ASR "Myth" regarding "discrete" circuitry...

[mansr]

these things are can be achieved with discrete as well, albeit with more effort

So why bother?

hence my pointing out that the best measured DAC by ASR is a design which uses both a discrete converter stage and a discrete output stage

Do you realise that the Moolah Moolah uses three large DSPs and dozens of other ICs to reach that performance level?

 

[solderdude]

you can build your own discrete 741 opamp.

 

[Blumlein 88]

This analogy is not really relevant: Vacuum tubes and transistors/FETs use fundamentally different principles for their transfer function. IC OPAs are just regular transistor/FET circuits, implemented in a chip. A discrete circuit is not fundamentally different (necessarily) from the circuit in a chip, it can even be identical. ICs are an advantage because they are smaller, cheaper, and easy to implement for a designer (the circuit is mostly already defined inside the chip). None of those advantages make them actually perform better.
ICs do have the advantage of thermal tracking, and as mentioned by some folks, they (usually) do not require matching-but these things are can be achieved with discrete as well, albeit with more effort. For example, discrete circuits can use dual package transistors/FETs for matching and thermal tracking.

The disadvantage of using ICs are that they are already a defined circuit, perhaps a good circuit, if they suit the exact needs of the design at hand, but they may not be the best circuit for the task at hand. A really good audio design engineer, may, be able to design a discrete circuit which better suits the needs of a given component, in other words, the design engineer can be more creative with discrete circuitry, which may allow for an improvement in performance (by not being locked into the design of the available ICs).

My intent with this thread is not for it to become a discrete vs. IC thread (or to suggest that discrete or IC is "better"), the point is only to show that there is not a fundamental performance advantage to using ICs (other than ease of design and cost factors), hence my pointing out that the best measured DAC by ASR is a design which uses both a discrete converter stage and a discrete output stage. Another example of a discrete circuit performing really well is the NC-400 amplifier modules, which use a fully discrete input stage design and discrete voltage regulators as well.

Long bunch of writing to state something basically obvious and informative of pretty much nothing. As some others have said, the performance is the performance however it was achieved.

Also the Philbrick tube op-amp wasn't a chip of course, but it was an op-amp.
http://www.philbrickarchive.org/

 

[barrows]

The fact remains that the cheaper 'discrete' designs (for the sake of offering the word discrete because so many desire it) do not perform better than most much cheaper devices using off-the-shelf opamps/IC's.
This is what counts in the evaluations here.
When something measures well it is irrelevant whether it is discrete or 'chip' based or combinations there-off.

Also one must realize that when designing opamps one can use components that are not available in discrete components.
Also shorter paths can be beneficial for fast switching designs.

One cannot glorify discrete over chip based nor chip based over discrete.
In the end it is the final result of the device itself and its performance which matters.

As mentioned more than once on this thread, I am not "glorify(ing)" any design approach, I am just pointing out that blaming discrete circuitry for poor measurements is in error. If a component with a discrete circuit design measures poorly, the cause of that poor measurement is the engineering, and the fact that it happens to use discrete circuitry, that is all I am pointing out here.

 

[Blumlein 88]

As mentioned more than once on this thread, I am not "glorify(ing)" any design approach, I am just pointing out that blaming discrete circuitry for poor measurements is in error. If a component with a discrete circuit design measures poorly, the cause of that poor measurement is the engineering, and the fact that it happens to use discrete circuitry, that is all I am pointing out here.

Yes, you've been pointing it out while I've not noticed much in the way of people saying otherwise.

 

[pkane]

I am just pointing out that blaming discrete circuitry for poor measurements is in error.

You are repeating this, but why? Who here dismissed discrete circuitry as always causing poor measurements? As far as I can tell, most state that both ways can be made to work well, but ICs are easier to implement with good results. Do you disagree?

 

[mansr]

You are repeating this, but why?

Because he wants to have an argument.

 

[SIY]

You are repeating this, but why? Who here dismissed discrete circuitry as always causing poor measurements? As far as I can tell, most state that both ways can be made to work well, but ICs are easier to implement with good results. Do you disagree?

When a peddler touts "discrete circuitry" as an advantage, it is far more probable than not that it's a gimmick and compromises performance. Of course there's a few rare exceptions, but the key word is "rare."

 

[mansr]

When a peddler touts "discrete circuitry" as an advantage, it is far more probable than not that it's a gimmick and compromises performance. Of course there's a few rare exceptions, but the key word is "rare."

In the cases where it doesn't compromise performance, it's still a gimmick and compromises price.

