Regarding the Audio Science review of the JC2, it seems they went ahead a measured it without regard to the "philosophy" from which it was designed; instead they measured it according to their own philosophy - the lower the better, even if that's well into the nanoscopic...
One of the reply's mentioned something along the lines of perhaps this is the product of a good designer, one who can hit the target of what he wants to achieve, through skillful application of electronic circuit design. So for all we know the JC2 does exactly what John wanted it to do - but the guy wielding the AP analyzer doesnt understand that - all he understands is 0.00006 is better than 0.0001.
Now if the design philosophy in the context of "How It Woks" were etched in cursive script into the top cover of each JC2, perhaps the reviewer would read it first, before applying sine waves and looking for deviations from mathematical ideals. Then perhaps the conclusion would be "it does exactly what the designer intended it to do". Then he could offer his rating based on agreement with the design philosophy - or not.
I mean, would the AS reviewer with his AP analyzer agree with an idea like insertion of a vacuum tube into the audio signal path fixes certain "sins" of audio circuit design - and there's practically no other alternative? I doubt it...
One of the reply's mentioned something along the lines of perhaps this is the product of a good designer, one who can hit the target of what he wants to achieve, through skillful application of electronic circuit design. So for all we know the JC2 does exactly what John wanted it to do - but the guy wielding the AP analyzer doesnt understand that - all he understands is 0.00006 is better than 0.0001.
Now if the design philosophy in the context of "How It Woks" were etched in cursive script into the top cover of each JC2, perhaps the reviewer would read it first, before applying sine waves and looking for deviations from mathematical ideals. Then perhaps the conclusion would be "it does exactly what the designer intended it to do". Then he could offer his rating based on agreement with the design philosophy - or not.
I mean, would the AS reviewer with his AP analyzer agree with an idea like insertion of a vacuum tube into the audio signal path fixes certain "sins" of audio circuit design - and there's practically no other alternative? I doubt it...
There has been a phono stage with LED's and PV for years. There is also an old patent for laser powered retinal implants. We custom fabbed some 55V PV arrays and patented an optically isolated MEM's switch that ran off a cell phone battery (no switching converter). We thought they would be all over it for antenna switching but no dice.
Nope just checked. Both the original JC2. Although why is there a photoshopped black box over the serial number? What are they afraid of?
Looking at the stereophile numbers distortion starts to rise after 3.5v output but JA got numbers 1/10th of ASR. I think drive level too high and waaaaaaay above domestic audio levels. Different gain settings, very low load for ASR (Does anyone expect a domestic unit to drive 300 Ohm power amplifiers?)
Or maybe they want their preferred reference to look better?
The operative word here is "works" as a former member here pointed out in that thread any carefully controlled listening comparison according to any mutually agreed upon protocol has not (and probably never will) happen.
That said I was very disappointed in the review myself. I have not seen evidence that -90dB H3 at full scale is audible while the data presented did not make clear or discuss the train of low frequency IM products (10k/20k IM test) visible in the Benchmark pre-amp. It was not clear, because the plots were superimposed if they were also on the JC-2. A train of equally spaced spurs can have an inverse transform that is a discontinuity which can have a peak amplitude more than one would compute from a simple rms addition. My efforts to get some tests done like THD without noise vs level or showing residual signal after nulling out the input fall on deaf ears, I assume because in some cases you can't just push a button on the AP.
300 Ohm is a typo, it was 200K. They have an edit with the correction. But testing at 4V is just wrong IMO. Hopefully John can comment on that.
Good points. But then AS could also have a plaque in front of the "entrance" that says something like "We don't give a #### what the designer intended, we just measure and report".
I hoped you not looked at 1kHz only. 1kHz result is pointless. I care about that rapid rise of distortion with frequency.
https://www.audiosciencereview.com/...-n-vs-frequency-audio-measurements-png.74050/
This shows poor engineering, and reflects in IM and multitone plots. This may add some color.
