|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
61.239.75.90
In Reply to: RE: It would be a deal-breaker for me indeed.....to this reviewers credibility. posted by dave789 on January 02, 2015 at 12:04:11
Hi,
> I never saw a product in Stereophile with higher power supply
> related jitter than the V90 reviewed by KenRockwell.
That is because they use an AP2, which goes a long way towards eliminating ground loops from the test gear (care is still needed).
And lo and behold, when Stereophile measured the V90, there was no such "power supply related jitter" to be measured...
I wonder why?
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Follow Ups:
Stereophile review sample: RRY0135 (JUNE 2013)
KenRockwell review sample: RRY0038 (JUNE 2013)
There can be product by product variation within the same model, V90.When V90 was replaced by Benchmark, there was no such jitter, according to KenRockwell. I do not know why the person who connected things should be blamed.
Connecting a DAC to an analyzer or an amp is not aerospace engineering. For an ordinary audiophile to use V90 jitter free, what do yo suggest?
With a power supply without ground prong, there is no ground loop. There can be of course ground noise. Anyway, if that V90 needs aerospace engineering level care to avoid "power supply related jitter," it just indicates poor quality control or poor design by Musical Fidelity.
Edits: 01/02/15
Hi,
> There can be product by product variation within the same model,
> V90.
The serial numbers and production dates are very close.
> I do not know why the person who connected things should be blamed.
I am not blaming the person for connecting things. But for publishing test results with insufficient fact checking. I for one would have tried to eliminate these mains related components from my measurements much harder and likely would have succeeded.
I guess that is the difference between a consummate professional who desires to understand what really goes on and an Amateur bend on panning products.
> Connecting a DAC to an analyzer or an amp is not aerospace
> engineering. For an ordinary audiophile to use V90 jitter free,
> what do yo suggest?
Use as is. As I remarked,all I see are mains related side-bands which are NOT jitter, just hum. And at a very low level. If the hum bothers you, find the source.
> With a power supply without ground prong, there is no ground loop.
For a power supply without an earth connection there is still a ground somewhere. In the kind of plug-top SMPS as used in this review it is via a low value capacitor, called a "Y" capacitor. Values of 10nF are common. Similar capacitors are also found inside mains powered gear without an earth connection, simply because you HAVE TO HAVE A GROUND somewhere, even if this one is not obvious.
This capacitor is usually connected to one blade of the Power plug only. If this side is connected to the neutral line of the mains (which should be the same as earth), all is well. If it is connected to the life connector (a 50% chance) then 115V/60Hz will applied across this capacitor in the USA and 230V/50Hz in Europe.
As a result 50/60Hz currents of up to around 1mA will flow through this capacitor, into the ground of the DAC and from there into the ULP analyser used to test. Now if there is any inductance and resistance in the cabling used and in the connectors (rest assured there is plenty), you can easily get fault voltages of a few 100 microvolt, more then enough to bollox up the test results.
> There can be of course ground noise. Anyway, if that V90 needs
> aerospace engineering level care to avoid "power supply related
> jitter," it just indicates poor quality control or poor design by
> Musical Fidelity.
It indicates neither. It indicates a lack of competence in the person performing and interpreting the the tests. There is a reason why only licensed and competent professionals are allowed to practice law and medicine, sadly similar restrictions do not apply to audio measurements.
You may argue that the published test indicates the real world hum noise performance of the DAC in a careless set-up with the wrong AC Plug polarity (if my guess is any good), but what he measured is still not jitter, just hum and normally these levels of hum should be a warning sign to the performing the tests that something in the test set-up is amiss.
Anyway, I see no point in defending competitors products, Imerely wanted to point out that just because "A measured K and claims it is X" does not mean that K is real and really represents X.
And as said, it happens often enough that K is not "real" and that it does not represent X.
Only last week I had one of my junior Engineers chasing instrument ghosts for two days trying to find noise that did not originate in the device where he supposedly measured the noise. And this feller did press the school bench for 3 years to get his EE degree...
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
Mains related side-bans are jitter.
It is different from noise at 60Hz, 120Hz, 180HZ, 240Hz, . . . . . .
Hi,
Jitter is clock related.
These side-bands with a fair bit of certainty (if not 100%) do not come through clock contamination, but from simple analogue noise.
So they are not jitter in the literal sense of the word.
Unless we re-define "Jitter" to mean "any noise from any source", in which case we no longer need to treat jitter separately and instead can subsume it under THD&N...
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
. . . . .
Power supply related noise can disturb clock.
If the disturbance has some periodicity, there will be sidebands.
....how many DAC's have you measured, for example for jitter? One does not need an AP2, ULP or dScope for that either (though something better than a 3rd octave RTA is needed in most cases)?
What you dig out is old news to me. And I will repeat, from experience the most likely cause of the results measured is not actual jitter, not PSU noise disturbing the clock or any of this, it is a simple and old fashioned operator error by the person setting up the measurement.
And I got that experience by making similar mistakes on occasion and learning to avoid repeating them subsequently.
I understand that you do not like my conclusions and that you want desperately believe that the particular test is not in error (even though alternative tests that fail to show the same problem exist in the public domain.
But honestly, that is your problem, not mine.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
I think your claim is completely wrong.
If the measurement set up invites so much 60Hz, 120Hz, 180Hz, 240Hz, . . . . . noise so that it can cause such a high jitter from a non-defective well-designed hi-fi DAC,
it is impossible to have this graph: THD at 0 dBFS (2.2V RMS), 22kHz measurement bandwidth.
