I still fail to see why we're hung up on out of band CMRR--isn't that why we build non-antenna pcb's and carefully ensure that the inputs and outputs are RF filtered? I care much more about CMRR within the audio band, as there are prudent things I can do to mitigate RF. That's not to say I don't enjoy an opamp with greater GBW, especially in certain, critical, locations. (DAC I/V comes immediately to mind)
Likewise, governing relationships like this:
Tell us that VFA amplifiers with similar GBW (and internal dominant pole compensation) are *probably* going to have pretty dang similar CMRR and PSRR's, especially at high frequencies.
I'm not necessarily defending the NE5532, just someone who doesn't pay much attention to datasheets and does "design by lore" rather than knowing what he/she needs the circuit to do and builds accordingly. Yes, I do wish that we had far, far better data sheets for said NE5532, but we take what we can. Samuel Groner's excellent work highlights some of the less-than-ideal things about the NE5532 (CM distortion probably dominates at low gains).
Likewise, governing relationships like this:
Tell us that VFA amplifiers with similar GBW (and internal dominant pole compensation) are *probably* going to have pretty dang similar CMRR and PSRR's, especially at high frequencies.
I'm not necessarily defending the NE5532, just someone who doesn't pay much attention to datasheets and does "design by lore" rather than knowing what he/she needs the circuit to do and builds accordingly. Yes, I do wish that we had far, far better data sheets for said NE5532, but we take what we can. Samuel Groner's excellent work highlights some of the less-than-ideal things about the NE5532 (CM distortion probably dominates at low gains).
CMRR is a linear small signal measurement RFI/EMI detection/nonlinear conversion is not strictly related - likely some correlation - but if you need RFI detection resistance you don't rely on manu datasheet numbers for CMRR
and in general jfet input are better on RFI detection than non-degenerated bjt op amps at similar input stage bias current
and in general jfet input are better on RFI detection than non-degenerated bjt op amps at similar input stage bias current
I do not really feel that here on DIYaudio. When I used to talk to Bob Pease, Walt Jung, Oscar Heil, Paul Klipsch, and many others I was never treated as poorly as here. Actually always with great respect with only John Kurl being as rude as almost everyone here is. There are good people here but, there are a boat load of wannabes who rely on some half baked or fully baked computer model and not on actual laboratory data. John Kurl once said, "I wish my stuff had as low of distortion as yours." About the only nice thing I have every heard from him. I try and nicely explain things and most of what comes back is a bunch of data from models having little to nothing to do with reality. This is why my designs and those like Bob Pease become almost universal and these wannabes strike out against the test of reality. I have huge experience and decades in instrumentation and pro audio but, no matter. Some wannabe goes off on the values of seriously inferior parts like this 5532 thread and then goes on and on about how somebody once used it. Well I once used it also but, that was several decades ago because there are far far better choices now no mater their worship of the should of been deleted part. Theoretical knowledge and imaginary computer based models versus real world test in real circuits. Mine is almost entirely the latter. That is the long and short of it. Use a junk part if you want but, do not try and convince me nonsense based on theoretical knowledge and computer models is somehow related to facts.
FYI- CMRR is exactly the opamps ability to have a useful and functional feedback loop at the frequency of interest. Often this includes the noise and interference feedback components in high performance systems. Plus and minus inputs are always involved in the feedback without exception.
Man, sumaudioguy, you're running the gamut of logical fallacies. Appeals to authority, direct insults, n=1 sighted listening comparison, unsubstantiated claims...
Why, again, are we supposed to follow your assertions? What "meat" are you actually providing? And what magical properties does your favorite NJM2068 provide? Nothing at *major* all jumps out. (Which is to say, it's a perfectly decent and inexpensive opamp for many applications and a few dB lower noise than the 5532, which is certainly commendable)
I get it, you don't like the NE5532 for some reason and you'll go out of your way to rag on it. Yes, there are plenty of chips that are better for this or that (and several modern chips for this AND that) but, all in all, it's really does a lot of things well. And under most common uses, these advancements would *probably* have a hard time being heard. I'm feeling particularly allergic to hyperbole.
If you're going to ask for the DiyAudio community for decorum, best to start from the source.
