Inverting gives lower distortion because there is no common mode signal. The positive input is tied to ground. Common mode distortion is a type of distortion that the feedback loop in a non inverting stage can not suppress because it appears on both inputs. The OPA134 is notorious for that. It depends on the internal structure of the Opamp. Opamps with cascoded inputs like the OPA627 show less of this distortion. To learn more about this read Douglas Self or download Samuel Groners survey of Opamp distortion. The disadvantage of inverted is that when you want a high input impedance one feedback resistor appears in series with the signal so raises noise. You can work around this when you use an input buffer. Self has shown how that works in his Precision Preamp. I made a successful DC coupled version of it with a current feedback OPamp and have described this in detail on my MPP thread.
I could than use a jfet voltage follower as a buffer and an inverting configured opamp.. But the signal comes out inverted... Is it irrelevant or do I need to have an inverting amp to get the signal on the correct fase ?
You never know how the source material is recorded. Some people claim though that they can here absolute phase with impulsive material like drum stokes and the like. The most simple would be to invert the speakers.
Joachim hi, what 5532s and 5534s you use and find good from what is offered today as far as brands are considered?
The OPA134 is good.
Apart from the obvious 180 degree phase shift the inverting configuration does have it's merits... I prefer it and used it here post #2 adding another stage to bring the phase shift back to zero,
http://www.diyaudio.com/forums/solid-state/119151-my-mosfet-amplifier-designed-music.html
Would 180 degree phase shift be audible... some say yes and others no. It's easy to try at no cost, just swap both leads (the - and +) around to both the speakers on any audio system. That is exactly the same... you have inverted the phase.
If you have a system (preamp) that inverts then swapping the speaker leads is another way to bring the phase back.
There are many reasons for and against the inverting stage.
As Joachim Gerhard mentions, the inverting mode eliminates "common mode" distortion (where the same signal is seen on both input pins as in the non inverting amp). That's just one though.
There's quite an interesting thread here, but don't get to hung up on it all, either configuration correctly designed can be truly excellent.
http://www.diyaudio.com/forums/solid-state/6558-opamp-inverting-input-sounds-better.html
Apart from the obvious 180 degree phase shift the inverting configuration does have it's merits... I prefer it and used it here post #2 adding another stage to bring the phase shift back to zero,
http://www.diyaudio.com/forums/solid-state/119151-my-mosfet-amplifier-designed-music.html
Would 180 degree phase shift be audible... some say yes and others no. It's easy to try at no cost, just swap both leads (the - and +) around to both the speakers on any audio system. That is exactly the same... you have inverted the phase.
If you have a system (preamp) that inverts then swapping the speaker leads is another way to bring the phase back.
There are many reasons for and against the inverting stage.
As Joachim Gerhard mentions, the inverting mode eliminates "common mode" distortion (where the same signal is seen on both input pins as in the non inverting amp). That's just one though.
There's quite an interesting thread here, but don't get to hung up on it all, either configuration correctly designed can be truly excellent.
http://www.diyaudio.com/forums/solid-state/6558-opamp-inverting-input-sounds-better.html
Have anyone tried LF442ACN to replace NE5532P?
I just did that this morning after getting some of them at a local shop. I replaced four NE5532P in the output filter boards for both channels in my Denon CD3520. The IV converter is already using OPA627 for couple years.
The difference are:
1. The sound stage is about 25% larger
2. The sound stage is many time more stable for loud and complicated music
3. Tighter bass
4. Sweeter mid but I think they are close though.
5. More sparkles at high
6. Overall the sound is sweeter.
7. It handles low sound level music much better that I could tell the type to of instrument. NE5532P can only tell me an instrument exists there.
The difference was heard instantly by swapping the amps back and forth.
Of course, you mileage may vary... just wonder if anyone of you have tried this amp and is willing to share your experience.
A3K
I just did that this morning after getting some of them at a local shop. I replaced four NE5532P in the output filter boards for both channels in my Denon CD3520. The IV converter is already using OPA627 for couple years.
The difference are:
1. The sound stage is about 25% larger
2. The sound stage is many time more stable for loud and complicated music
3. Tighter bass
4. Sweeter mid but I think they are close though.
5. More sparkles at high
6. Overall the sound is sweeter.
7. It handles low sound level music much better that I could tell the type to of instrument. NE5532P can only tell me an instrument exists there.
The difference was heard instantly by swapping the amps back and forth.
Of course, you mileage may vary... just wonder if anyone of you have tried this amp and is willing to share your experience.
A3K
Salas, i have not answered post 44. I am using NE5534A from TI. In the FPS i did for Linear Audio they are used inverted. I also build my TDS time domain subsonic filter with it. I only use A grade because noise and offset are better specified. Both circuits work excellent. The TDS filter was presented at CES in a 30.000 $ system and most people preferred the sound with the filter in the system without actually knowing whats inside. Self published measurements on various NE5532 in Elektor and the TI version was actually the worst. If you can hear 0.002% distortion instead of 0.004% i do not know. Some say the japanese ones are the best.
audio3000, that can not be. The LF442 is slow and noisy. Maybe you have another problem in your CD player that is hidden by the noise and slow slew. I can recommend the OPA1641 when you want to try a modern Fet Opamp instead of the NE.
audio3000, that can not be. The LF442 is slow and noisy. Maybe you have another problem in your CD player that is hidden by the noise and slow slew. I can recommend the OPA1641 when you want to try a modern Fet Opamp instead of the NE.
Hi Joachim, Thanks for your comment "audio3000, that can not be. The LF442 is slow and noisy. Maybe you have another problem in your CD player that is hidden by the noise and slow slew. I can recommend the OPA1641 when you want to try a modern Fet Opamp instead of the NE."...
I did wonder about this too when I compared their spec... 5532 is much better in terms of Slew rate and GBW... the only spec that LF442 is much better is the input current (not voltage) noise... And it seems on the spec that LF442 is a much simpler design. That could make a difference in sound.. But I would not know for sure as the spec doesn't show the exact circuit diagram..
I have no good explanation why LF442 sounds better in my CD purely from technical spec perspective...
I don't think I can completely rule out the possibility my CD causes this... but the chance is not high though... it's still playing music in the background.. ;-)
Will enjoy the nice sound it produces for the moment...and that's what matters to me..🙂
Will try OPA1641 when I have a chance to get some... and will be trying some other idea in the mean time...
thanks again...
A3K
I did wonder about this too when I compared their spec... 5532 is much better in terms of Slew rate and GBW... the only spec that LF442 is much better is the input current (not voltage) noise... And it seems on the spec that LF442 is a much simpler design. That could make a difference in sound.. But I would not know for sure as the spec doesn't show the exact circuit diagram..
I have no good explanation why LF442 sounds better in my CD purely from technical spec perspective...
I don't think I can completely rule out the possibility my CD causes this... but the chance is not high though... it's still playing music in the background.. ;-)
Will enjoy the nice sound it produces for the moment...and that's what matters to me..🙂
Will try OPA1641 when I have a chance to get some... and will be trying some other idea in the mean time...
thanks again...
A3K
My guess would be input stage intermodulation distortion. LF442 is a JFET input type and as such these demonstrate better resistance to EMI. A CD player is electrically noisy owing to the preponderance of digital circuits. If you protect the 5532 from RF, (for example with judicious use of an inductor on the input and correct decoupling and PSU arrangements) then in my experience its sound tends to beat JFET types. Unprotected though, its definitely worse.
- Status
- Not open for further replies.