I have a JC-2 clone from China and it drives my HD650's just fine. I would stop short of any accusations like that, though some of Benchmark's "technical" copy leaves much to be desired.
Agreed.
Depends on your definition of "engineering". To me, the topology and the design are good, so I see only 2 options:
- Either ASR did something wrong
- Or it's a faulty unit, maybe mismatched JFets: I've seen it happen, once (not with Parasound); so it could be a production issue.
Without a harmonic structure it is hard to tell, could be a mis-biased output stage too.
JA didn't, at least not to such dramatic extend as ASR, hence my option #1: maybe ASR did something wrong.
Not only was the measurement done at a bit high of a voltage but the other issue is modeling the actual music spectrum to get realistic distortion results.
If a CD player is producing a maximum of 2 volts with a real music signal, we could expect that to drop to 1 volt by 1,000 hertz and .25 volts by 16,000 hertz, although the Meyer M noise data suggests the peak to average level increases with increasing frequency.
The point being measuring the wrong things doesn't yield useful data.
Bill what power amplifiers can produce less than .001% THD at a watt or do? (Often the output power mostly used for real music.)
If a CD player is producing a maximum of 2 volts with a real music signal, we could expect that to drop to 1 volt by 1,000 hertz and .25 volts by 16,000 hertz, although the Meyer M noise data suggests the peak to average level increases with increasing frequency.
The point being measuring the wrong things doesn't yield useful data.
Bill what power amplifiers can produce less than .001% THD at a watt or do? (Often the output power mostly used for real music.)
Simon: The challenge as you know is meausuring the THD* on its own as you are generally noise floor limited at that level. But the LM3886 based amps TomChr does are one example. Hans has just published his amplifier schematic that does fractional ppm. I still prefer to look at the 32 tone test results.
Attached some measurements from Bruno's G_word preamp at +18dBu. Basically measuring the AP noise floor. That is what I would like my preamps to do. Total and utter overkill but if you can...
* Whilst accepting that some of the greybeards here just think thats yoof being lazy and expecting the analyser to do it all with no brain required.
Attached some measurements from Bruno's G_word preamp at +18dBu. Basically measuring the AP noise floor. That is what I would like my preamps to do. Total and utter overkill but if you can...
* Whilst accepting that some of the greybeards here just think thats yoof being lazy and expecting the analyser to do it all with no brain required.
Attachments
A lot of these discussions rest on assumptions of monotonicity of distortion with level, and the additional assumption that distortion components below the noise floor must be unimportant. The first is almost always wrong, and the second is a WAG.
For a conventional preamp output level at a conventional amplifier's sensitivity of, say 1 Volt for +17dBW, a conventional speaker sensitivity of, say 88dB SPL 1W1M, +3dB for two non-correllated speakers, minus, say 8dB for room gain, 0VU signal level at a quite loud 75dB SPL at seated position is -8dBV. This is "VU" forte level, and not peak.
Music from acoustic sources should be expected to reach down at least 30 or 40 dB below this and very good arguments can be made that these levels are more important to get right than forte. My takeaway here is that preamps should be measured at -38dBV and considerably lower to get measurements that mean anything remotely useful.
Too hard to do? That's why I'm looking for my wallet under the streetlamp.
All good fortune,
Chris
For a conventional preamp output level at a conventional amplifier's sensitivity of, say 1 Volt for +17dBW, a conventional speaker sensitivity of, say 88dB SPL 1W1M, +3dB for two non-correllated speakers, minus, say 8dB for room gain, 0VU signal level at a quite loud 75dB SPL at seated position is -8dBV. This is "VU" forte level, and not peak.
Music from acoustic sources should be expected to reach down at least 30 or 40 dB below this and very good arguments can be made that these levels are more important to get right than forte. My takeaway here is that preamps should be measured at -38dBV and considerably lower to get measurements that mean anything remotely useful.
Too hard to do? That's why I'm looking for my wallet under the streetlamp.
All good fortune,
Chris