Why, again, are we supposed to follow your assertions? What "meat" are you actually providing? And what magical properties does your favorite NJM2068 provide? Nothing at *major* all jumps out. (Which is to say, it's a perfectly decent and inexpensive opamp for many applications and a few dB lower noise than the 5532, which is certainly commendable)
I get it, you don't like the NE5532 for some reason and you'll go out of your way to rag on it. Yes, there are plenty of chips that are better for this or that (and several modern chips for this AND that) but, all in all, it's really does a lot of things well. And under most common uses, these advancements would *probably* have a hard time being heard. I'm feeling particularly allergic to hyperbole.
If you're going to ask for the DiyAudio community for decorum, best to start from the source.
Last edited:
I would have thought that attention would be payed by the main suspect to the link i gave above, for more insight here two datasheets.
yes, these are exactly the same circuit, so now we can add the TL07x serie to the list of op amp allegedly better than the 5532, in the waiting of the 741 and the 709..
yes, these are exactly the same circuit, so now we can add the TL07x serie to the list of op amp allegedly better than the 5532, in the waiting of the 741 and the 709..
I have always said everyone should do what they believe is right. I made several test beds (different configurations) with sockets which put 10 op amps in a row and tried a bunch of different op amps. This is where the sound differences became less than subtle. Has anyone here but that one guy done any testing remotely similar? Not that anyone has stated. Seems most are basing their views on either some opamp buried in a circuit or what someone else says. How about a little DIY and build some test bed and try these different amps out in your system? Is that to much trouble to make a decision? Probably for most.
As for TI parts, I know a boat load of engineers who will not use them including me. Does not matter at all to me if you want to use them. We will not and will always find another way to avoid any TI part. Goes for Illinois Capacitor also. Both junk. People purchase Mini Cooper cars in spite of the fact that car has made the least reliable highest maintenance car sold in America for years. Just because many do does not make it a good idea.
No I am not a nice guy. I deal in facts and what works in a laboratory and installation setting. This makes a lot who worship trends, popularity, and gurus angry and resentful. I cannot change that.
Derfnofred- No appeal to authority. Sighted listening test? Where do those come from anyway? I agree with many systems the differences would often be to small to note with a single op amp. On the other hand, in a great system they are not to small to note. There is the assumption, a great system. Not that many of those around with none using 5532s.
Wahab- the TL-072 and the LF-353 used to be by different manufacturers. This changed when TI purchase (absorbed) National Semi. Now that LF 353 is made by TI I will not use them. What is you point? Did you not know these were different manufacturers for a very long time or are you just making waves? FYI- I have heard if you take a TL-072 and run it on ±22 volts for two days, the ones that still work are actually pretty good. Everyone I know that used the TI part graded them this way. Some 80% or more did not make the grade is my understanding.
cliffforrest- I can tell you drive a Ford or VW. Nuff said.
As for TI parts, I know a boat load of engineers who will not use them including me. Does not matter at all to me if you want to use them. We will not and will always find another way to avoid any TI part. Goes for Illinois Capacitor also. Both junk. People purchase Mini Cooper cars in spite of the fact that car has made the least reliable highest maintenance car sold in America for years. Just because many do does not make it a good idea.
No I am not a nice guy. I deal in facts and what works in a laboratory and installation setting. This makes a lot who worship trends, popularity, and gurus angry and resentful. I cannot change that.
Derfnofred- No appeal to authority. Sighted listening test? Where do those come from anyway? I agree with many systems the differences would often be to small to note with a single op amp. On the other hand, in a great system they are not to small to note. There is the assumption, a great system. Not that many of those around with none using 5532s.
Wahab- the TL-072 and the LF-353 used to be by different manufacturers. This changed when TI purchase (absorbed) National Semi. Now that LF 353 is made by TI I will not use them. What is you point? Did you not know these were different manufacturers for a very long time or are you just making waves? FYI- I have heard if you take a TL-072 and run it on ±22 volts for two days, the ones that still work are actually pretty good. Everyone I know that used the TI part graded them this way. Some 80% or more did not make the grade is my understanding.
cliffforrest- I can tell you drive a Ford or VW. Nuff said.
There was Analog Systems back in the 1980s that took TL-072 and the NE5534 (neither from a big vendor), graded then, rejected 90% of them. And, at least the NE5534 variant did sound good. It was sold under the part number - MA-362. And went for around $11/ea. A lot of them went into mixers and microphone preamps. The MA-372 replaced the TL-072 and the MA-332 replaced the NE5532. Same parts just binned from the top 10%.
