A capacitor operates only on AC. The author of the capacitor article somehow doesn't seem to understand this. He says:
Of course there is no sound associated with DC, but the reference to DC is inappropriate here and simply begs the question. A capacitor is intended as a means of rendering that DC rail pure DC by shunting AC to ground. To the extent the capacitor fails in that function will AC be present on that DC rail.
How does a capacitor fail to shunt AC to ground? When it doesn't operate as an ideal capacitor---such as when its DA/Dk creates distortion as seen in the diagram I posted from Bateman's article.
Power supply capacitors and their shunts are theoretically in series with signal amplifying devices for this reason. I see no proper basis to claim that use of a better bypass capacitor is "sheer lunacy," in the author's words. I see badly thought theory.
Your language was similarly this:
Please justify your statement that a capacitor sees no AC signal when its sole job is to pass AC signal.
Of course there is no sound associated with DC, but the reference to DC is inappropriate here and simply begs the question. A capacitor is intended as a means of rendering that DC rail pure DC by shunting AC to ground. To the extent the capacitor fails in that function will AC be present on that DC rail.
How does a capacitor fail to shunt AC to ground? When it doesn't operate as an ideal capacitor---such as when its DA/Dk creates distortion as seen in the diagram I posted from Bateman's article.
Power supply capacitors and their shunts are theoretically in series with signal amplifying devices for this reason. I see no proper basis to claim that use of a better bypass capacitor is "sheer lunacy," in the author's words. I see badly thought theory.
Your language was similarly this:
Please justify your statement that a capacitor sees no AC signal when its sole job is to pass AC signal.
Last edited:
DF, what do you actually mean by "bad"? Ceramics don't stop operating as ceramics---their DA/DF/Dk ratings don't change---because they're placed here or there. They're not "good" here and "bad" there. They are what they are, or in other words, they are characteristically nonideal for that type of capacitor wherever you put them.
Abrasiveness is good as a stage of cyclic development. Dreaming, construction, criticizing. Circle closed, stop criticizing, go back to dreaming. When Walt Disney learned this strategy he become very productive. He taught his staff. He used 3 different rooms for 3 different tasks. Only one of them was meant for criticizing.
That's correct. A (PSU) bypass capacitor sees all the AC ripple, but a coupling capacitor sees almost no AC in case the frequency is far above 1/2*pi*R*C, for the C-R hipass filter.
The capacitor sees little or no AC VOLTAGE (if your supply rails have 3V AC at 1kHz it is a bad design). No voltage no distortion. Also again DA does not make distortion the voltage coefficient of capacitance does. The AC displacement current does NOT have the distortion. If you don't differentiate between AC signal as current vs voltage, any circuit analysis is difficult.
A properly designed supply would use ceramics as RF bypass since the current loop and parasitic L are minimized. The rails are already stiff at audio frequencies with larger caps of other material.
Last edited:
There used to be a "definition according to Paul" of a design review in memo form (1975 or so). Quite something.
Impulse ripple (e.g. during contactor switching in the power distribution net), burst shaped with damped HF oscillations in MHz or tens of MHz, is much higher. Depends on filter and PSU design, and on interconnection of the audio components (instruments) as well.
Depends on the "story." If you have something colorful that the chimp reviewing the amp can cut and paste or reword, there's a good chance that the review will be positive. If it's just one more amp from Panasonic (or whatever), the best you could hope for is "ok for a mass market amp, but not really high end." The good measurements would probably give it away in the Atkinson section, and in the conclusion, you'd see something like, "The measurements are impeccable and give no clue as to why the amp sounded flat and lifeless in AK's system. Once again, we have a long way to go before measurements can distinguish the great from the merely competent."
Sadly, I suspect that is, by and large, true.
Years ago the astronomer Robert Kraft was at Lick Observatory, the same time I was there with the photodiode array spectrometer, and gave us his quick rundown on how to get a Nobel Prize. The crucial ingredient: "You gotta have a gimmick."
To that end I have a few gimmicks in mind for some bits of equipment. Whereas they will also measure impeccably, I know that at best those measurements will have a neutral influence when it comes to sparking interest in product. What will matter is the gimmick factor, which will provide material for reviewer copy and stimulate the subjective evaluator to listen for something distinctive. And if you are particularly fortunate, you may be blessed with a cover photo, perhaps suitably embellished with some artwork tying into how the editorial staff interprets one or more of your gimmicks.
It is difficult not to become a bit cynical. The attempt to introduce a human side to audio equipment is also rather a stretch. I see people in here and elsewhere pronouncing on branded stuff in a way that suggests a mental picture of activities at some imagined company, say Harman-Kardon. But the reality is usually drastically different than the image, for larger conglomerates at least, with the brand name having no distinctive character and little or nothing to do with any particular product. The engineers and designers are deployed over the brands, with usually a precious few competent ones doing most of the work.
There are exceptions, where a company or company division has remained relatively insular. But even then... I did work for one famous midwest firm, the bust of whose founder was prominent in the lobby. I was assured that there was something very special and distinctive about the way they did speakers and other equipment. Perhaps, but they were unable or unwilling to articulate this. I came to suspect it was something akin to a Masonic lodge, with the essential doctrine imparted in secret ceremonies, were one deemed sufficiently worthy.
