[scootsit]

Quote:

I think you guys are confused as to what a preamp is/does. a preamp is primarily for 4 things
1. providing multiple inputs and source switching (not all do, but most do)
2. providing a high input impedance and a low output impedance, thus buffering the source from the amp
3. provide volume control
4. provide voltage and/or current gain
 
the ODAC does not include a preamp, but it does not need one
no, most dacs do not have an internal opamp, only the voltage output ones do, perhaps most do fall into this category these days, but many high performance dacs are still current out.
 

You are absolutely right. Sorry, I haven't slept...at all.

 

[Chris J]

Quote:

Speaking of technical advances, a device the size of a Buick was just hurdled 8000 mph at another planet about 154 million miles away, and lowered at less than 1m/sec via sky crane. Also, I believe it landed a few meters off mark.
http://blogs.plos.org/retort/2012/08/05/satisfying-curiosity-preparing-for-the-mars-landing/
 
Sorry, I know this is very off topic, but my mind is blown.

 
Consider my mind blown as well!
 
(sorry, just lurking!)

 

[ulogin]

Just thought some of you might be interested:
How to: O2 with Triple Gain Switch

 

[DefQon]

Quote:

 
 Also, I don't think you can build and case up a B22/S22 for $500.  Just sayin'.
 
 

You can build a 2 channel passive ground B22/S22 for $100 more. Enclosure is a basic one but yet again, it's no GS-X look alike at the end. Just sayin'.

 

[Sandman65]

Read through about half this and all drama aside I still intend to try a O2 or C421.
 
I'd really like to see an impartial measurement shootout vs. the QuickStep.... accompanied by a blind listening panel.
 
At it's pricepoint, options and intended use, I consider the QuickStep to be TOTL, SOTA, etc...

 

[fubar3]

Did anyone notice, the NJM chips in the O2, are not said to be for Audio? The TI/National Audio buffers like BUF634 cost around $12,  the NJM4556 cost 63 cents. So anyone who cannot discern a significant difference in the listening experience is getting a bargain.
 
I can recognize the difference between my modest headphones but not an amplifier difference.  Perhaps some listeners are sensitive to slew rate which the audio IC buffers chips have in abundance.
 

 

[scootsit]

Quote:

Did anyone notice, the NJM chips in the O2, are not said to be for Audio? The TI/National Audio buffers like BUF634 cost around $12,  the NJM4556 cost 63 cents. So anyone who cannot discern a significant difference in the listening experience is getting a bargain.
 
I can recognize the difference between my modest headphones but not an amplifier difference.  Perhaps some listeners are sensitive to slew rate which the audio IC buffers chips have in abundance.
 

Is it an NJM4556? If so, the datasheet is speced for audio. It is also the IC in the Grado headamp.

 

[stv014]

Quote:

 
Is it an NJM4556?

 
Yes. More precisely, it is two NJM4556 buffers in parallel, with a 1 Ω serial resistor on the output of each.

 

[nick_charles]

Quote:

 
Yes. More precisely, it is two NJM4556 buffers in parallel, with a 1 Ω serial resistor on the output of each.

..... and it measures extremely well for the purpose even if it is deprecatingly refereed to as a "jelly bean" chip, but in circuit it performs well even if it does not have the caché of a Burr-Brown or Analog devices opamp, but I cannot mention where I discovered that 

 

[fubar3]

Quote:

 
Yes. More precisely, it is two NJM4556 buffers in parallel, with a 1 Ω serial resistor on the output of each.

The O2 has the NJM4556A which can drive 70ma into a 150ohm load, according to the JRC datasheet.  It has an 8mhz gain-product and 3v/ms slew rate. So I am curious about the sonic difference between these economy chips and the expensive audio buffers. Can anyone hear the difference in an objective listening test?

 

[mikeaj]

Quote:

The O2 has the NJM4556A which can drive 70ma into a 150ohm load, according to the JRC datasheet.  It has an 8mhz gain-product and 3v/ms slew rate. So I am curious about the sonic difference between these economy chips and the expensive audio buffers. Can anyone hear the difference in an objective listening test?

 
People say they can but don't really prove it (because doing controlled listening tests is a pain, regardless of whether or not it is possible).
 
Actually, if you want something like 300 mA out of it, then you should certainly hear a difference.  NJM4556 can do plenty over 70 mA into lower impedances, still with low distortion, more like 150-200 mA (peak).  I don't know which modern-day headphones other than HE-6 need current like that, though.

 

[fubar3]

Quote:

 
People say they can but don't really prove it (because doing controlled listening tests is a pain, regardless of whether or not it is possible).
 
Actually, if you want something like 300 mA out of it, then you should certainly hear a difference.  NJM4556 can do plenty over 70 mA into lower impedances, still with low distortion, more like 150-200 mA (peak).  I don't know which modern-day headphones other than HE-6 need current like that, though.

What about slew rate?  There seems to be some hype or emphasis on that factor. The LM49610 can do 2000v/m-sec which is 600 times faster the NJM4556. What difference would it make to a headphone enthusiast?

 

[stv014]

Quote:

 
What about slew rate?  There seems to be some hype or emphasis on that factor. The LM49610 can do 2000v/m-sec which is 600 times faster the NJM4556. What difference would it make to a headphone enthusiast?

 
Slew rates in the range of thousands of Volts per second are mostly a marketing feature for driving dynamic headphones. The slew rate required for a sine wave at frequency 'f' and voltage 'Vp-p' (peak to peak) is ℼ * f * Vp-p V/s. Therefore, the NJM4556 in the O2 allows for a maximum frequency of about 47.7 kHz at its 20 Vp-p output before slew limiting occurs. Of course, in practice the distortion usually starts to rise already before the limit is reached, but for typical CD quality music containing frequencies up to 22050 Hz, the O2 has enough "headroom" as far as slewing is concerned (it is also useful for not perfect reconstruction of the signal by the DAC filters, but modern oversampling DACs with a decent analog filter stage are not too bad in this aspect).
 
In the case of buffers that are intended to be used in a feedback loop, like the LME49600, the very high speed is useful to minimize the loss of phase margin due to the addition of an extra stage (i.e. the open loop frequency response of the op amp hopefully rolls off to 0 dB before the buffer starts to introduce a significant phase shift).

 

[fubar3]

Quote:

 
Slew rates in the range of thousands of Volts per second are mostly a marketing feature for driving dynamic headphones. The slew rate required for a sine wave at frequency 'f' and voltage 'Vp-p' (peak to peak) is ℼ * f * Vp-p V/s. Therefore, the NJM4556 in the O2 allows for a maximum frequency of about 47.7 kHz at its 20 Vp-p output before slew limiting occurs. Of course, in practice the distortion usually starts to rise already before the limit is reached, but for typical CD quality music containing frequencies up to 22050 Hz, the O2 has enough "headroom" as far as slewing is concerned (it is also useful for not perfect reconstruction of the signal by the DAC filters, but modern oversampling DACs with a decent analog filter stage are not too bad in this aspect).
 
In the case of buffers that are intended to be used in a feedback loop, like the LME49600, the very high speed is useful to minimize the loss of phase margin due to the addition of an extra stage (i.e. the open loop frequency response of the op amp hopefully rolls off to 0 dB before the buffer starts to introduce a significant phase shift).

Well then, if I may extrapolate a bit, one can spend a certain amount of money for "good" hi-fi, and then spend 10 times more, for a possibly tiny improvement.

 

[antberg]

Sub.
Even thought a little late..