 

[SIY]

In the cases where it doesn't compromise performance, it's still a gimmick and compromises price.

Likely true.

 

[March Audio]

This analogy is not really relevant: Vacuum tubes and transistors/FETs use fundamentally different principles for their transfer function. IC OPAs are just regular transistor/FET circuits, implemented in a chip. A discrete circuit is not fundamentally different (necessarily) from the circuit in a chip, it can even be identical. ICs are an advantage because they are smaller, cheaper, and easy to implement for a designer (the circuit is mostly already defined inside the chip). None of those advantages make them actually perform better.
ICs do have the advantage of thermal tracking, and as mentioned by some folks, they (usually) do not require matching-but these things are can be achieved with discrete as well, albeit with more effort. For example, discrete circuits can use dual package transistors/FETs for matching and thermal tracking.

The disadvantage of using ICs are that they are already a defined circuit, perhaps a good circuit, if they suit the exact needs of the design at hand, but they may not be the best circuit for the task at hand. A really good audio design engineer, may, be able to design a discrete circuit which better suits the needs of a given component, in other words, the design engineer can be more creative with discrete circuitry, which may allow for an improvement in performance (by not being locked into the design of the available ICs).

My intent with this thread is not for it to become a discrete vs. IC thread (or to suggest that discrete or IC is "better"), the point is only to show that there is not a fundamental performance advantage to using ICs (other than ease of design and cost factors), hence my pointing out that the best measured DAC by ASR is a design which uses both a discrete converter stage and a discrete output stage. Another example of a discrete circuit performing really well is the NC-400 amplifier modules, which use a fully discrete input stage design and discrete voltage regulators as well.

You are doing a lot of hand waving. Your points lack any real basis. I don't care if a discrete designer can be "more creative", I just care about the final technical performance.

Your points are moot because no one has said discrete is bad and I don't know where you got this from.

It's also no good repeatedly mentioning Bruno Putzeys designs as examples when he himself has said that the only reason to use discrete is when the circuit is not available off the shelf.

It should also be pointed out that his Purifi design using op amp input buffer performs just as well as his nc400 design.

The only myth that needs to be countered, one that still persists amongst audiophiles, is that op amps are lacking in performance and sound quality and that discrete is fundamentally better. It is often used as a marketing ploy by manufacturers. This is completely untrue.

BTW you didn't mention who you work for in "the high end industry".

 

[direstraitsfan98]

western australia doubled support payments

 

[0bs3rv3r]

The next step is back to tubes. ICs < Discreet < Vacuum Tubes......ha!

You could go for the tube version of an IC. There were tubes with multiple elements. One such had a triode and 2 output pentodes - a power amp in one glass "IC".

https://www.radiomuseum.org/tubes/tube_ecll800.html

 

[mansr]

Your points are moot because no one has said discrete is bad and I don't know where you got this from.

He pulled it out of his arse because he wanted an argument.

BTW you didn't mention who you work for in "the high end industry".

I believe he works for Sonore. Maybe he has a second job with a high-end company.

 

[Chrispy]

Well I just wasted a bunch of time on what seems like an audio salesman's whine. Why on earth ever consider $10k dacs in the first place when there's so many actually reasonably priced alternatives of similar performance?

 

[BDWoody]

Well I just wasted a bunch of time on what seems like an audio salesman's whine. Why on earth ever consider $10k dacs in the first place when there's so many actually reasonably priced alternatives of similar performance?

1. Someone told them after some uncontrolled purely subjective impressions thst it sounded better, and since they don't know enough to realize they are being taken advantage of, they go along with whoever has the best ad copy, and pay whatever they are told is required to be in the club.

2. Hmmm... Can't think of anything else really...other than It gives them instant 'credibility' in the woo club, and allows them to work on their own creative writing as they describe subtle nuances, that only exist in their minds, in an almost uncomfortable sexual breathlessness...

And...that used to be me. (First one...sheesh)

 

[Chrispy]

1. Someone told them after some uncontrolled purely subjective impressions thst it sounded better, and since they don't know enough to realize they are being taken advantage of, they go along with whoever has the best ad copy, and pay whatever they are told is required to be in the club.

2. Hmmm... Can't think of anything else really...other than It gives them instant 'credibility' in the woo club, and allows them to work on their own creative writing as they describe subtle nuances, that only exist in their minds, in an almost uncomfortable sexual breathlessness...

And...that used to be me. (First one...sheesh)

I'm more value oriented in any case rather than a particular implementation method, which too often in the high end doesn't mean much. High end is more a selling and aesthetic environment than anything meaningful in quality/performance. I think many of us here have been "there", too.