What HiFi humiliation and QUAD discussion

[davebmw]

Still, that was to point out that transmission at typical DC voltages would require a rather significant conductor indeed.

Please don't try and tell me you honestly thought we didn't get that?!?!

I would NEVER use a silver cable myself, people that believe in skin factor at audio frequencies should be taken out and shot IMO

Your'e telling me! I had a huge standing and very public argument with the foremost What Hi Fi columnist at one of the UK's biggest Hi Fi expo's, he who shall remain nameless because he knows he talks out of his backside.
I caught him spouting about the quality of cables feeding a set of Quad ESL 63's were far better than the ESL's could ever hope to produce in audio terms, he said that the sound came solely from the cables and the speaker meant nothing.
I took great pride in telling him and I quote "Quite frankly sir, you really have spent too much time listening to the outright waffle that comes out of your front door, I'm just glad that the discerning British public knows this and laugh at your.....writings, had you had any technical knowledge whatsoever, you would know that as long as the cable has sufficient current carrying capacity and that the power amplifier has control over the speaker, the sound reproduction will be as faithful as the combined efforts of the sum of the Hi Fi's parts.
At which point the crowd applauded.

I felt brilliant and not bad for a 15 year old!!! my old man was chuffed!

[Fred]

Awesome work Dave :-) I'd have given my left nut to see that.

The issue comes when you try to define "sufficient current capacity" with the inductive and capacitive nature of multiway speakers. The actual current demands can be fairly significant at peak. Plus, as you go larger the capacitance of the cable increases too...

The main thing is to keep the leads as short as possible.

For the amplifier such speakers are even worse. Due to phase shifts the current can often peak when voltage is close to zero, net result is the transistors doing the hardest work where they least like it.

Reality is :

Most important : Speakers
Second/Third : Source and Amp in no particular order
distant fourth unless you are living next to a cell transmitter : cables

[davebmw]

That's all well and good but this particular setup was a set of QUAD ESL-63 electrostatics driven by Quad 606 current dumping amplifier, controlled by a 66 control unit and the brand new at the time 67 CD player.
The ESL-63 takes very little driving to produce enormous SPL. there is not any crossover distortion because there are no crossovers there can be no phasing issues as there are no multiway speakers. it basically a flat sheet of mylar moving back and forth.

Besides that the 606 is an amazing power amp design with outstanding control over anything you decide to attach to it.
the feed forward and feedback design actively damps the speaker much like a servo. these still outperform most of todays mosfet amps.
I have a 606 and a 405mk2 and i would give my left nut to keep them as they are the only thing that is capable of controlling my Kef cantata's which have B139 driver units.

[Fred]

You seem to have misunderstood my comment on phase. Electrostatic is almost worst case for phase load. It has impedance down to 2 Ohms, and the amplifier has a schematic from the dark ages http://quad405.com/606schematic.pdf

What's your extension number at what hifi ? :-p ;-)

I'm sure it was good at the time, but it would appear to be a class A-B design which is the worst of both worlds.

I prefer a BJT design for under 150rms / channel, but to say that amp kicks modern designs is ummm misleading at best.

[davebmw]

Ah look closer, it has both the classic class A stage and a class B. At lower power the Class A (very high quality amplifier) has total control. when the going gets a bit tougher, its class B colleague gets involved and provides the grunt the rest of the way.
The combination of the 2 is usually called AB, but the QUAD is not. it is a class all of its own, a Current dumper.
There have been many attempts to copy the design none of which have really improved on the original 405 design. when the 606, 707 and 909 were released they just used more output transistors to provide more grunt.

BTW anyone who says that a QUAD Current dumper can't hold its own in this day and age hasn't actually heard one.
If you ever have a chance to sit down and grill one for a couple of hours take the opportunity and a huge variety of your music and you will be surprised.
The oddly designed quite frankly ugly chunks of equipment somehow have that little something.

[Fred]

Ah look closer, it has both the classic class A stage and a class B.

It has only one connection to the output transistors, and thus has a switch point from A to B, and therefore has switching distortion (perhaps inaudible) in that region.

At lower power the Class A (very high quality amplifier) has total control. when the going gets a bit tougher, its class B colleague gets involved and provides the grunt the rest of the way.

The transition is the "problem". Perhaps the "problem" is not a problem at all. Still, A-B or whatever you choose to call it till you define the difference more carefully has no advantage over a good B and only a power output advantage over a pure A with the consequence of worse distortion performance (assuming the A is well designed as they often aren't)

The combination of the 2 is usually called AB, but the QUAD is not. it is a class all of its own, a Current dumper.

