Ditto, looks like a typo.
Funny, I thought I got that idea from Nelson, but I can't find the source now.
Maybe I just extrapolated from the inverting version, which he showed here in the "7 Easy Pieces" thread.
@Mooly
Here's another thread re throwing away gain, with some interesting links: http://www.diyaudio.com/forums/chip-amps/232233-lm3886-unity-gain.html#post3429334
Here's another thread re throwing away gain, with some interesting links: http://www.diyaudio.com/forums/chip-amps/232233-lm3886-unity-gain.html#post3429334
Ditto, looks like a typo.
Funny, I thought I got that idea from Nelson, but I can't find the source now.
Maybe I just extrapolated from the inverting version, which he showed here in the "7 Easy Pieces" thread.
Well, wherever you got your idea from I think it should work
Next question then... if this is an idea worth pursuing for using opamps and extracting the best subective performance, what sort of values might you put on R1 and R2 as a ballpark starting value ?
Attachments
As I posted before, it is a completely stupid idea to place a resistor between inverting and non-inverting inputs, exactly because you throw away feedback=error correction. The major effect is that distortion will increase!
Manufacturers recommend this setup to increase distortion in order to push distortion above the measuring limits of commercially available gear.
Only wacko audiophiles would consider this an advantage. If you want distortion in the first place, I suggest you start out with the worst possible opamp you can find.
Manufacturers recommend this setup to increase distortion in order to push distortion above the measuring limits of commercially available gear.
Only wacko audiophiles would consider this an advantage. If you want distortion in the first place, I suggest you start out with the worst possible opamp you can find.
Pavel
I would like to ask you if you find the following example correct, that is to say, to conform if I understood you well.
I do a wave recording of a preamplifier output with my sound card, 48KHz sampling.
I set the 0db as 1Vrms.
The recording I get with no signal on pre input is the one on the upper part of your attachment:
This noise signal has Vmax:0.348Vp+, Vmin:-0.340Vp-, Vpp=0.688V, Vrms=0.243Vrms.
Is it ok to say that this recording has a S/N ratio of –12.29dbV [ SNR db=20LOG (Asignal/Anoise) ] ?
Now, doing an FFT (65536 points) over this recording, I get a result as on the lower part of your attachment. This shows a "noise bottom" of almost –58db. Subtracting the 45db which is the difference between "noise bottom" and "integrated noise over 24000 Hz" as you say, we get to a figure of –13db. This is the "noise over the 24KHz band" and is very close to the S/N figure I wrote above.
George
I would like to ask you if you find the following example correct, that is to say, to conform if I understood you well.
I do a wave recording of a preamplifier output with my sound card, 48KHz sampling.
I set the 0db as 1Vrms.
The recording I get with no signal on pre input is the one on the upper part of your attachment:
This noise signal has Vmax:0.348Vp+, Vmin:-0.340Vp-, Vpp=0.688V, Vrms=0.243Vrms.
Is it ok to say that this recording has a S/N ratio of –12.29dbV [ SNR db=20LOG (Asignal/Anoise) ] ?
Now, doing an FFT (65536 points) over this recording, I get a result as on the lower part of your attachment. This shows a "noise bottom" of almost –58db. Subtracting the 45db which is the difference between "noise bottom" and "integrated noise over 24000 Hz" as you say, we get to a figure of –13db. This is the "noise over the 24KHz band" and is very close to the S/N figure I wrote above.
George
Last edited:
A bit odd that, even at a thread with engineering appeal, all talk is about the 'best' opamp.
After 22 years of OPA627/637, an OPA827 at half retail cost looks a step up (despite that I get BP's, that somewhere somehow fell out off the truck, for €5/pc).
More comprehensible would be to list the minimum requirements of an opamp for various applications.
(Mr Morin : beauty of living here is that everything on TV is broadcasted in the original language version. Subtitles, no Voice-over. Result is that some cope with English and German at highschool level by age 12, French/Spanish/Italian at age 16, including a range of dialects & slang, e.g. some Gaelic/Welsh. Just think, an average American gloats 35 hours of TV a week, almost twice as much as the average here. US citizens could be the polyglots of the world, instead of scoring a high grade on the functional illiteracy list. Y'all have a spanking good time, Ya hear !
)
After 22 years of OPA627/637, an OPA827 at half retail cost looks a step up (despite that I get BP's, that somewhere somehow fell out off the truck, for €5/pc).
More comprehensible would be to list the minimum requirements of an opamp for various applications.
(Mr Morin : beauty of living here is that everything on TV is broadcasted in the original language version. Subtitles, no Voice-over. Result is that some cope with English and German at highschool level by age 12, French/Spanish/Italian at age 16, including a range of dialects & slang, e.g. some Gaelic/Welsh. Just think, an average American gloats 35 hours of TV a week, almost twice as much as the average here. US citizens could be the polyglots of the world, instead of scoring a high grade on the functional illiteracy list. Y'all have a spanking good time, Ya hear !
)
Last edited:
George, I always run a calibration with white noise well above the floor since the noise BW per bin depends also on the windowing function. 604K = 100nV/rt-Hz.
yes, but we are not discussing the many and varied permutations of soundstage and musical transparency resulting from swapping one opamp for another. obviously, there are preferred technical reasons for using one opamp or another, thats just good engineering, nobody ever said that all opamps of all types sound/function the same in every situation, or did they?
that would be pretty dopey
LME49990 is a bit of a bargain at less than $3
that would be pretty dopey
LME49990 is a bit of a bargain at less than $3
Yes, the measurement in post
http://www.diyaudio.com/forums/anal...ch-preamplifier-part-ii-3722.html#post3426423
#37218 was white noise well above measurement system noise floor.
George, I am sorry, I am in time pressure now.
http://www.diyaudio.com/forums/anal...ch-preamplifier-part-ii-3722.html#post3426423
#37218 was white noise well above measurement system noise floor.
George, I am sorry, I am in time pressure now.
A note on IC op amps:
It is almost impossible to effectively recommend a specific op amp, unless you know the conditions that it is being used.
For example:
Is the drive Z low, medium, or high?
Is the output load resistance low, medium or high?
What is the output swing that is necessary, and what are the available power supply voltages?
Is this application for battery operation?
I know it sounds simple, but nobody here DEFINED the problem, before entering their choices, AND your choices are all over the map. '-)
It is almost impossible to effectively recommend a specific op amp, unless you know the conditions that it is being used.
For example:
Is the drive Z low, medium, or high?
Is the output load resistance low, medium or high?
What is the output swing that is necessary, and what are the available power supply voltages?
Is this application for battery operation?
I know it sounds simple, but nobody here DEFINED the problem, before entering their choices, AND your choices are all over the map. '-)
Somehow over the rainbow it did sound like that to me.
"AD797/OPA827/AD4627/OPA1642/LME49990/LM4562/NE5532/OPA627/AD8066/LT1358/OPA1641/OPA1644". Hurses for Curses.
The 49990 is a good example, in a fully balanced bit-high DAC a bucket cheaper than AD797, and preferable. For phono stage input, not so. Without listening.
A moment of major agreement. And here you have tweak audio at its best, no-one bothering to define the problem or the target, but full of opinions anyway.
- Status
- Not open for further replies.