Best opamp for I/V conversion? (DAC)

Jump to Latest
#321
Ken Newton said:
Firstly, please spare me your silly condecensions.
Secondly, hyperbole much?

You sound like the kind of person who let's spectrum analyzes tell them to what degree their ears should be enjoying the music being reproduced some given component. Better yet, just visit eBay and order whatever are the cheapest components claiming the lowest static THD, yes?

Regarding DAC chip design, the resistor class of DACs (R-2R, ladder, etc.) don't use signal feedback. They typically consist of a quiet precision buffered voltage reference feeding resistors, or switch precision current sources through those resistors. You will see feedback employed to maintain D.C. levels, or to minimize the output impedance of the reference buffer. If, however, they employed signal feedback at the quantizer block you'd see much better THD figures listed for the resistor class of DACs, which greatly depend on the precision of their resistor networks (or on averaging the lack of precision to noise via DEM) for low distortion (particularly obvious problem at low signal levels), just as you see low THD figures for the feedback based sigma-delta types. The sigma-delta class of DACs inherently use discrete-time digital feedback, which is not the thing thing as the actual quantizer block which can and often is a small multibit resistor class converter.
Click to expand...

Funny, I guess i better just buy the farm :wchair: call the enjoyment of 2nd harmonic rich (being kind) IV conversion what you like, but dont call it more transparent.... its by definition LESS transparent. nothing wrong with enjoying it, people including myself for a couple of months; have enjoyed it for many years before and many years to come, but dont try and cover up those sort of deficiencies in transparency and house it all under the banner of 'as yet unknown'. I 100% agree with the ears being the last arbiter as we only have assisted and collaborative subjective experience, but it should be a final tuning process and our ears are afterall attached to our heads

besides you can get -115 THD + N with a single stage discrete class A mosfet IV, doesnt have to be an opamp.

i'm glad someone else mentioned buffering the transformer.

but i'll leave you guys to compare the length of your academic record 🙄
 
Last edited:
#322
qusp said:
Funny, I guess i better just buy the farm :wchair: call the enjoyment of 2nd harmonic rich (being kind) IV conversion what you like, but dont call it more transparent.... its by definition LESS transparent. nothing wrong with enjoying it, people including myself for a couple of months; have enjoyed it for many years before and many years to come, but dont try and cover up those sort of deficiencies in transparency and house it all under the banner of 'as yet unknown'. I 100% agree with the ears being the last arbiter as we only have assisted and collaborative subjective experience, but it should be a final tuning process and our ears are afterall attached to our heads

besides you can get -115 THD + N with a single stage discrete class A mosfet IV, doesnt have to be an opamp.

i'm glad someone else mentioned buffering the transformer.

but i'll leave you guys to compare the length of your academic record 🙄
Click to expand...

You've yet to step outside the intellectual box you've placed your mind (and ears) in. Just re-read that first, er, sentence in which you wrote: "Funny, I guess i better just buy the farm :wchair: call the enjoyment of 2nd harmonic rich (being kind) IV conversion what you like, but dont call it more transparent.... its by definition LESS transparent."

Your position that an I/V converter featuring mostly 2nd harmonic distortion will not sound as transparent as one without (something I never alluded to, by the way) is an intellectual position, not a perceptual one. We do not appreciate music through our intellectual understanding of engineering theory, we appreciate it through being human.
 
Last edited:
#323
Ken Newton said:
You just said it yourself, your belief is that lower THD will sound better than higher THD.
Click to expand...
Where did I say that? "Measure better" is not equal with "lower THD". 0.1% sounds better that 10%, but 0.00001% doesn't sound necesarily different that 0.1%. It's all in th emind that READS those measurements. BTW, measurements are not equal either with the ads in 80's newspapers that everyone keep quoting as argument. Yes they where lying, so? How that makes measurements responsable? How do you think they desing the DAC's if not using measurements and formulas? Or they are using alienes, magic, wodoo?

I DID listen to the stuff that is posted around here like the best thing since invention of the wheel. That is: resistor I/V, NOS, tube output (just one branch from a DAC that has differentials), no filtering.
It sounds anywhere from bad to horrid. Sounds that should not be there apear due to aliasing process. Distortion and noise are annoying. And yes, that stuff measures horrid too.
Only somebody that desperatelly wants to belive that "different" is always "better" can argue the opposite.
 
Last edited:
#324
Sonic,

If you find NOS, or passive resistor i/v, or zero feedback, or whatever to sound unpleasant to you, fine. I have no argument with this. That has been my point. The problems begin when you then seek to justify what are your own personal preferences and perceptions with engineering theory. Which, obviously, has the intended objective of discrediting the preferences and perceptions of any that differ from your own.

You are very quick with the strawmen arguments, such as the nonsensical suggestion that there is the belief (by me or whomever) that any difference is always better. Apart from the illogic of such a statement, I truly wish that you could see the arrogance which underlays that, and many of your arguments. Such, unfortunately, exemplifies much that is wrong with the supposed scientific mind today.

Too many of us in the scientific/engineering community, IMHO, seem to have forgotten, or never learned, that when theory conflicts with observed reality it is the theory which is either incorrect or incomplete (or simply mis-applied), not the observed reality.
 
Last edited:
#326
tunrning self reported sense perceptions into data reflecting reality is another part of science and technology - why is it that many audiophile "tweakers" seem ignorant of and/or hostile towards any listening controls, bilnding protocols, any suggestions that human perception is variable, less than "infinetly resolving" - essentially claiming that their hearing is a comic book style super power
 
#327
cliffforrest said:
That may be true if the "observed reality" is established scientifically.

