Hi Dejan,
Would you accept that those new devices compared to the old devices are not the problem? You are comparing a specific couple of cases to the entirety of the market today. I'll move that there were plenty of hideous sounding amplifiers back then, and not a few of them were very unreliable on top of that!
I think that if we look at the averages, the bar has been on a steady climb up. Reliability, well that has decreased steadily. But this is only due to design practices and not the parts involved.
Yes, we really do have our priorities mixed up. But these priorities have to do with the poor manufacturing practices we have deemed to be acceptable.
-Chris
Would you accept that those new devices compared to the old devices are not the problem? You are comparing a specific couple of cases to the entirety of the market today. I'll move that there were plenty of hideous sounding amplifiers back then, and not a few of them were very unreliable on top of that!
I think that if we look at the averages, the bar has been on a steady climb up. Reliability, well that has decreased steadily. But this is only due to design practices and not the parts involved.
Yes, we really do have our priorities mixed up. But these priorities have to do with the poor manufacturing practices we have deemed to be acceptable.
-Chris
Personally, I don't know much about 'clocks' but I DO KNOW that no cost effective manufacturer, including OPPO, will give you the best 'clock' that money can buy. So, if someone changes the clock (not my first choice with my OPPO, there are so many other compromises), I take them at their word that they are trying to improve the product, and they probably are improving the product. Just because Jan or SY can't believe that any difference is possible, doesn't make it so.
Let's be sure we are on the same page. We are talking about clock jitter. This is easily coupled into an FM tuner. So if one has a broadband receiver that can pick up FM at the clock frequency, then all that is required is a coaxial cable input that can just be loosely coupled into the clock signal.
Simplest case is the shield is grounded and the inner conductor stripped bare for a few inches and placed near a clock lead. Worst case use a few pF of coupling capacitance.
The tuner only needs microvolts the clock should be volts.
Often there is enough leakage you can hear the jitter from outside the case.
This will give you a relative measurement not an absolute value.
ES
Chris, I'll agree that reliability has improved, and not by a small margin. No argument there.
However, quality of workmanship, inluding strudiness, has definitely declined overall. Just look at those funny "shark fin" heat sinks and compare with similar models previous to them.
Power supplies haven't fared all that well either. Being expensive, they were always the weak point of most consumer orientated audio, but these days they show little, if any, progress.
Admittedly, relatively speaking, modern audio is cheaper than bafore. Modern manufacturing has taken care of that. In that respect, some good strides have been taken.
In the end, I feel that the average sound quality has not moved much, if at all. What has moved along are the prices of solid mid priced audio, from well known names. I think most of that fare has relatively increased in price, but offers little in sound quality in comparison with vintage fare from the same sources. True, some old names, like say Marantz, still try to give one a realistic choice, but unfortunately what is really good from them has moved along in terms of prices and is now more of High End pricing for results which are to me mostly modest. You might not agree.
I understand perfectly that audio is not half as lucrative today as it was in the say late 70ies, in its heyday, too many other things have intruded, like the VCR now cum CD, the PC and the latest gadget in form of phones. Anyway, this decreases the backbone of what used to progressive audio, small companies which must be good to barely survive, and we are left with what SY calls "fashon audio" (an excellent term indeed). There is little impetus pushing them on, these days its all about how imaginative your designer is and how good are your CNC machines in making sculptures.
While this rant is necessarily at least partly personal, I have what I wanted to have, it produces results I'm very happy with and from listening to other people's systems, many of them costing several times more than mine, I know I listen to better sound, as I think of it. Much of this is taste, to be shure, but I don't miss out on the expression on people's faces when they hear my relatively modest setup.
However, quality of workmanship, inluding strudiness, has definitely declined overall. Just look at those funny "shark fin" heat sinks and compare with similar models previous to them.
Power supplies haven't fared all that well either. Being expensive, they were always the weak point of most consumer orientated audio, but these days they show little, if any, progress.
Admittedly, relatively speaking, modern audio is cheaper than bafore. Modern manufacturing has taken care of that. In that respect, some good strides have been taken.
In the end, I feel that the average sound quality has not moved much, if at all. What has moved along are the prices of solid mid priced audio, from well known names. I think most of that fare has relatively increased in price, but offers little in sound quality in comparison with vintage fare from the same sources. True, some old names, like say Marantz, still try to give one a realistic choice, but unfortunately what is really good from them has moved along in terms of prices and is now more of High End pricing for results which are to me mostly modest. You might not agree.
I understand perfectly that audio is not half as lucrative today as it was in the say late 70ies, in its heyday, too many other things have intruded, like the VCR now cum CD, the PC and the latest gadget in form of phones. Anyway, this decreases the backbone of what used to progressive audio, small companies which must be good to barely survive, and we are left with what SY calls "fashon audio" (an excellent term indeed). There is little impetus pushing them on, these days its all about how imaginative your designer is and how good are your CNC machines in making sculptures.