Define the difference explicitly please or we may as well get you a plaque for the door of your what hifi office ;-)

BTW anyone who says that a QUAD Current dumper can't hold its own in this day and age hasn't actually heard one.

Here's the thing, that is not required. I'm not arguing that it sounds bad. I'm not arguing that it doesn't sound good enough. I'm not even arguing that it doesn't sound great. I'm not saying that there aren't bad amps today. I am saying that on a scope and spectrum analyser under load where it counts you will either not pick the difference or the quad will lose to a superior up to date design. It is feasible that it was below the 0.1% ear detectable range way back then, but not really that it's at the 0.001% range achievable now ;-)

Fred.

(I'm going to split this topic out.)

[davebmw]

It has only one connection to the output transistors, and thus has a switch point from A to B, and therefore has switching distortion (perhaps inaudible) in that region.

Christ knows what circuit diagram you are looking at!

The transition is the "problem". Perhaps the "problem" is not a problem at all. Still, A-B or whatever you choose to call it till you define the difference more carefully has no advantage over a good B and only a power output advantage over a pure A with the consequence of worse distortion performance (assuming the A is well designed as they often aren't)

OK class B have crossover distortion which is most noticeable in the low power end. Class A have excellent distortion characteristics in the low end but bugger all power and are very inefficient. Quad's design is a composite of the 2 and it really works, there obviously is no convincing the un-con-vincible so I will leave that there.

The combination of the 2 is usually called AB, but the QUAD is not. it is a class all of its own, a Current dumper.

Define the difference explicitly please or we may as well get you a plaque for the door of your what hifi office

I have no need to, the patent office took some convincing but granted many patents on the design as it was proven technology.

BTW anyone who says that a QUAD Current dumper can't hold its own in this day and age hasn't actually heard one.

Here's the thing, that is not required. I'm not arguing that it sounds bad. I'm not arguing that it doesn't sound good enough. I'm not even arguing that it doesn't sound great. I'm not saying that there aren't bad amps today. I am saying that on a scope and spectrum analyser under load where it counts you will either not pick the difference or the quad will lose to a superior up to date design. It is feasible that it was below the 0.1% ear detectable range way back then, but not really that it's at the 0.001% range achievable now

OK I'm not quite sure how to put this without sounding like a complete cunt! so here goes!
What I am trying to get at is the quad is a unique design using the best characteristics of 2 proven designs.

It sounds great, its durable will drive virtually anything and lasts decades my dad has 2 405's that are 28 years old and still kick butt!

And no in your last comment you are wrong, what counts is the end experience, what counts is the music sounds great, that's what you expect, not some distorted pile of shite.
The human hearing is most sensitive at lower SPL not when the Amp is at full tilt.

And are you trying to say that over the last 30 years human evolution has moved that fast that our ears have evolved into a super high definition realm of audio perception. I think not, that really is in the realm of What HiFi BS. The average Joe can't even tell one frequency band from the next, let alone fractions of a percent of THD.

Fred my man we are destined to dis-agree for many a year to come on this topic, all I can do is invite you to my home to have a listen to my QUAD 606 driving my Kef Cantata prototypes and see what you honestly think.
you'll have to wait a couple of months though because we are about to move house and would love to end our hate filled relationship with our neighbors with the loudest send-off possible.

Dave

[Fred]

Christ knows what circuit diagram you are looking at!

The one I linked Dave :-) If I'm reading it incorrectly, please help me out! :-)

OK class B have crossover distortion which is most noticeable in the low power end.

Let me fix that...

OK class B have crossover distortion which is most significant in the low power end but not audible anywhere on a good design.

Fixed :-)

Class A have excellent distortion characteristics in the low end but bugger all power and are very inefficient.

True, they can also have other distortion mechanisms due to the OS load. The paralleled OS BJTs in that design probably prevent such modes of distortion though.

Quad's design is a composite of the 2 and it really works, there obviously is no convincing the un-con-vincible so I will leave that there.

I'm not saying it doesn't work. I'm saying that if you scoped it etc then you would prove that it doesn't work as well as a good A or B design. OR it isn't A-B in which case I want to learn about it! I'm convince-able, but not with subjective claims that we were taking the piss out of when we started this thread.

There is no aspect of HiFi that isn't measurable with a tool. It is purely science and nothing else.

The combination of the 2 is usually called AB, but the QUAD is not. it is a class all of its own, a Current dumper.

