Hi,
Actually, you need to look at MK II version (I think you where looking at the LS-25 schematic at ARCDB.WS). Current ARC stuff seems mostly J-Fets for gain, 6N30PE Tubes as cathode followers on the output. Maybe the wrong way around if you ask me, current CJ products do it the other way tube gain, fet buffer.
The Dallas/Maxim digitally controlled pots ARC uses are 128 step btw, appx. log law. I have tried them and correctly applied (there are one or two crucial tricks, miss them and the results suck) they are among the better digital attenuators. I am familiar with most of them BTW and would not use most.
Sadly Maxim has marked them NRND and Lifetime Buy, so one will have to find something else...
Ciao T
Actually, you need to look at MK II version (I think you where looking at the LS-25 schematic at ARCDB.WS). Current ARC stuff seems mostly J-Fets for gain, 6N30PE Tubes as cathode followers on the output. Maybe the wrong way around if you ask me, current CJ products do it the other way tube gain, fet buffer.
The Dallas/Maxim digitally controlled pots ARC uses are 128 step btw, appx. log law. I have tried them and correctly applied (there are one or two crucial tricks, miss them and the results suck) they are among the better digital attenuators. I am familiar with most of them BTW and would not use most.
Sadly Maxim has marked them NRND and Lifetime Buy, so one will have to find something else...
Ciao T
People who have used it have made very kind comments. As you know, it's not exactly universal in application, so no danger of me ever trying to compete with you...
Solid state amplifiers with GNFB are PROVED as being better than either
tubes amps or SS fedbackless amps.
The ones that contest such scientifically established constatation
always rely on unscientific discourse , going as far as pretending that
their ears are better suited to measure the distorsions of an amplifier
than instrumentation apparatus that are capable of measuring imperfections
several orders of magnitudes below human ear discrimination capabilities.
Hi,
First, better implies a metric to be used by which it is possible to define better/worse. It would be good if, instead of making UNPROVEN and UNSUPPORTED statements, you could state the metrics used.
Second, in order to claim something is "better" one would probably have to prove that it is better for the intended purpose.
Amplifiers in Audio are intended explicitly for the reproduction of Speech and Music, so you need to provide proof that your selected metrics reliably correlate with a greater utility for the stated purpose.
Unless all these are presented you are merely making a statement that is not supported by fact.
It is in fact a direct inversion of the equally unsupported statement "Tubes sound better" or "Zero feedback sounds better".
However, there are very serious academic tests that show at least some support for for the "tubes/ZGNFB better" position that I am aware of, though they are usually not cited by the proponents of "Tubes/ZGNFB".
So, care to provide some evidence?
Maybe then I can then be arsed to provide some for the contrary position, but then again, maybe instead I'll take the advise of my teacher instead...
Calling something "scientifically self-evident" does not make it so. Where is your evidence?
As it seems do those who oppose them, based on your writing.
So it is all about distortion?
You are trying to advance the old chetsnut that "lower distortion correlates with improved sound quality"? Would you care to provide evidence?
Ciao T
First, better implies a metric to be used by which it is possible to define better/worse. It would be good if, instead of making UNPROVEN and UNSUPPORTED statements, you could state the metrics used.
Second, in order to claim something is "better" one would probably have to prove that it is better for the intended purpose.
Amplifiers in Audio are intended explicitly for the reproduction of Speech and Music, so you need to provide proof that your selected metrics reliably correlate with a greater utility for the stated purpose.
Unless all these are presented you are merely making a statement that is not supported by fact.
It is in fact a direct inversion of the equally unsupported statement "Tubes sound better" or "Zero feedback sounds better".
However, there are very serious academic tests that show at least some support for for the "tubes/ZGNFB better" position that I am aware of, though they are usually not cited by the proponents of "Tubes/ZGNFB".
So, care to provide some evidence?
Maybe then I can then be arsed to provide some for the contrary position, but then again, maybe instead I'll take the advise of my teacher instead...
Calling something "scientifically self-evident" does not make it so. Where is your evidence?
As it seems do those who oppose them, based on your writing.
So it is all about distortion?
You are trying to advance the old chetsnut that "lower distortion correlates with improved sound quality"? Would you care to provide evidence?
Ciao T
Torsten re your post on the ARC pre 's. Ok I stand corrected on the technical details of the volume pot. My point was ZJ was anti SS but used a solid state device for what most would consider a critical component.
That said, I have no beef when people claim a sounds better than b. I just don't accept that the difference is because opamps are used, GF or any of the other things that usually come up.
I am going to step out of this specific discussion now - positions are entrenched and I doubt either side will budge an inch.
