Worth noting John Caldwell's comments here, as well: http://www.diyaudio.com/forums/head...eadphone-amp-comments-advice-appreciated.html (Posts 61 & 65).
What? Are you saying the plus reference input does not matter and CMRR therefore does not matter? If this were true opamps would always be used in inverting mode. Out of curiosity, where did that notion come from?
CMRR is a frequency dependent measurement and not a "number." Please refer to a graph of CMRR versus frequency to know the difference between a very good op amp and that piece of antique junk 5532.
No - he meant that noninverting gain circuits were not that common. Which I cannot agree with, in fact the only area dominated by inverting circuits that I can think of would be mixers, oh, and very low-voltage IC designs (think +Vs = 5 V max, where common-mode voltage swing can become very small). Elsewhere their noise penalty often tends to be undesirable.
It is true that inverting mode disables several important distortion mechanisms though. If you think about it, inverting operation has both inputs at a fixed potential. Hence, common-mode voltage swing = zero. +input impedance is also being kept very low by feedback (virtual ground). So it's bye-bye to input impedance distortion and all the distortion hinging on common-mode voltage.
BTW, the design of the old TL07x suddenly makes all the sense in the world if you primarily keep inverting use in mind. You wouldn't need a massive output stage as feedback resistor values are limited by desired input impedance anyway, but highish slew rate and good transfer linearity would be desirable as signal levels would have to be reasonably high for a good amount of dynamic range.
Of course that it s frequency dependent, and the 5532 has better numbers to start with, so the evolution in respect of frequency can only be better as well..
Indeed using deragotary terms to qualify an op amp is of no scientifical value and is just aknowlegment of lack of aguments, i guess that you did throw this CMRR blunder because you didnt even bother checking the numbers..
FTR the LF353, or NJM2968 for instance, share the same topology as the 741, wich was famed in discrete design under the blameless moniker by Douglas Self, linearity wise the NE5532 is one step above thanks to its three stages topology.
It is not a notion. It is called engineering, about which you appear to know little.
Oh thanks cliff. I see you confuse theory with practice as many do. Grounds are never every stable and always move with load signal. Many like to look at their computer models, especially using idealized parts, and then believe they know about real circuits in the real world. No so. CMRR always comes into play with any op amp though, sometimes it is not an issue however, many times it is. Inverting or not it still can easily be an issue. Maybe you could try some real world testing?
You know something, if you are in the U.S. it is almost impossible you have not be directly exposed to my design work. I know a lot about design with millions of pieces in circulation with other designs widely used throughout the industry of audio. So widely you can almost not escape being exposed therefore, all you have done is insult yourself and your lack of grasp of reality in favor of some notion of theory and computer models. That kind of engineer is 10¢ a dozen. Am trying to educate those who wish to learn and understand here however, there are many like you who already know everything, not.
Someone who tries to throw their weight around would not be routinely inclined to write:
Funny enough, it's not too different from this very good op amp (with very similar GBW) http://www.ti.com/lit/ds/symlink/opa1642.pdf
Funny enough, it's not too different from this very good op amp (with very similar GBW) http://www.ti.com/lit/ds/symlink/opa1642.pdf
I wonder what the use of CMRR at 6 MHz is? An interesting choice of frequency for audio and op amps with appropriate gbw.
Wouldn't a much more sensible value be 50/60 and 100/120Hz and 1kHz and say 20kHz?
Well - I've only been in this gig for a few decades now, and that's where I'd start. Along with a filter to ensure the amplifier never saw RF.
Hey which op amp works well against those specs? (Flip back a page, the graph is there)
Wouldn't a much more sensible value be 50/60 and 100/120Hz and 1kHz and say 20kHz?
Well - I've only been in this gig for a few decades now, and that's where I'd start. Along with a filter to ensure the amplifier never saw RF.
Hey which op amp works well against those specs? (Flip back a page, the graph is there)
I don't feel immortal as you seem to arbitrarily qualify people... And don't feel inclined to test it out.
I do however feel inclined to call someone out when I see a load of bollocks presented as fact.