Define the difference explicitly please or we may as well get you a plaque for the door of your what hifi office ;-)

I have no need to, the patent office took some convincing but granted many patents on the design as it was proven technology.

Great, then link the patent. While you are at it, show me where I went wrong reading the schematic of the device in question.

What I am trying to get at is the quad is a unique design using the best characteristics of 2 proven designs.

And what I am trying to get at is that until you show me how it's not A-B, it is the case that by merging the designs you end up with something that has the worst characteristics of both designs. If/when you show me that it is in fact not A-B then we will talk further on the matter :-)

Randy Slone on A-B :

CLASS AB Class AB is not really a class but rather a poor marriage of both class A and class B characteristics.

Since class AB operation provides no advantages whatsoever and only serves to degrade linearity and create additional heat problems, it should be entirely dismissed as a "good idea that just didn't work out."

Of course, if it isn't A-B then that doesn't apply, but if it isn't you need to show us what it is ;-) No subjectivism please, just objective scientific discussion!

It sounds great, its durable will drive virtually anything and lasts decades my dad has 2 405's that are 28 years old and still kick butt!

I deny none of that at all :-)

And no in your last comment you are wrong, what counts is the end experience, what counts is the music sounds great, that's what you expect, not some distorted pile of [shizzle].

Exactly where did I say that that isn't the case??? Are we discussing whether you are happy with it? Are we discussing whether the distortion it outputs (ALL amps output distortion) is audible? Are we discussing whether or not it is as good as it possibly could be/as good as modern pure B designs? I'm certainly not trying to argue the first two of those :-)

The human hearing is most sensitive at lower SPL not when the Amp is at full tilt.

Agreed, to quote myself : "Perhaps the "problem" is not a problem at all."

And are you trying to say that over the last 30 years human evolution has moved that fast that our ears have evolved into a super high definition realm of audio perception.

No, where did I say that? I was saying that over the last 30 years audio circuit design has changed significantly enough to render the performance of the quad less than the best. Note, not less than good enough, just less than the best. In 1975 this kiwi ran the fastest mile in the world :

http://en.wikipedia.org/wiki/John_Walker_(runner)

Today, people run faster, he is still very fast, and more than fast enough, but he is no longer the fastest ;-)

I think not, that really is in the realm of What HiFi BS.

I am the last person you will get whathifibs from! :-p

The average Joe can't even tell one frequency band from the next

Are you joking? That must be a joke.

Fred my man we are destined to dis-agree for many a year to come on this topic

Based on what you just posted, I'm not sure we disagree all that much at all. Only on :

1) whether it is A-B or not, and I'm not saying it is A-B, only that I'll continue to believe it is until you prove otherwise. I'll base that belief on the schematic posted above and my understanding of audio circuitry :-)
2) whether or not it is the best in 2008 or not

I look forward to enjoying it :-) In NZ I was using some KEF uni-q speakers and though a little light in the bottom end (typical for properly designed bookshelves without a big base resonance) they were supurb :-) I still prefer a B&W though.

Fred.

[Fred]

No need to explain what it is, I've done my own homework out of interest.

It is in fact a Class C output stage. It is in NO way class B and the class A section does NOT use the output devices AT ALL but DOES drive the load at lower levels, and controls the load (as an ordinary input VA stage pair would) through the output devices the rest of the time.

A clever design and I'm sure it works very well, but at the end of the day, DEFINITELY no better than any good Class B amp design. Reason being that the switch on/switch off is still occurring and feedback is still being used to nullify the effects of that switching. Exactly like a B amp in that respect, and hence no better. Just different :-) I have nothing against B, I think A is highly over rated and thus I have nothing against your current dumpers either :-)

References :

http://ourworld.compuserve.com/homepage ... urdump.htm

http://www.st-andrews.ac.uk/~www_pa/Sco ... page3.html

http://www.quadesl.org/Album/Interviews ... sW1975.doc

[Fred]

I'll take back one thing :

In a practical sense or even percentage sense it may actually be better. Still as discussed better when already way below audible levels is pretty much irrelevant.

Because you are right about the human ear being more sensitive to distortion at lower levels, it is probably true that if the percentage distortion was equal across all volume levels then it would be most audible when in the crossover region rather than somewhere else. That entirely depends on the levels it is reaching when the dump occurs.

Also, because the error from switching is most likely fairly linear regardless of what sort of level it is at, the percentage error will be lower the higher you go.

I'd like to see your amps distortion level plot as frequencies got up around the 20 - 40k region.

Fred.