That said, I have no beef when people claim a sounds better than b. I just don't accept that the difference is because opamps are used, GF or any of the other things that usually come up.
I am going to step out of this specific discussion now - positions are entrenched and I doubt either side will budge an inch.
Hi,
Well, he does not seem to have any problems with SS gain devices either (see ARC LS-25 MKII schematic below).
Maybe he learned something new since that interview (charitable view) or he decided that commercially it was better, sound quality be damned (uncharitable view).
As I have used both open loop J-Fet gain stages and the DS-1666 used above I can attests that neither are reliable causes for bad sound.
Well, I think it may be that statements are not well qualified. For example, GNFB alone is not is a reliable cause for bad sound, however, they way it was commonly used in the 70's and 80's (and still is used most often) it can very well be.
Equally, Op-Amp'd alone do not assure bad sound, however many of those supposedly specifically designed for Audio have pretty poor subjective and objective performance (I find I BY FAR prefer video Op-Amp's for audio) and often the application is questionable, so that again we get many cases where "op-amp = bad sound".
It is tempting and easy to generalise from these frequent but isolated cases...
I can assure you that they will not budge. Each side feels it is reasonable and has good points.
What I do notice, we never get massive and extensive threads trying to copy generic Op-Amp preamps or boring generica Poweramp's, no matter how low the THD.
In fact, I rarely notice massive threads about "objectivist designed" HiFi, suggesting that either these designs not excite people intellectually, sonically or both.
Ciao T
Well, he does not seem to have any problems with SS gain devices either (see ARC LS-25 MKII schematic below).
An externally hosted image should be here but it was not working when we last tested it.
Maybe he learned something new since that interview (charitable view) or he decided that commercially it was better, sound quality be damned (uncharitable view).
As I have used both open loop J-Fet gain stages and the DS-1666 used above I can attests that neither are reliable causes for bad sound.
Well, I think it may be that statements are not well qualified. For example, GNFB alone is not is a reliable cause for bad sound, however, they way it was commonly used in the 70's and 80's (and still is used most often) it can very well be.
Equally, Op-Amp'd alone do not assure bad sound, however many of those supposedly specifically designed for Audio have pretty poor subjective and objective performance (I find I BY FAR prefer video Op-Amp's for audio) and often the application is questionable, so that again we get many cases where "op-amp = bad sound".
It is tempting and easy to generalise from these frequent but isolated cases...
I can assure you that they will not budge. Each side feels it is reasonable and has good points.
What I do notice, we never get massive and extensive threads trying to copy generic Op-Amp preamps or boring generica Poweramp's, no matter how low the THD.
In fact, I rarely notice massive threads about "objectivist designed" HiFi, suggesting that either these designs not excite people intellectually, sonically or both.
Ciao T
Last edited:
We have technically established metric using the usual apparatus,
the very same whose results are never denied when computing a device
caracteristic.
By all possible measurements, GNFB amps have better caracteristics
in respect of accuracy of the amplified signal.
The only response of those contradicting such evidences is to propose
to remove any scientificaly computed number and rely solely on the ears,
since they know that these latters are not accurate enough to the point
that a distorted signal can be confused as "better sounding".
Hi,
Not tried this one, it appears linear law, so not very useful for audio, as the available range is only 48dB best case and the the steps at low volume are very large. This is something the DS1666 already struggles with.
Ciao T
Not tried this one, it appears linear law, so not very useful for audio, as the available range is only 48dB best case and the the steps at low volume are very large. This is something the DS1666 already struggles with.
Ciao T
FWIW, I have had good results with the CS3318 and PGA2320. A cousin of the CS3318 (IIRC CS3308) was used by Jeff Rowland in some of their amps. If you are interested in the issues involved, JR has a white paper on their website explaining how Cirrus designed the on-chip ladder geometry to avoid voltage dependence of the resistances which are typical problems in these ICs.
jan didden
Hi,
This is a purely technical position.
You fail to show any link between your metric (which I understand to be "Distortion" - is that THD btw or something more useful?) and fitness for the purpose.
Speakers are considered high fidelity despite having having well over 0.1% distortion at 1W input and easily several percent at rated power, at low frequencies easily two figure levels of distortion.
So you can argue that according to your measurement a GNFB SS Amp is more "accurate", however if an amplifier with comparably high distortion (say several % THD, 2nd HD dominant at rated power) is combined with a normal speaker in the correct polarity, we observe materially lower distortion for the complete system.
Again, you make claims that lack support.
To question the validity and usefulness of established measurements is not the same as rejecting the use of measurements.
