The HDAM in Marantz was nothing more than a name for the metal can they put over their small signal circuitry.
I had a PM6800(?) and now have a PM7200 that does workshop duty blah blah. Both tout HDAM. When the PM6800 blew (years ago now) I lifted the can off in the power amp section to see what was underneath - it was just the front end to a very run of the mill VFA amplifier.
HDAM - marketing but I will refrain from taking that statement to its obviously logical conclusion!
I had a PM6800(?) and now have a PM7200 that does workshop duty blah blah. Both tout HDAM. When the PM6800 blew (years ago now) I lifted the can off in the power amp section to see what was underneath - it was just the front end to a very run of the mill VFA amplifier.
HDAM - marketing but I will refrain from taking that statement to its obviously logical conclusion!
With the greatest of respect, that's just not correct John.
If you want a bit more dynamic range then go discrete - but almost any amp on the planet will clip with 2V RMS input anyway.
You can't beat the distortion (ppb), PSRR (100dB+), slew rate (20V/us to lightspeed), DC performance (10uV/C, <1mV offset) or the fact that an IC is super compact so its less susceptible to hum pickup etc. As for RFI - use a filter or a FET input opamp. If you have RFI, you most likely have an implementation problem anyway.
This is 2019 - get with the program man! Embrace the future!
Douglas has been a jobbing consultant most of his career. He gets called in, designs something and moves on. Not sure if he ever had a permanent role at either of those companies.
What do you call it when you incessantly diss IC's with no more proof than it 'sounds better'?
As you well know, the inner conviction that it "sounds better" is unfortunately enough for many here.
Everyone has their reasons. Maybe they truly believe it. Maybe it's a deep-rooted defense mechanism to avoid the harsh truth: that the cheap, mass-market op-amp has brought better performance in many cases than discrete circuits can achieve, and it's brought it to less talented engineers, too. It must be hard to accept (subconsciously?) for many designers that devoted their creativity and passion to the pursuit of goals which are now better realized by tiny ICs from a giant corporation at dirt cheap prices.
Since there is no real effort to prove these audible differences exist, it'll forever remain a holy war. Until the entire industry fades into complete obscurity, anyway.
It is almost exactly like arguing about religion. Those in the so-called objectivist camp will be labeled as closed minded, but is that really the truth? I think most objectivists would gladly accept the "proof" of a well-conducted test. Meanwhile, I do not think ANY argument could change the mindset of subjectivists.
You could create an op-amp, let's say it has 100V rails, 1A output current, 0.1 nV/rtHz JFET inputs, -160 dB distortion @ 100 kHz, 1 GHz unity gain bandwidth and 5000 V/uS slew rate and a bunch of people would still say it's inferior to their discrete circuits. Arachnophobia
.
Last edited:
indeed. But sometimes we have to attempt to set the record straight.
(I do use discrete circuits - but for line stages, opamps reign supreme IMV and for bog standard MM you’d be hard pressed to beat them if you have a practical, engineering hat on - and given the noise levels coming off vinyl).
(I do use discrete circuits - but for line stages, opamps reign supreme IMV and for bog standard MM you’d be hard pressed to beat them if you have a practical, engineering hat on - and given the noise levels coming off vinyl).
Agreed, and obviously there are still places where discrete has advantages.
I don't have an insider's perspective, but it seems like the consolidation of the semiconductor industry is going to force even greater dependence on ICs as they continue to kill off low volume discrete parts.
Last edited:
Isn´t it also true, that the "inner conviction" (based also often purely on subjective evaluation) that something doesn´t sound differently is unfortunately enough for many around?
Let´s face it, to discover the real cause and effect relationship for any (maybe) audible difference is often a difficult task.
But i totally agree that the generalizations are often based on questionable data. Be it in case of audibility or inaudibility.
We should avoid to talk about "proof" as it isn´t possible to get this kind of certainty from any empirical process.
Furthermore, isn´t it unreasonable to demand evidence without being able to specify which kind of evidence would be accepted?
It does appear to be two sides of a coin, sure. Everyone has their default position.
My bar is pretty low for evidence. I'd like to see some people pick out differences in a live (blind) test in a repeatable way.
Anyway, I'd rather not get into a really deep existential debate with you on this. It's probably something we'll never agree on and I don't need to spend more time arguing on the internet
.Really, what we should be discussing is how bad the writing for Game of Thrones has been since Season 4.
Last edited:
<snip>
My bar is pretty low for evidence. I'd like to see some people pick out differences in a live (blind) test in a repeatable way.
Anyway, I'd rather not get into a really deep existential debate with you on this. It's probably something we'll never agree on and I don't need to spend more time arguing on the internet
As it is in this case just about your personal bar, there is no need for a "deep existential debate..."
, but does "live (blind)" mean that you have to be personally present during the test event? Can you specify a criterion for the "repeatable way" ? Something like a (statistical) significance level?And imo most important, are you sure about your "pretty low bar" ?
It happened over the years quite frequently that in case such evidence occurred that additional demands came up.
Ah but on PSRR is precisely where discrete can beat ICs. The reason being very simple - ICs have their Ccomp to the substrate typically and there's no 0V pin on an opamp. Samuel Groner builds discrete opamps with a 0V connection, presumably because he recognizes the -ve rail PSRR is compromised by taking Ccomp to the -ve rail.
I've just remembered there is one fairly recent TI opamp which breaks with the 8pin tradition and does have a 0V connection. Can't recall its number though.
Last edited:
As it is in this case just about your personal bar, there is no need for a "deep existential debate..."
And imo most important, are you sure about your "pretty low bar" ?
It happened over the years quite frequently that in case such evidence occurred that additional demands came up.
I'd have to think about it, but get me even a few people that can do it to a statistically significant degree, and yeah - being present would be nice.
It's true that the PSRR of the rail attached to the Ccomp is compromised, but there's still the OLG of the op-amp. For a modern op-amp, that's going to be a pretty big number in the audio band.
Yes, i missed that part in Bonsai´s post.
Ahuja wrote about these effects in 1983:
Ahuja, B. K. (1983). An improved frequency compensation technique for CMOS operational amplifiers. IEEE Journal of Solid-State Circuits, 18(6), 629–633.
I don't have IEEE access anymore, but I know OPA1622 has a GND pin as one example.
Yep granted, but I for one don't consider merely the in-band PSRR to be relevant.
- Status
- Not open for further replies.