Hi,
> If the measurement set up invites so much 60Hz, 120Hz, 180Hz, 240Hz, . > . . . . noise so that it can cause such a high jitter from a
> non-defective well-designed hi-fi DAC,
>
> it is impossible to have this graph: THD at 0 dBFS (2.2V RMS),
> 22kHz measurement bandwidth.
Please, based on your understanding of the R&S ULP analyser and the specific way the above measurement (note it is THD NOT THD & N) is operated, provide an adequate explanation as to why mains related hum would show up in the test you reference.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
If there is high level 60Hz signal, how can it have low THD at 20Hz?
"The effect of power supply ripple on jitter is quite straightforward."
Silicon Laboratories Inc.
http://www.silabs.com/Support%20Documents/TechnicalDocs/AN491.pdf
"The serial numbers and production dates are very close."
That does not indicate anything.
A product can be non-defective, but the product that was packed in the factory right after it can be defective.
It is possible that two adjacent units off a production line might have a critical component from two very different manufacturing lots. This could account for a large variation in performance between two successive units. I know this from personal experience. I was managing a line of hardware products back in the 1970's and there was one unit that was poorly designed. It had a poor quality hardware design and the firmware was designed and coded by an incompetent dishonest man. (Obviously we didn't realize the full extent at the time, but later we wasted millions of dollars redesigning the product using a super-competent engineer who did both the hardware and firmware.) We built a pilot run of a few dozen units and tested them. The product appeared to work, not at full speed, but it did work. At that point we cranked up the manufacturing line. As soon as we started shipping products to customers we began to get complaints. All the pilot production units came from one lot number. Another lot number had slightly different timing performance and there was a race condition in the firmware that caused intermittent crashes. I remember this situation quite clearly, because it seemed rather mysterious at the time. Now, I would dismiss it completely as a possibility because over the years I learned to distinguish between two kinds of "engineers", those who don't know what they don't know and so run into problems and those who do know what they don't know and take steps to ensure their ignorance doesn't create problems.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
People who connect things aren't to blame. People who publish things without reconnecting things and figuring out what is going on are to blame.
I learned this lesson in high school physics lab where we were using a pendulum to measure the acceleration of gravity but we ended up using the pendulum to measure the power line frequency. Collecting measurements without knowing what you are doing is worse than useless.
It could be a defective DAC, defective test equipment or defective wiring. Before publishing gross measurements this should have been tracked down. This would include, as a minimum, using a different product sample and/or different test procedure.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
He put Benchmark DAC in the place of Musical Fidelity V90.
"I popped my Benchmark DAC1 HDR in to reality-check my lab setup; it was as clean as always." © 2013 KenRockwell.com.
If there was some kind of a ground loop it might be hard to track down the cause. It might disappear when different equipment is used. Or it could be hum in the digital source that is causing jitter that the ASRC in the benchmark filtered out. Or, (most likely) it could be defective power supply components in the Peachtree DAC clock circuitry.
I would bust the reviewer for calling the product's sound "neutral" while it had such gross hum modulation of the audio signal. He should have heard something wrong with the DAC's sound. One would hope that a reviewer who can't hear problems would at least be competent at measuring and interpreting results. This does not appear to be the case. (I've been to the web site and previously reached a conclusion about his reviews.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Do you think the reviewer did something (connection or placement) that no reasonable-minded audiophile in the world will ever do at home?
It is hi-fi designer's responsibility to make it jitter free in wide range of possible settings.
The reviewer is not wrong about not feeling the jitter sidebands. They are below the well known masking curve. Nothing dishonest about that.
It is of course possible that some other people hear that kind of jitter.
Hi,
> The reviewer is not wrong about not feeling the jitter sidebands.
Why not? Anyone who cannot hear hum at -60dBFS is pretty much deaf.
And FM at -60dB the same.
Of course, there is a good chance that when the unit was auditioned these hum components were absent (as I think anyone would have tried to eliminate such obviously audible problems).
As said, the problems are likely the result of not setting up the test correctly and overlooking something that would have been obvious to a competent professional.
Ciao T
At 20 bits, you are on the verge of dynamic range covering fly-farts-at-20-feet to untolerable pain. Really, what more could we need?
They are not -60dbFS hum.
They are (11000 +/- 60)Hz tones below 60dB of 11000Hz tone. Please do internet search about masking curve.
Now what is the level 60Hz tone you are claiming to be induced by wrong set up? See the whole review and figure it out. Do not imagine!
The spectra with the spuria shows only a narrow frequency range. We don't know what the level was at 60 Hz during this test. This presentation, by itself, shows the incompetence of the reviewer. He should have given a complete spectra, followed by zooming in to the high frequency peak. I've seen more than my share of doctored graphs and charts. Shifting the axis is a standard method of lying by statistics. (These graphs are statistics, by the way.) In this case, however, I suspect incompetence rather than dishonesty. Given that the DAC didn't have a polarized plug, something as simple as reversing the power plug might have cured the problem.
Note that this reviewer is not a soft-touch, feel-good non-technical subjectivist audiophile. The web site trys to project technical expertise and is written from an objectivist viewpoint. I came upon this web site a few months ago and after reading an number of articles, I concluded that the web site was worthless, a relic of the Julian Hirsch reviews in Stereo Review.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"Given that the DAC didn't have a polarized plug, something as simple as reversing the power plug might have cured the problem."
If reversing the plug can cause such a high jitter, then it is design defect or manufacturing defect. It is not at all fault of the user or the reviewer.
If the reviewer were competent he would have understood the possibilities and would have discovered and reported that the product sample was sensitive to ground loops. This would have been useful information. However, he did not do this.
I find it hard to know whether a product is good if I am forced to deal with information that comes through an obviously incompetent reviewer.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: