Quad HIFI Ideas?

[Mike Stranks]

When I first met my father-in-law-to-be I commented with approval on his Sugden amp and tuner set-up.... after that I was home and dry!

[ef37a]

"cheap generic IC's are used a lot these days."

You are in for a bad time then Arpangel because just about everything you have ever heard, hear now and ever likely to hear was amplified, buffered or processed through many dozens of those "cheap generic ICs" you seem to despise.

The NE5532 was/is the mainstay of the audio electronics industry. The distortion is so low that even the best analysers struggle to resolve it above noise. Their noise level is extremely low (and the LM4562 even better but more expensive) to the point that you will find it hard source a program nearly as good!

The most common FET input chip, the TL072 is not as good if you want to drive low loads (under 5k say) but is otherwise a perfectly acceptable device. Has a higher slew rate than the NEs but that is just tweaky bllx and matters not for audio applications.

The only area where discrete components are better is for noise critical applications like mic pre amps and (very posh!) summing amplifiers but that is just a noise issue, has nothing to do with subjective sound quality and if you can run to a decent input transformer the NE5534 will beat a transistor for noise. (RED pre amp!)

If a 20dB discrete amplifier sounds differently from a 5532 set for 20dB the discrete amp is faulty!

Dave.

[Arpangel]

"cheap generic IC's are used a lot these days."

You are in for a bad time then Arpangel because just about everything you have ever heard, hear now and ever likely to hear was amplified, buffered or processed through many dozens of those "cheap generic ICs" you seem to despise.

The NE5532 was/is the mainstay of the audio electronics industry. The distortion is so low that even the best analysers struggle to resolve it above noise. Their noise level is extremely low (and the LM4562 even better but more expensive) to the point that you will find it hard source a program nearly as good!

The most common FET input chip, the TL072 is not as good if you want to drive low loads (under 5k say) but is otherwise a perfectly acceptable device. Has a higher slew rate than the NEs but that is just tweaky bllx and matters not for audio applications.

The only area where discrete components are better is for noise critical applications like mic pre amps and (very posh!) summing amplifiers but that is just a noise issue, has nothing to do with subjective sound quality and if you can run to a decent input transformer the NE5534 will beat a transistor for noise. (RED pre amp!)

If a 20dB discrete amplifier sounds differently from a 5532 set for 20dB the discrete amp is faulty!

Dave.

Interesting, I have an old Trio KA1500, I can't see any IC's in it at all. It's all, by the looks of it, discreet components and hand wired. Sounds very good though, and it was a budget amp at the time.
There was a time in the 80's lots of BADA shops, all had the same dogged philosophy, and I thought a lot of that Naim, Linn, Rotel stuff sounded pretty bad TBQH. It was sort of all down hill a bit from then.

[ef37a]

"cheap generic IC's are used a lot these days."

You are in for a bad time then Arpangel because just about everything you have ever heard, hear now and ever likely to hear was amplified, buffered or processed through many dozens of those "cheap generic ICs" you seem to despise.

The NE5532 was/is the mainstay of the audio electronics industry. The distortion is so low that even the best analysers struggle to resolve it above noise. Their noise level is extremely low (and the LM4562 even better but more expensive) to the point that you will find it hard source a program nearly as good!

The most common FET input chip, the TL072 is not as good if you want to drive low loads (under 5k say) but is otherwise a perfectly acceptable device. Has a higher slew rate than the NEs but that is just tweaky bllx and matters not for audio applications.

The only area where discrete components are better is for noise critical applications like mic pre amps and (very posh!) summing amplifiers but that is just a noise issue, has nothing to do with subjective sound quality and if you can run to a decent input transformer the NE5534 will beat a transistor for noise. (RED pre amp!)

If a 20dB discrete amplifier sounds differently from a 5532 set for 20dB the discrete amp is faulty!

Dave.

Interesting, I have an old Trio KA1500, I can't see any IC's in it at all. It's all, by the looks of it, discreet components and hand wired. Sounds very good though, and it was a budget amp at the time.
There was a time in the 80's lots of BADA shops, all had the same dogged philosophy, and I thought a lot of that Naim, Linn, Rotel stuff sounded pretty bad TBQH. It was sort of all down hill a bit from then.

The Trio would probably have "sounded" even better for the use of a couple of NEs! (I put sounded in " " because I doubt anyone could hear the difference in practice)
It is for example very difficult to make a really linear Baxandall tone control using the low loop gain of discrete transistors. Accurate RIAA equalisation and adequate headroom for phono preamps is also better done with op amps, the NE5534 for instance is hard to beat for noise. The only area where discretes score is the use of parallel devices in MC front ends.

