QUAD 521F Power Amp

[RustyB]

That's disappointing. Pretty much the only major annoyance with the UK built Audiolab gear was broken sockets, you'd think they'd have sorted that by now.

One of my pet peeves is expensive, usually European built electronics that have PCB-mount socket arrays that don't provide a solid connection, are too close together for easy connection/disconnection, and fall apart.

 

[JohnW]

Does stuff 115-120db+ down really matter? I've got a Rob serviced 306 that is deathly silent even into my 105db efficient Klipsch La Scalas. If it's fine in that context noise can't possibly be an issue in normal usage with typical box speakers etc. Most amps sound like white-noise generators by comparison!

Over the years I've found that reducing any spurie product no matter how low down they are always impacts the sound quality. You don't normally directly hear these products as tones or Hiss unless really bad, but once they are solved the signature of the sound changes, Bass tightness, Glare / Brightness, Sound stage are all effected...

As a designer with today's measurement tools I would consider such artefacts as a substandard design... I'm sure if the designers at Quad had access to FFT analysers and the power of todays computer simulators, they would have designed out these issues.

I'd have much admiration to what has been achieved in the past with such limited design tools... Pure brain power over measuring then tweaking...

Also any visible crossover distortion spurie ALWAYS adds a hardness to the sound quality (which I associate with the signature of "transistor sound").

 

[JohnW]

The little 240 is acquitting itself rather well driving my JBL H9800Be horns -somewhat ruthlessly revealing transducers-if it was staying I'd investigate some of the potential improvements.
Corrosion aside(not a problem here btw)he nice thing about Neutriks be they XLR or Speakon is they tend not to fall apart when you try to unplug them-I've had to superglue all the rca phono's on my MDAC.

Yep - I hear you, I'll use those chunky RCA's (as used for the MDAC's audio output) for the SPDIF inputs on the MDAC2 design.

I'd be VERY interested in buying the unit... you can consider it already "Sold" when your ready to pass it on.

 

[ian r]

Cooky thanks for the superglue idea

 

[JohnW]

This isn't a dig at John since he's a designer looking to develop his own products and hey, if he can bash distortion down to 0.00001% then I'm the last person to moan and very much welcome him sharing his ideas! For a different design and a different design ethos to Quad that's all very welcome.

Thank you for understanding my intentions - its not to criticise the design but to gain further knowledge and experience with hardware - over the years I've return to look at current dumping with the "Plan" to realise a design when time permits... Playing with the 240 was just for fun, like a detective trying to solve its design issues and to confirm some of my earlier simulation work.

From as early as I can remember I pulled EVERYTHING apart to understand how they worked - I can recall "fixing" my first radio at 4 years old - only a bad battery connector, but still I was very proud - I never seemed to put anything back together again though...

I've always had huge admiration for Peter Walkers ESL's and the Current Dumping topology. While I don't agree with the Quad ethos, to me all “competently” designed Amplifier don't sound the same - I do feel its such a shame that the Current dumping topology has not been further developed - the 405 / 606 circuits while "functional" and adequate - are just that... adequate, but as you say we have a different design ethos but that’s

I'd like nothing more then to publish the designs with my circuit ideas - but I cannot bring myself knowing that if the engineers at IAG China discovered the circuits they would just copy and implement these ideas as they badly need a revamp of the Quad current dumping designs.

I’ve become allergic to the constant “theft” of intellectual property by the Chinese – only then to “proudly” proclaim everything has been developed by themsleves… (China's High speed trains and “there” space program as an example) – even Chinas new jet liner is manufactured on MD tooling that China refused to return (Russia helped design a new super critical wing, but the body is a carbon copy of the DC9/MD80 family).

 

[Arkless Electronics]

Some excellent posts there John! Largely agree.
We are indeed spoilt by the ease of CAD using SPICE etc but as you will be only TOO aware it speaks with forked tongue and it all HAS to be done for real.... Then we can do the iterative process, comparing measured with simulated results and tempering both with the results we get from those wiggly bits of skin on the side of the head

Total respect to the likes of Walker, Baxandall etc who had to do it the hard way... (as I did until about 10 years ago and I'm sure that applies to most designers. Multisim mainly here cos I love the GUI but Microcap and LT Spice also).

Current dumping and its ilk (Sandman, Technics etc) truly was the last of the genius designs and I'm sure it has a long way to go yet

TonyL. Yes I'm starting to post again... missed it! I'll be good

 

[RustyB]

In a nutshell, Quad had some clever ideas and produced some ingeniously well thought out and nicely built casework.