That s right that the NS LFs were generaly prefered by the audio crowd over the TLs but those latter were much more used in commercial products, guess that cost was the reason.
Dont know if TI has inferior process and if it showed in their products, the 5532 is quite sensitive to transistors caracteristics in that linearity can vary by 10-15dB from a crappy process to a good quality one, but still, the circuit is so linear that even the worse samples from the worse foundry should be within 4 ppm THD within the audio band.
In the late 70s Etienne Lemery from the deceased french review Le Haut Parleur did test it with a bunch of other op amp as comparisons, his Signetics samples yielded less than 200ppm THD at 20KHz, 60dB gain and +20dBm power output (7.75V RMS/600R) , they could theoricaly do 0.2ppm at 0dB gain and at theses settings.
NE5532 or better ne5534an have decent voltage noise characteristics and very good current noise characteristics for bipolars so with low input signals, as in phono, they offer superior noise characteristics to lm4562, Noise Factor of ne5534an is almost 5dB better that that of lm4562 with standard 47k input load resistance used with MM cartridges.
However, with balanced inputs that resistance can be lowered to 1k or below and then noise advantage of ne5534an disappears.
cheers,
cheers,
However, with balanced inputs that resistance can be lowered to 1k or below and then noise advantage of ne5534an disappears.
cheers,
cheers,
Must admit I like the sound using LF353 in my active crossover. LM4562 seems a bit clinical and 5532 sounds a bit "dirty".
I think I'll try to get hold of some NJM2068Ds and give them a whirl.
Ah! Is this where their lowish differential mode input impedance comes in? I was always in what kind of circuit that would matter.
The 5532 does not work well in high impedance circuits, 40K tops (IMO).
http://www.marchandelec.com/ftp/xm1man.pdf
Most commercial crossovers are similar to the Marchand, look and see how many opamps the signal has to go though!
http://www.marchandelec.com/ftp/xm1man.pdf
Most commercial crossovers are similar to the Marchand, look and see how many opamps the signal has to go though!
All very valid comments!
My XO is a Salen and Key type and since the XO frequency is low (120Hz) the resistor values are indeed highish, therefore loading by the input resistance of the 5532 (30K min) could be a problem. The minor concerns about the sound quality of the 553x parts happens well above the XO frequency, although it's possible changes in response at LF can affect the way we perceive higher frequencies.
If I can be bothered, I will make some measurements of the actual performance of the filters with/out 5532s. I doubt that minor response deviations will be easy to hear at these frequencies due to the large effect the room has below 300Hz. FWIW I also formed my opinion of 553x over may years of playing around with preamp, RIAA, I/V etc. ciruits.
It is also worth keeping this in context:- sound differences between opamps are quite small and to some extent application dependent. I am also prone to changing my opinion like the weather can change here in the UK
.
djk - I will try some 2604s. I have not been keen on these in the past. The fancy version, the 627 sounds even worse IMO. Anyway, I am happy to give them another chance and it looks like the weather might be brightening up
My XO is a Salen and Key type and since the XO frequency is low (120Hz) the resistor values are indeed highish, therefore loading by the input resistance of the 5532 (30K min) could be a problem. The minor concerns about the sound quality of the 553x parts happens well above the XO frequency, although it's possible changes in response at LF can affect the way we perceive higher frequencies.
If I can be bothered, I will make some measurements of the actual performance of the filters with/out 5532s. I doubt that minor response deviations will be easy to hear at these frequencies due to the large effect the room has below 300Hz. FWIW I also formed my opinion of 553x over may years of playing around with preamp, RIAA, I/V etc. ciruits.
It is also worth keeping this in context:- sound differences between opamps are quite small and to some extent application dependent. I am also prone to changing my opinion like the weather can change here in the UK
.djk - I will try some 2604s. I have not been keen on these in the past. The fancy version, the 627 sounds even worse IMO. Anyway, I am happy to give them another chance and it looks like the weather might be brightening up
Last edited:
It's not quite as simple as all that.
The 47k is the -load- resistance and this a small part of the equation.
If you look here Magnetic Phono Pickup Cartridges
there is an approximation of the self impedance of a MM cartridge.
Essentially we are looking at approx 220mH in series with 1k2.
As such the total noise will be opamps voltage noise + current noise reacting
with cartridges self impedance which will be rising with frequency.
Using a low noise jfet IP opamp will probably net the best results.
Then there is the whole subjective side of the equation
T.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.