - it just shows that what people *like* and what is objectively transparent are very different things. But real hipsters prefer cassette tapes anyway...
Don't know the answer to that, but Resistors can make a difference from personal experience ...and was relieved when Mr Westlake (backed me up after a comment) was bemused to find, when designing the AUDIOLAB MDAC that his prototype boards were not all sounding the same...hence discovered it boiled down to the Resistors!? He is the one claiming the Vishay ZFOIL are the best he has come across since; I've only tried Dale-Vishay which warm things up and refine treble compared to SMD that were there.
Personally I appreciate the designers and engineers who make the effort to figure out why something happens or affects the end result - the only way to achieve an optimal result is to understand the underlying mechanisms. After all, without Bruno making the effort of figuring out what was wrong with the way people thought about feedback in class D amplifiers, working out a better model, and sharing it with the engineering community, we wouldn't have the ncore amps we have today.
Is that a Discrete OpAmp you have on there Bavmike!?
Any change for the better?
Ps the Resistors look suspiciously like Dale!
It's an older Burson. Does it sound better? Well it would be a matter of personal taste. Some may like the LM4562 better than the Burson.
I think it's V3 circa 2012 anyways. I like it better. Pretty much the same way everyone reports how the Burson's change the sound vs IC opamps. Plenty of info out there on that.
Dragging this back imto the light as I came over some new infomwhen looking closer at it.
The datasheet for the smps3k states that the power suppøy is not suited for usage below 100hz. A regular speaker operate below 100hz. That would indicate that the smps3k is not suited to drive several nc400 that is going to be used to drive regular full range speakers...
Here's a noise floor comparison of the stock LM4562 buffer compared to my discrete pure class A buffer.
Those look extremely suspicious... The stock LM4562 should look more like the second one. This deviates so greatly from what one would expect that makes to question also the 140dB SNR claims of the 'class A' buffer that were made before. How was that measured? What is the total noise level in Volts at the output of a buffer or the NC500 with this buffer?
I'm waiting to get more answers on how the testing was done. The SMPS1200A700 wasn't used. Agilent E3634A supplies were used to power the buffers. The measurements were taken from the outputs.
These are the 3 plots I received.
1st is stock buffer LM4562 board untouched, besides upgrade to HxR Voltage regs:
2nd plot is same board with a few mods and same LM4562 opamp:
3rd plot is the new discrete class A buffer:
I'm wondering if it was picking up noise from the input during the testing? I sent the boards in with a short cable connected with a Neutrik XLR connector soldered on.
I'm waiting to hear what was connected to the input while the tests were being conducted.
My guess is that the left scale of graphs just referenced/positioned arbitrarily. So, there is no noise pickup or something of this sort on the first one, it is just shifted in vertical scale.
Attachments
Last edited:
- Status
- Not open for further replies.