"I heard (insert your hyperbole here)" does not and never will, cut it.

You are welcome to trust your ears, but don't expect me or anyone else to so do.
Click to expand...

Eventually, yes, one expects observed reality and scientific explanation to coincide - completely. However, just because they may not currently coincide completely doesn't mean that present scientific explanation therefore has invalidated some observed reality. That is the point of maintaining an critical yet open mind. Many scientific theories have undergone, and continue to undergo, revision as our understanding of observed phenomena increases. But, when there is disagreement between observed reality and existing theory, it is the existing theory which must be questioned before the observed reality. To do otherwise would seem irrational.

I would never ask you to trust my ears. I would ask you to trust your own. Related to this is the question of whether current instrumented measurements dictate a particular perceived human reality, the enjoyment of electronically reproduced music, or rather serve to help explore an explanation to such enjoyment.
 
Last edited:
#328
jcx said:
tunrning self reported sense perceptions into data reflecting reality is another part of science and technology - why is it that many audiophile "tweakers" seem ignorant of and/or hostile towards any listening controls, bilnding protocols, any suggestions that human perception is variable, less than "infinetly resolving" - essentially claiming that their hearing is a comic book style super power
Click to expand...

That is a good question. I suspect that it might have something to do with their suspicion that so-called objectivists are too eager to accuse them of illusion, delusion, lying, insanity, and other forms of public ridicule. Heaven save us from all forms of closed-minded attitudes, perceptual and intellectual.
 
#330
Its a matter of taste - I reckon myself that you need wide bandwidth and freedom from intermodulation in that critical first stage. So standard LTP (undegenerated) input stages are ruled out. That leaves, inter alia, CFB topologies - but some people don't like the sound of them. 😕
 
#332
tvicol said:
LME49713
Regards,
Tibi
Click to expand...
thank you - good advice.
http://www.diyaudio.com/forums/digital-line-level/117238-ess-sabre-reference-dac-8-channel-132.html (post #1315)
LME49713, could be a killer opamp
http://audio.com.pl/testy/cary-audio-design-xciter-dac:25
http://audio.com.pl/pdf/testy/przetworniki_d_a/Cary Audio Design XCITER DAC.pdf
http://www.elforum.ro/viewtopic.php?f=35&t=59477&start=15

Who can say the lowest possible supply voltage without loss in audible quality for +/- 2Vss output swing (the mentioned value of +/-5V from datasheet is probably too low)?

Where are to find the internal simplified schematic diagram?
 
Last edited:
#334
dirkwright said:
Keeping RFI out of that chip is going to be a challenge for anyone.
Click to expand...

I built a headphone amp using two of these that was powered by a DC/DC converter. No RFI problem that I could tell; excellent sound.
 

Attachments

  • 49713.jpg
    49713.jpg
    75.1 KB · Views: 1,289
#336
tvicol said:
LME49713

Regards,
Tibi
Click to expand...

1. inverting input of the current feedback OP is a poor choice for virtual ground summing

and

2. transimpedance current feedback OP cannot have a cap across the feedback resistor


Datasheet specifies the Rf value of 1.2Kohm to be optimal; this value could give less than desired voltage output level with well-regarded current-out DAC's that have only around 1mA output…. which means a second gain stage…. or…..the choice of a sub-optimal DAC (with higher Iout)… or even worse – the Vout DAC

There are OP's that can give outstanding results if used as I/V, due to very high slew rates, extremely low input bias current and very low offset voltage + they are voltage feedback OP's. The only thing to pay attention to is output oscillations, but once done right – this combination can really show what preceding current-out DAC is capable of. If the oscillations are not controlled properly the end result could be quite bad.

49713 will most definitely sound very nice (at least the oscillations would be easier to control compared to a fast voltage feedback OP) - however, the PCB trace cutting (cap should be placed at the input of the current feedback OP, in combination with the resistor for the desired cut-off frequenc, and damping) and a bit of adjustments are needed to allow this chip to be used in place of voltage feedback OP. Not many DIY-ers are ready to do this... many expect drop-in replacement should just work irrelevant of the type of the OP at hand.


Boky
 
#337
is this series RC at the input of the CFB chip considered part of the dac output, or the opa input? ie. i take it you must know the CFB chips inputZ to enable proper choice for the size/value of the component? or does the resulting LPF rely mainly on the dac's output impedance? i expect its a combination of the 2, but which has the more profound effect? i'm on mac so cant run spice unfortunately
 
#338
Bump

Curious if there has been any more experience with the LME series opamps referenced since these last posts. My interest is in a comparison of the performance, and subjective stuff as well, to some discrete solutions using current conveyors, and current conveyors with amps wrapped around them to reduce the input impedance.

If this is being covered in other more recent threads, I'll be happy to have them pointed out. It's a big site.

Brad
 
#339
I would try fet input or AD's "highly linear" front end op amps - seem to be poorly advertized but input diff lineariztion should be great for DAC I/V objective performance - high GBW doesn't hurt either

I seem to recall the some of the LME being faulted by some for less than stellar input EMI/RFI rectification/sensitivity rejection
 
Last edited:
#340
Thanks jcx. Yes I could imagine the bipolar inputs of the LME series to have some issues with well-out-of-band energy.

The results I'm getting, in sim at this point, with cascoded JFET/DMOS a la EUVL's approach, and with similarly cascoded amplifiers for feedback around the common-gate input, look very promising. As a comparison I looked at the old dog 5534 just as a point of reference for what it manages with an app note recommendation in terms of input impedance with stiffish R-C feedback and it definitely underperforms compared to the circuits under investigation --- by a whole bunch.
 
You must log in or register to reply here.
Top Bottom