In any event I did my best to divine the mysteries, and deduced from an existing design what was permissible or anathema. And in the process I took about 8 dollars out of the bill of materials (which was, percentage-wise, a lot), while realizing something that looked and sounded (and measured) exactly the same. The prototype lasted long enough in their tests for them to agree with that assessment, before probing blew it up. But they didn't follow through, despite sales running about ten thousand units a month. At least they paid my invoice.
I'm sorry if my frustration at continually trying to point out the same technical points appears abrasive. Mr. Elliot might exagerate to make a point but technically he does not make any major mistakes in his article.
Here is an experiment you can try at home, you can even use the poly capacitor feature in LTSPICE to do it in software. Take a really bad .1uF X7R and a 1600 Ohm resistor and sim or measure at 1kHz, it should be bad enough to not need the Bateman oscillator. In LTSPICE you can just set the voltage coefficient high enough to get lots of distortion.
Now parallel it with a good 1uF and reduce the resistor to 160 Ohms. Ok now you have the same voltage (close enough, tweaking to exactness is not necessary) across your X7R, but lo and behold the seconds are 22dB or so lower. The driving impedance of the distortion is simply the reactance of the .1u capacitor but it forms a capacitive divider with the 1u. Now extrapolate that to .1u in parallel with 20,000 - 30,000 uF in a power amp power supply.
Other issues will turn up in the real world but that's what design is all about.
bcarso - cool, opto-electronics is my thing these days.
Last edited:
In defense of audio reviewers, I find that calling them 'chimps' to be an insult. I also find that impugning the motives of the reviewers is misleading as well.
It is a difficult position that reviewers are in. They are supposed to be 'cultured people' with a love of classical music, yet technical people who know their way around test equipment, who are faced with evaluating audio equipment that they neither designed or completely know the design decisions of the equipment's designers. They are forced to chose between a wide array of audio products, and try not to hurt some newcomer's (or well established company's) feelings, by being overly critical, and even drawing a lawsuit from it.
Now, John Atkinson was been put up here for criticism. I hardly know John Atkinson, and I usually speak to him, once a year, yet I first met him about 30 years ago.
Let's look at JA's vitals: Degree in physics, (gee that is a start), has used the fanciest audio test equipment available for decades. Does his own audio measurements. Records classical music on the side. Plays a musical instrument. Maybe, just maybe, he is reasonably well qualified to be an 'objective' audio reviewer.
And what does he find? He finds that mere audio measurements, no matter how precise, do not well correlate with audio subjective reality. Furthermore, he finds that he can also be 'fooled' by double blind tests, such that the conclusion derived from them (usually a null) is utterly useless (just like me) with further long term listening in his own audio system. Yet he has to put out a 'credible' publication every month and try to please everybody. I would not want his job, too much stress.
It is a difficult position that reviewers are in. They are supposed to be 'cultured people' with a love of classical music, yet technical people who know their way around test equipment, who are faced with evaluating audio equipment that they neither designed or completely know the design decisions of the equipment's designers. They are forced to chose between a wide array of audio products, and try not to hurt some newcomer's (or well established company's) feelings, by being overly critical, and even drawing a lawsuit from it.
Now, John Atkinson was been put up here for criticism. I hardly know John Atkinson, and I usually speak to him, once a year, yet I first met him about 30 years ago.
Let's look at JA's vitals: Degree in physics, (gee that is a start), has used the fanciest audio test equipment available for decades. Does his own audio measurements. Records classical music on the side. Plays a musical instrument. Maybe, just maybe, he is reasonably well qualified to be an 'objective' audio reviewer.
And what does he find? He finds that mere audio measurements, no matter how precise, do not well correlate with audio subjective reality. Furthermore, he finds that he can also be 'fooled' by double blind tests, such that the conclusion derived from them (usually a null) is utterly useless (just like me) with further long term listening in his own audio system. Yet he has to put out a 'credible' publication every month and try to please everybody. I would not want his job, too much stress.
The thing I see over and over is an engineer type will think a design has a nice gimmick/feature but the average public does not see it. Most average Joe's just don't think like an engineer.
Case in point; Alesis Studio Electronics. After Russell Palmer left the company it flipped from being sales driven to engineering driven. That lead to a straight, 3 year path from $10 million a month to BK.
Also the flip side; Brad Plunkett with what would become the Wah Wah pedal. He didn't see it. Basically it took Del Casher to see the potential of that and run with it.
Gosh, now I wonder how many gimmicks have just wispped right past me.
Case in point; Alesis Studio Electronics. After Russell Palmer left the company it flipped from being sales driven to engineering driven. That lead to a straight, 3 year path from $10 million a month to BK.
Also the flip side; Brad Plunkett with what would become the Wah Wah pedal. He didn't see it. Basically it took Del Casher to see the potential of that and run with it.
Gosh, now I wonder how many gimmicks have just wispped right past me.
- Status
- Not open for further replies.