While this rant is necessarily at least partly personal, I have what I wanted to have, it produces results I'm very happy with and from listening to other people's systems, many of them costing several times more than mine, I know I listen to better sound, as I think of it. Much of this is taste, to be shure, but I don't miss out on the expression on people's faces when they hear my relatively modest setup.
With all the criticism about 'classic' audio designs, (you know, the ones that have won lots of awards over the decades), I am surprised that these same people want to contribute to this thread. You know, why bother?
It is true that, in general, trimpots are rather impractical. That is why we often Idss match jfets in advance, and use servos for ultimate output. Even IC's often get laser trimming, what is wrong with an occasional trimmer? Most trimmers, like those who set output bias in power amps, can be cermet, and rather cheap to buy. Only, at extremely critical levels, like 0.4nV/rt Hz, only the best military grade wirewound pots will be necessary, or you can initially trim with fixed resistors (relatively easy) to remove the need for the trimpot. Usually, the initial setting is OK, with the Vendetta, for example, over the life of the product, (more than 30 years), because a servo will take up the slack, if necessary.
It is true that, in general, trimpots are rather impractical. That is why we often Idss match jfets in advance, and use servos for ultimate output. Even IC's often get laser trimming, what is wrong with an occasional trimmer? Most trimmers, like those who set output bias in power amps, can be cermet, and rather cheap to buy. Only, at extremely critical levels, like 0.4nV/rt Hz, only the best military grade wirewound pots will be necessary, or you can initially trim with fixed resistors (relatively easy) to remove the need for the trimpot. Usually, the initial setting is OK, with the Vendetta, for example, over the life of the product, (more than 30 years), because a servo will take up the slack, if necessary.
To Jan's question
I think we have to ask what freedom of design we get when we can use a mix of discrete components and integrated ones. Consider for the moment just small-signal design.
Among the attributes of discretes: unlimited power dissipation; unlimited voltages; unlimited physical size. In a sense these are not true limitations of ICs either, but look right now for a JFET-input op amp in integrated form with really low voltage noise. The selections are limited, and one reason is JFETs tend to take up a lot of chip area. So even if yields from a wafer are good, there are only so many devices. Who is willing to pay for quantities of such things? As Scott W. pointed out, some ADI video amp chips are about the same size as a 2SK170 die. The long-discontinued 2SJ72 and 2SK147 (two of each in the Blowtorch per channel) are even larger.
Perhaps with the transition to more three-dimensional structures we can get there, but I suspect it won't be easy or cheap.
There is the school of thought that prizes simplicity. There is also among many the desire to avoid global feedback. So when a design like the Blowtorch is inspected, this seems to fulfill some of these aims. But some component matching and trimming is useful to assure the best performance, and as we know, the approach to complementarity is sound in principle, but difficult to achieve in practice because of the asymmetry in P and N device characteristics. Nonetheless, the performance can be quite suitable for low impedance sources like MC cartridges. The capacitive loading is tractable for such source impedances with low resistances and inductances.
The paucity of decent P parts is another problem.
I think we have to ask what freedom of design we get when we can use a mix of discrete components and integrated ones. Consider for the moment just small-signal design.
Among the attributes of discretes: unlimited power dissipation; unlimited voltages; unlimited physical size. In a sense these are not true limitations of ICs either, but look right now for a JFET-input op amp in integrated form with really low voltage noise. The selections are limited, and one reason is JFETs tend to take up a lot of chip area. So even if yields from a wafer are good, there are only so many devices. Who is willing to pay for quantities of such things? As Scott W. pointed out, some ADI video amp chips are about the same size as a 2SK170 die. The long-discontinued 2SJ72 and 2SK147 (two of each in the Blowtorch per channel) are even larger.
Perhaps with the transition to more three-dimensional structures we can get there, but I suspect it won't be easy or cheap.
There is the school of thought that prizes simplicity. There is also among many the desire to avoid global feedback. So when a design like the Blowtorch is inspected, this seems to fulfill some of these aims. But some component matching and trimming is useful to assure the best performance, and as we know, the approach to complementarity is sound in principle, but difficult to achieve in practice because of the asymmetry in P and N device characteristics. Nonetheless, the performance can be quite suitable for low impedance sources like MC cartridges. The capacitive loading is tractable for such source impedances with low resistances and inductances.
The paucity of decent P parts is another problem.
Just for good measure, John, you can throw in amp PSUs as well. They are still the favorite part for "saving" money. Still weedy, still not even satisfactory, simply built to power the device, in most cases. How - who cares?