I am not a 5532 nutter. Why would I be? But I also won't swallow the load of bollocks that presents an lf353 as an equivelant.
Choosing a device to support a crap design is a poor reason for going into bat for it!!!
That said, my goto op amp of choice right now is the lm4562.
I wonder how long this post lasts....
I do however feel inclined to call someone out when I see a load of bollocks presented as fact.
I am not a 5532 nutter. Why would I be? But I also won't swallow the load of bollocks that presents an lf353 as an equivelant.
Choosing a device to support a crap design is a poor reason for going into bat for it!!!
That said, my goto op amp of choice right now is the lm4562.
I wonder how long this post lasts....
Last edited:
I'll throw my 2 cents worth into the frying pan.......... I have not used opamps in several decades.... but recently I needed them to upgrade some test equipment. Last year I bought a used Mastering recorder for some testing. I listened to its recordings and it sounded pretty poor compared to descretes I had used...... It was lifeless and dull etc. It could be the coupling caps (electro's) or other..... I thought I might find out why it was so poor sounding and opened it up and looked inside. All the analog used 5532/4. I thought about replacing them all and other parts but decided not to do so for the time involved and I would use it as a transport and take the digital output to a great DAC (BenchMark's DAC2). I wont go so far as to say it was all in the opamp.... but they are now suspect IMO.
THx-RNMarsh
THx-RNMarsh
RNMarsh
I am a big fan of discrete. Where sheer volume allows that is "how I roll". I also recognise that in a lot of real life as I meet it, even SMD discrete just doesn't fit.
I suffer from being an engineer (and I guess scientist if you dig too far) and struggle to see how a working, decent, op amp will kill an analogs signal.
That said, I have encountered serious errors in design in seriously high end gear before, but to say that a "decent" op amp is singularly at fault in and of itself breaks me up.
Cas, pots, switches etc are all varying levels of evil. Oh, and plugs, sockets and connectors....... I'll add to that list the tweakers favourite IC sockets.
I am a big fan of discrete. Where sheer volume allows that is "how I roll". I also recognise that in a lot of real life as I meet it, even SMD discrete just doesn't fit.
I suffer from being an engineer (and I guess scientist if you dig too far) and struggle to see how a working, decent, op amp will kill an analogs signal.
That said, I have encountered serious errors in design in seriously high end gear before, but to say that a "decent" op amp is singularly at fault in and of itself breaks me up.
Cas, pots, switches etc are all varying levels of evil. Oh, and plugs, sockets and connectors....... I'll add to that list the tweakers favourite IC sockets.
yes... any and all could be contributors. I dont have enough experience listening to opamps nor the 5534 in particular to say it is an opamp sound. I am curious now to hear other gear using this opamp.
THx-RNMarsh
With all the radio and TV broadcast plus noise one must always deal with frequencies in the megaHertz region with filters and of course the feedback of the op amp which will typically have at least gHz transistors used in the build. Pretending "nothing above 20kHz" matters is just plain stupid. It is the bandwidth of the device being used to be the lowest frequency of interest. Out of band signals also cause numerous distortions when applied to a circuit. Who has not heard radio interference in an audio circuit? How did that get in there? Poor design and wide bandwidth parts.
I used the op-amp the lack of whit compared to the LF353, that OPA thing, saying great HF CMRR is not that rare in op amps. Not true unless looking at differential video amps. The NE5532 modern datasheets do not show a CMRR graph for an NE5532 because that would be one pitiful graph. If I come across this graph will post it.
When truth appears the ignorant will line up in a row against that truth. Try and learn something here and not try and be right with ignorance.
I used the op-amp the lack of whit compared to the LF353, that OPA thing, saying great HF CMRR is not that rare in op amps. Not true unless looking at differential video amps. The NE5532 modern datasheets do not show a CMRR graph for an NE5532 because that would be one pitiful graph. If I come across this graph will post it.
When truth appears the ignorant will line up in a row against that truth. Try and learn something here and not try and be right with ignorance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.