Naim? Linn sound bad? Depends who you ask. These products were reckoned to be the dogs whatsits back in the day.

Dave.

[Arpangel]

"cheap generic IC's are used a lot these days."

You are in for a bad time then Arpangel because just about everything you have ever heard, hear now and ever likely to hear was amplified, buffered or processed through many dozens of those "cheap generic ICs" you seem to despise.

The NE5532 was/is the mainstay of the audio electronics industry. The distortion is so low that even the best analysers struggle to resolve it above noise. Their noise level is extremely low (and the LM4562 even better but more expensive) to the point that you will find it hard source a program nearly as good!

The most common FET input chip, the TL072 is not as good if you want to drive low loads (under 5k say) but is otherwise a perfectly acceptable device. Has a higher slew rate than the NEs but that is just tweaky bllx and matters not for audio applications.

The only area where discrete components are better is for noise critical applications like mic pre amps and (very posh!) summing amplifiers but that is just a noise issue, has nothing to do with subjective sound quality and if you can run to a decent input transformer the NE5534 will beat a transistor for noise. (RED pre amp!)

If a 20dB discrete amplifier sounds differently from a 5532 set for 20dB the discrete amp is faulty!

Dave.

Interesting, I have an old Trio KA1500, I can't see any IC's in it at all. It's all, by the looks of it, discreet components and hand wired. Sounds very good though, and it was a budget amp at the time.
There was a time in the 80's lots of BADA shops, all had the same dogged philosophy, and I thought a lot of that Naim, Linn, Rotel stuff sounded pretty bad TBQH. It was sort of all down hill a bit from then.

The Trio would probably have "sounded" even better for the use of a couple of NEs! (I put sounded in " " because I doubt anyone could hear the difference in practice)
It is for example very difficult to make a really linear Baxandall tone control using the low loop gain of discrete transistors. Accurate RIAA equalisation and adequate headroom for phono preamps is also better done with op amps, the NE5534 for instance is hard to beat for noise. The only area where discretes score is the use of parallel devices in MC front ends.

Naim? Linn sound bad? Depends who you ask. These products were reckoned to be the dogs whatsits back in the day.

Dave.

Thanks, my Trio is fairly noisy, I must admit, but, Noise hasn't ever been an issue for me, we all have different priorities, but I'd rather have a bit of noise and a good tonal balance than a completely quiet clinical sound.
That's the thing, Linn/Naim had great PR, their reputation was phenomenal, and they were very clever about dealer networking, demo evenings, and how they presented their stuff, but it just wasn't my cup of tea, it was all about the last ounce of detail at the expense of everything else, I much prefer a richer, slightly coloured sound, to the colder and more clinical sound of some of that 80's stuff.
And now prices for those things are quite frankly, stupid, £18,000 for a Linn Sondeck? No record deck is worth that no matter what it's made of or how it's designed.

[ef37a]

"cheap generic IC's are used a lot these days."

You are in for a bad time then Arpangel because just about everything you have ever heard, hear now and ever likely to hear was amplified, buffered or processed through many dozens of those "cheap generic ICs" you seem to despise.

The NE5532 was/is the mainstay of the audio electronics industry. The distortion is so low that even the best analysers struggle to resolve it above noise. Their noise level is extremely low (and the LM4562 even better but more expensive) to the point that you will find it hard source a program nearly as good!

The most common FET input chip, the TL072 is not as good if you want to drive low loads (under 5k say) but is otherwise a perfectly acceptable device. Has a higher slew rate than the NEs but that is just tweaky bllx and matters not for audio applications.

The only area where discrete components are better is for noise critical applications like mic pre amps and (very posh!) summing amplifiers but that is just a noise issue, has nothing to do with subjective sound quality and if you can run to a decent input transformer the NE5534 will beat a transistor for noise. (RED pre amp!)

If a 20dB discrete amplifier sounds differently from a 5532 set for 20dB the discrete amp is faulty!

Dave.

Interesting, I have an old Trio KA1500, I can't see any IC's in it at all. It's all, by the looks of it, discreet components and hand wired. Sounds very good though, and it was a budget amp at the time.
There was a time in the 80's lots of BADA shops, all had the same dogged philosophy, and I thought a lot of that Naim, Linn, Rotel stuff sounded pretty bad TBQH. It was sort of all down hill a bit from then.

The Trio would probably have "sounded" even better for the use of a couple of NEs! (I put sounded in " " because I doubt anyone could hear the difference in practice)
It is for example very difficult to make a really linear Baxandall tone control using the low loop gain of discrete transistors. Accurate RIAA equalisation and adequate headroom for phono preamps is also better done with op amps, the NE5534 for instance is hard to beat for noise. The only area where discretes score is the use of parallel devices in MC front ends.

Naim? Linn sound bad? Depends who you ask. These products were reckoned to be the dogs whatsits back in the day.

Dave.

Thanks, my Trio is fairly noisy, I must admit, but, Noise hasn't ever been an issue for me, we all have different priorities, but I'd rather have a bit of noise and a good tonal balance than a completely quiet clinical sound.
That's the thing, Linn/Naim had great PR, their reputation was phenomenal, and they were very clever about dealer networking, demo evenings, and how they presented their stuff, but it just wasn't my cup of tea, it was all about the last ounce of detail at the expense of everything else, I much prefer a richer, slightly coloured sound, to the colder and more clinical sound of some of that 80's stuff.
And now prices for those things are quite frankly, stupid, £18,000 for a Linn Sondeck? No record deck is worth that no matter what it's made of or how it's designed.

Well of course, you are entitled to listen to what you want but if you KNOW it has limitations you cannot call it "Hi-Fi" !!

Dave.

[Ian Shaw]

Swap you then The Quad was the one when I were a lad (33/303 back then) when I was buying the Sugden was more 'affordable' and considered pretty good. I don't run a 'hifi' any more as I do any listening in the studio.

Do you still have the electrostatics? A mate of my dad had a pair and they were awesome.

I also have a Quad 522 (a 405 in a rack case with BBC mods applied) which was my main PA amp for several years before settling down to studio life. It's awaiting some TLC as I type before returning to the studio.

I don't have electrostatics I'm afraid. They cost him around £1500 in the early 80s & he sold them for more about 10 years later when my parents downsized. I only have a small pair of Missions he bought after that.

[Sam Spoons]

Still a formidable system. Maybe I'll dig out my Sugden again and set it up somewhere with my LS3s (kit built LS5/5As).

[Ian Shaw]

So, early results are in:

https://www.dropbox.com/s/r394jmsiwxaieae/Quad%20Setup.png?dl=0



It works but only in the left channel no matter what input I use

Tuner working fine ( left channel only) and sounds good, but have to laugh at the BBC bias. The BBC1 button is' in anticipation of a fourth BBC channel' according to the manual.

Plugging my phone into the CD input works but mostly bass with little treble

CD player in the same input is silent on 0 or very loud & distorting when turned to 1 on the volume knob

There is something sliding around inside the pre-amp! Shall investigate that first!

[Martin Walker]

Good progress Ian, and at least there were no bangs!

Martin

[John Willett]

So, early results are in:

https://www.dropbox.com/s/r394jmsiwxaieae/Quad%20Setup.png?dl=0



It works but only in the left channel no matter what input I use

Tuner working fine ( left channel only) and sounds good, but have to laugh at the BBC bias. The BBC1 button is' in anticipation of a fourth BBC channel' according to the manual.

Plugging my phone into the CD input works but mostly bass with little treble

CD player in the same input is silent on 0 or very loud & distorting when turned to 1 on the volume knob

There is something sliding around inside the pre-amp! Shall investigate that first!

Ooh - so the 34 *does* have that wonderful exclusive Quad tilt control like the 44.

That is a "magic" system to keep.

Just get it checked out (I still have my 44.405-2).

[Ian Shaw]

https://www.dropbox.com/s/ku1g00ius5acb4s/405%20Insides.png?dl=0

So, I traced the dead outputs to the circuit board on the right & the 2 fuses were clearly blown, so I threw caution to the wind & put in the ones from the good channel...

...switched on & they blew immediately.

That being the limit of my knowledge it's time to get it to a repair shop.

Still, it's been fun getting all set up & avoiding work on this rainy day

Anyone know of a good independent repair centre in the South West? Martin?

Ian

[Hugh Robjohns]

Ooh - so the 34 *does* have that wonderful exclusive Quad tilt control like the 44.

Yes, but it has fewer inputs, no input customisation options, and a much less flexible and adjustable set of vinyl filters.

The first question to answer is whether the lack of a channel is down to the power amp or the preamp.

Unless the OP is experienced in electronic fault finding and repair, I'd strongly recommend getting the whole system to someone who is -- otherwise there's a risk of doing more harm than good.

I can definitely recommend AmpLabs in Worcester (http://www.amp-labs.co.uk/servicing.htm), but there are countless others with the appropriate experience around the UK and elsewhere...

H

[Ian Shaw]

Thanks Hugh, those guys look like they know what they're doing. Useful having the guide prices. Might have to do one piece at a time, starting with the power amp.