For example, there came up recently a quite astute and well implemented piece of research on distortion audibility, which has it's origin ultimatly in the controversy of better measuring audio devices (lower distortion) which have been criticised as failing to offer improved performance and indeed have been at times cited as offering poorer performance:
Section of Acoustics Institute of Electronic Systems - Aalborg University -
Perception & Thresholds of Nonlinear Distortion using Complex Signals
Ciao T
This is a purely technical position.
You fail to show any link between your metric (which I understand to be "Distortion" - is that THD btw or something more useful?) and fitness for the purpose.
Speakers are considered high fidelity despite having having well over 0.1% distortion at 1W input and easily several percent at rated power, at low frequencies easily two figure levels of distortion.
So you can argue that according to your measurement a GNFB SS Amp is more "accurate", however if an amplifier with comparably high distortion (say several % THD, 2nd HD dominant at rated power) is combined with a normal speaker in the correct polarity, we observe materially lower distortion for the complete system.
Again, you make claims that lack support.
To question the validity and usefulness of established measurements is not the same as rejecting the use of measurements.
For example, there came up recently a quite astute and well implemented piece of research on distortion audibility, which has it's origin ultimatly in the controversy of better measuring audio devices (lower distortion) which have been criticised as failing to offer improved performance and indeed have been at times cited as offering poorer performance:
Section of Acoustics Institute of Electronic Systems - Aalborg University -
Perception & Thresholds of Nonlinear Distortion using Complex Signals
Ciao T
Hi,
Good for you.
I must admit finding these chip's quite bad (I'd take < USD 1 Alpha pot over them any day), compared to my usual solutions and indeed the DS1666.
Maybe I dislike the results the on-board CMOS Op-Amp's produce.
Ciao T
Good for you.
I must admit finding these chip's quite bad (I'd take < USD 1 Alpha pot over them any day), compared to my usual solutions and indeed the DS1666.
Maybe I dislike the results the on-board CMOS Op-Amp's produce.
Ciao T
I am using the PGA2320 you can see some pics and an interim write up on my website. Still have to build the case but that will take some months. So far, I am impressed with the sound and pleased with the result.
If you take a look on the web at the new Esoteric preamplifier, you will see that it uses a PGA2320 as the main volume control and gain element. I have not configured my design like theirs though. Last time looked, there were a few reviews, and nothing bad about the product- all good in fact. Selling price is around $10k IIRC.
Last edited:
Great, a claim that an engineer can make some firm deductions from
reducing distortion is unimportant except when done by composing canceling distortions in different system components
no normal speakers use multiple drivers for different frequency ranges, or IMD products are inaudible – choose one
since IMD products from the amplifier won't be canceled even if multiple driver's nonlinearities individually match/exactly complement the amplifier – the crossover separates many of the frequencies to the individual drivers so they never mix to produce all of the IMD complementary to the amplifier's
no normal speakers use Linkwitz-Riley LF4 crossover networks, or somehow normal speakers different drivers just happen to have the correct alternating sign or compressive/expansive distortion terms going from woofer to mid to tweeter
I just keep on learning new stuff here
May I ask why they cant be beat.
Regards
Arthur[/QUOTE]
Well, there might something that beats it, but not my QuadII, Mac, Luxman class A, some other tube gear I own, or anything I have listened to that I can remember.
This is very disappointing to me. Class D amplifiers are nothing but institutionalized cross over distortion machines, albeit at very high frequencies. On my scope, the class D amps I have checked show a wide band of high frequency noise, so much so that I have not been able to do any THD measurements on them; they upset my measuring equipment. And they come in disappointingly small form factors. How can all this be good?
But it is, and I have not been able to find a satisfactory reason why this should be so.
One reason might have to do with the fact that class D sinks reactive energy coming from speakers in a very different way from conventional amps. I have seen others on this forum floating this theory. Another may have to do with the fact that remnants of the switching frequency remain in the signal fed to the speakers, even after filtering. What this might do is keep the speakers in ever so slight motion, just enough to overcome the hysteresis present in every driver.
Hi,
My point was merely to illustrate how silly all this "low THD in Amplifiers" monkey jabber really is.
BTW, my concrete example involved a single wideband driver driven by an SE Tube Amplifier, just BTB, no crossover between Amp and Speaker, active sub and supertweeters, wideband driver covering around 100Hz-10KHz, dipole arrangement.
Ciao T
My point was merely to illustrate how silly all this "low THD in Amplifiers" monkey jabber really is.
BTW, my concrete example involved a single wideband driver driven by an SE Tube Amplifier, just BTB, no crossover between Amp and Speaker, active sub and supertweeters, wideband driver covering around 100Hz-10KHz, dipole arrangement.
Ciao T
- Status
- Not open for further replies.