Hi Ed,
An iffy FM Front end maybe? A way-too-wide IF filter? You're right about the amplitudes but I would hope that the filtering in the signal chain in a tuner would knock the level of the interference way down. I do remember I had a Philips 'scope that made FM alignment impossible while the 'scope was running.
CD player clocks don't run at 10 MHz, but I can't remember where most of them sit off the top of my head. One Denon with a Sony chip set ran at 16.9344 MHz or 14.9344 MHz. The value wasn't clear. Another looked like 15 MHz. The value wasn't spec's even to one decimal place. I do know that during the adjustment on a 3.28 MHz clock, it would take an FM tuner off the air, but in that case a harmonic was in-band for the tuner. I guess you would need an analogue tuner to figure this out more easily.
-Chris
An iffy FM Front end maybe? A way-too-wide IF filter? You're right about the amplitudes but I would hope that the filtering in the signal chain in a tuner would knock the level of the interference way down. I do remember I had a Philips 'scope that made FM alignment impossible while the 'scope was running.
CD player clocks don't run at 10 MHz, but I can't remember where most of them sit off the top of my head. One Denon with a Sony chip set ran at 16.9344 MHz or 14.9344 MHz. The value wasn't clear. Another looked like 15 MHz. The value wasn't spec's even to one decimal place. I do know that during the adjustment on a 3.28 MHz clock, it would take an FM tuner off the air, but in that case a harmonic was in-band for the tuner. I guess you would need an analogue tuner to figure this out more easily.
-Chris
But as the BAS survey showed these circuits have up to .05% range THD's at full scale even worse than the PONO.
So what do they know from a small survey about what is possible??
-RNM
Not the point at all, right now JC is trotting out the 40yr. circuits once more. As I said last week it's ironic to mix the -120dB THD + noise stuff with the Curl, Hansen, Pass school of design.
I have a nifty AM/FM tuner goes from just about DC to 1 Mhz. I can set it to the clock frequency and see what actually rocks the boat.
Before I got that I used an old tunable vacuum tube wirless mic receiver tuned up to 30 Mhz and it had a nice magic eye tube to be sure you were centered on frequency.
Learned the trick on listening to your audio digital boards from a self taught tech who drifted into becoming a manufacturer.
Of course I can measure AES jitter with my AP system 2.
When you lump it all into one basket it may seem that way. -60 on second harmonic distortion is no big deal, it is a big deal on 17th or even a non-harmonic. Throw some noise into the same basket and confuse things even more.
Before FFT type analyzers were used it was hard to distinguish between low level noise and distortion. Now that we have them we ignore things they don't do well.
So when you look at a report of what can be perceived it is important to look at the test setup and what the measurements really mean.
To me the biggie is headphones or loudspeakers? If you are testing a single component (RLC level) and listening in stereo to loudspeakers there are so many variables that the test is useless for negative results. (If you can hear a difference under those conditions the parts really have issues.)
In general all generalizations are wrong.
ES
It's not that I say it is impossible, on the contrary, anything can be improved with enough time nd effort and money.
My point is that those mods demonstrably increase jitter, lower rise- and fall times, add inductance to connections and power supply lines, decrease PSRR and a host of other issues clear to any half-wit designer.
And it does sound better?
Is anybody thinking this through?? Are you John?
Jan
I wouldn't do mods or designs on speculation or assumptions re results. I invest in the test equip needed to Know the results.... so measure the jitter and what ever is done you can measure and see if it is going in the right direction. It takes all the quess work and assumptions out of the picture.
THx-RNMarsh
THx-RNMarsh
Who needs unlimited size or power? We have a requirement for some audio thingamegong. I posit that we can fill that much better, higher quality any way you define it (except psychologically*) for less money, more reliable and repeatable than half a century ago, IF only we act as we did when we also were a half century younger. THAT is the basic problem isn't it? We lament the demise of 40 years old JFETs but we don't see the rich landscape of opportunities that modern parts give us. Because we never really dig in, we just regurgitate 40 years old stuff.
Who was it again who said that new ideas, developments, opportunities do not grow smoothly, but the old generation has to die off before the new takes hold?
It's all fine but don't blame ' the engineers' who don't give us better audio. It's WE who fail.
Jan
But Richard, we know THD numbers don't tell us anything (or very little) about audio performance. So why buy better and better and more and more THD analysers? I can't see how that could help at all.
Jan
"Everything is illuminated by the aura of nostalgia, even the guillotine.”
Just a gentle fight between two opposite schools: Simple, and highly optimized to reduce noises and distortions or Distortions reduced by design (complex).
I have no idea of the best way. Somewhere in the middle, as usual ?
- Status
- Not open for further replies.