Best

Ian

[Hugh Robjohns]

It's actually just one guy, but he is very familiar with the whole classic Quad range and really knows his stuff. I've seen lots of examples of his work from other people's restored/repaired classic hifi, and he did a fabulous job repairing and upgrading a Quad 44/405-1 combo for me. His rates are a fair reflection of the time involved -- there's a lot of testing and disassembly involved -- as well as the quality of his work and any replacement components, I think. I'd definitely use him again.

[Arpangel]

Ooh - so the 34 *does* have that wonderful exclusive Quad tilt control like the 44.

Yes, but it has fewer inputs, no input customisation options, and a much less flexible and adjustable set of vinyl filters.

The first question to answer is whether the lack of a channel is down to the power amp or the preamp.

Unless the OP is experienced in electronic fault finding and repair, I'd strongly recommend getting the whole system to someone who is -- otherwise there's a risk of doing more harm than good.

I can definitely recommend AmpLabs in Worcester (http://www.amp-labs.co.uk/servicing.htm), but there are countless others with the appropriate experience around the UK and elsewhere...

H

I like the filters on my 33, and those plug in cards, but my God, whoever did the soldering on those? it's worse than mine, and that's saying something!
Looking good Ian......like.....PPHHHWWOOOOORRR!!!

[Ian Shaw]

Well, I seem to have turned my house into a HIFI graveyard this weekend. After reconnecting up my Arcam system one of my B&Ws made a farting noise & went dead, so I'm now using Dad's tiny old Missions with that. Strangely though, the display on my Arcam CD player has sprung back into life after being dead for some years. I guess that's what happens with the finely balanced & delicate nature of HIFI seperates

But I have a least decided to go the whole hog & get the Quad refurbed by the guy Hugh mentioned.

It's strangely comforting, staring at the insides of something that's been put together by hand & not just knocked out by the gazillion in a huge factory. My kids can't quite believe the weight of the power amp or why it's even needed to drive some old speakers.

Thanks for all the input.

BTW can someone easily explain 'current dumping'? It was obviously good enough to slap a patent on. From what I can gather it requires a high powered amp which gives whatever is needed to a lower powered amp?

Cheers

Ian

[James Perrett]

BTW can someone easily explain 'current dumping'? It was obviously good enough to slap a patent on. From what I can gather it requires a high powered amp which gives whatever is needed to a lower powered amp?

I've seen it described as feed forward rather than feedback. It uses two amplifiers - a big class B amplifier to provide most of the power and a small class A amplifier. Quad's simple explanation was that the big amplifier provided a signal that was roughly correct while the small accurate class A amplifier provided the final adjustment so that the output was exactly right.

[ef37a]

"BTW can someone easily explain 'current dumping'? It was obviously good enough to slap a patent on. From what I can gather it requires a high powered amp which gives whatever is needed to a lower powered amp?"

Hah! I spent a nifty on a book and although Current Dumping was only a small part of the many things I hoped to learn, there was next to naff all about it!

I do recall at the time the pages of Wireless World (we are not worthy) were rife with claims and counter claims that CD was just a fancy way of using negative feedback.

I probably just don't have the math and will never really understand it.

Dave.

[James Perrett]

Hah! I spent a nifty on a book and although Current Dumping was only a small part of the many things I hoped to learn, there was next to naff all about it!

That's probably down to Doug Self not rating it as a useful design. John Linsley Hood has a few paragraphs on it in his "Valve and Transistor Audio Amplifiers" book but there are also plenty of web resources covering the 405 and its successors.

PS - Amazon is listing a copy of the Linsley Hood book for under a tenner from one of their alternative suppliers at the moment.

[ef37a]

Hah! I spent a nifty on a book and although Current Dumping was only a small part of the many things I hoped to learn, there was next to naff all about it!

That's probably down to Doug Self not rating it as a useful design. John Linsley Hood has a few paragraphs on it in his "Valve and Transistor Audio Amplifiers" book but there are also plenty of web resources covering the 405 and its successors.

PS - Amazon is listing a copy of the Linsley Hood book for under a tenner from one of their alternative suppliers at the moment.

Thanks for that James, daughter can treat her old dad for that Dad's Day!

There is another amp design D S seemed a bit cold one? The "Edwin" amp. There you had the OP Transistors "just" not conducting and the speaker was driven at very low levels by the driver transistors*. As the drive current increase the voltage across the 100Rs in the power trans base circuits increase and caused them to start to conduct. All under feedback control. The design could not achieve distortion levels in 4 decimal places as do the best of Self's offerings but I am not sure we need such low figures, not for any but THE most esoteric purposes anyway?

*The Sinclcair Z50 was an Edwin op stage and was pretty good even if "50 watts" WAS rather optimistic! I actually built two "rips" using what I had available, TIP41/42 iirc and used them in bridge mode to drive a 10 foot column speaker. That system chucked it out I can tell thee! Sounded a hell of a lot better than the 15" disco "bins" you saw in the day, at least for vocal signals.

Dave.

[Hugh Robjohns]

BTW can someone easily explain 'current dumping'? It was obviously good enough to slap a patent on. From what I can gather it requires a high powered amp which gives whatever is needed to a lower powered amp?

There's lots of stuff across the web in varying levels of complexity about the current dumping concept. One of the most succinct is from Richard Brice: http://www.richardbrice.net/curdump.htm

But your description is broadly correct. A highly accurate, low-powered, class-A amplifier feeds into a passive network which ultimately feeds the speaker. And for low level signals, that's all there is...

But when a louder signal comes along, the class-A amp isnt powerful enough, so a simple class-B amplifier kicks in to deliver the appropriate oomph through the network to the speaker -- it 'dumps' the required current into the load.

However, a normal class-B amplifier suffers from gross cross-over distortion; as the signal moves from its positive side of the cycle to the negative side both of the class-B's output transistors switch off for a while and stop providing the required current to the speaker.

A class-AB amplifier cures that problem by biasing those transistors such that they never turn off completely... But that brings drawbacks with power and heat dissipation, as well as issues around controlling the bias current and maintaining stability.

The Quad design sticks with the simple class-B output stage, but cures the cross-over distortion problem by using that class-A stage to 'fill in' the gaps, feeding the corrective signal forward to the speaker through that passive network again.

It is an ingenious, very elegant, and extremely effective design, but Quad's patent prevented widespread adoption of it, so it never really received the praise it deserved.

H

[Ian Shaw]

BTW can someone easily explain 'current dumping'? It was obviously good enough to slap a patent on. From what I can gather it requires a high powered amp which gives whatever is needed to a lower powered amp?

There's lots of stuff across the web in varying levels of complexity about the current dumping concept. One of the most succinct is from Richard Brice: http://www.richardbrice.net/curdump.htm

But your description is broadly correct. A highly accurate, low-powered, class-A amplifier feeds into a passive network which ultimately feeds the speaker. And for low level signals, that's all there is...

But when a louder signal comes along, the class-A amp isnt powerful enough, so a simple class-B amplifier kicks in to deliver the appropriate oomph through the network to the speaker -- it 'dumps' the required current into the load.

However, a normal class-B amplifier suffers from gross cross-over distortion; as the signal moves from its positive side of the cycle to the negative side both of the class-B's output transistors switch off for a while and stop providing the required current to the speaker.

A class-AB amplifier cures that problem by biasing those transistors such that they never turn off completely... But that brings drawbacks with power and heat dissipation, as well as issues around controlling the bias current and maintaining stability.

The Quad design sticks with the simple class-B output stage, but cures the cross-over distortion problem by using that class-A stage to 'fill in' the gaps, feeding the corrective signal forward to the speaker through that passive network again.

It is an ingenious, very elegant, and extremely effective design, but Quad's patent prevented widespread adoption of it, so it never really received the praise it deserved.

H

Thanks Hugh, that'll do nicely

[ef37a]

Yes thanks Hugh I do now have a better grasp.

I would venture to say that the output stage is not class "B" as it is normally considered? The transistors are at zero bias current (class C? ) whereas conventional class B, both valve and transistor, always has a small current flowing in the output stage and the complications of this were what Quad were trying to eliminate. Many would say they succeeded?

Self is not keen perhaps because the distortion figures for the 405 look decided pedestrian compared to his own designs which typically achieve an order better. 0.001% up to 1kHz and still .005% at 10kHz.

The idea of complimentary transistor at zero bias in the loop of a high gain op amp has been used in countless cheap mixers and other circuits as a headphone amplifier. Never read a complaint in a review!

Dave.

[Forsty]

I found this thread while I was wondering if I should get a Quad system. I recently bought a friends pair of B&W CM9 and Im in need of an amp. I really like the design and simplicity of the Quad amps (Elite Pre/QSP, Artera Play/Stereo). Plus it seems like the prices are dropping the last weeks, for example I found an Elite QSP for under 700€ (brand new).

But it's really hard to find reviews about these modern systems, I found only 2-3 reviews which all were quite positive for the price class. Is Quad really so unpopular nowadays? If so, why exactly?