But: their penny-pinching attitude and philosophical blind-spots hamstrung the amps down to a very mediocre level of performance, one easily bettered even by competent Japanese integrateds, eg those from Harman Kardon.

Closest approach to the original sound? What about the music??

 

[Tony L]

In a nutshell, Quad had some clever ideas and produced some ingeniously well thought out and nicely built casework.

But: their penny-pinching attitude and philosophical blind-spots hamstrung the amps down to a very mediocre level of performance, one easily bettered even by competent Japanese integrateds, eg those from Harman Kardon.

I don't think that's fair at all - the 22/II was a pioneering tube amp, the preamp defining the concepts of tone controls, filters etc, the power amp still a sought-after reference today. The 33/303 a landmark of early transistor design that offered unprecedented stability and reliability, sold in ridiculous quantities and stayed in production for well over a decade. The 405 arguably the last real innovation in amplifier design prior to current chip-amps etc. The 34/306/606 etc being worthwhile refinements of Quad technology and thinking and still very decent sounding amps by any standard even decades after they were last built. The most innovative UK audio company by a huge margin IMO. It's great kit - I can think of hardly any modern kit that would stand comparison with a 34/306/ESL63 system. In fact I can think of little that could compete with a 22/II/ESL57 rig, and that's rather old!

 

[Robert]

In a nutshell, Quad had some clever ideas and produced some ingeniously well thought out and nicely built casework.

But: their penny-pinching attitude and philosophical blind-spots hamstrung the amps down to a very mediocre level of performance, one easily bettered even by competent Japanese integrateds, eg those from Harman Kardon.

Closest approach to the original sound? What about the music??

I think that's unfair TBH. It wan't penny pinching, more a case of using components they deemed fit for purpose and sonically fine. The number of still working Quad amplifiers still in circulation attests to this IMO.

The philosophical blind-spots can easily be taken as practical, tested realism.
They were one of the few manufacturers willing to subject their amplifiers to independent controlled listening tests, which they passed.

Don't be so dismissive of Japanese amplifiers They made some of the best IME.

The sound is the carrier for the music. If the sound is right the music is right surely?
I would say that I encounter many people not interested in the closest approach, i.e. they find neutrality in products not to their taste and prefer some spice. That's ok - Quad even gave them the ambler and other filters

 

[Robert]

I've always had huge admiration for Peter Walkers ESL's and the Current Dumping topology. While I don't agree with the Quad ethos, to me all “competently” designed Amplifier don't all sound the same - I feel its such a shame that the Current dumping topology has not been further developed - the 405 / 606 circuits while "functional" and adequate - are just that... adequate, as you say we have a different design ethos but that’s

I'd like nothing more then to publish the designs of the circuits - but I cannot bring myself knowing that if the engineers at IAG China discovered the circuits they would just copy and implement these ideas as they badly need a revamp of the Quad current dumping designs.

For me the beauty of the feed-forward design is the ability to null distortion, rather than simply suppress it using NFB. It's certainly true you can finesse the circuit and get some stunning results. With regard to the Chinese and IAG I think merely discussing the areas that interested DIYers might investigate is great - and it does't hand specifics to 'the enemy'

On a practical level in 2014 as opposed to 1975 I'm not so sure feedforward and current dumping has legs for new adaptation. It's a perfectly viable set of technologies but less of a benefit today IMO.

The ESL technology however certainly has legs. We're looking a much smaller 63 style mid/hf driver but its behind a few other products in the line.
It's the sort of technology which lends itself to very small scale work - much easier to make a working ESL driver than a conventional MC

Anyway, onwards and upwards, and a happy new year!

 

[RustyB]

The 405 arguably the last real innovation in amplifier design prior to current chip-amps etc.

I'd rather listen to an amp built around a carefully considered and properly implemented application of tried and tested circuitry that provides an engaging listen than an innovative design that reduces music to Muzak.

Which is what a stock 44/405 does.

Can't comment on the 33/303.

 

[RustyB]

I think that's unfair TBH. It wan't penny pinching, more a case of using components they deemed fit for purpose and sonically fine. The number of still working Quad amplifiers still in circulation attests to this IMO.

The problem is the components were not "sonically fine", as anyone who's upgraded a 405 knows only too well.

The philosophical blind-spots can easily be taken as practical, tested realism.

Utter bullshit, more like.

 

[JohnW]

The 405 arguably the last real innovation in amplifier design prior to current chip-amps etc.

I'd rather listen to an amp built around a carefully considered and properly implemented application of tried and tested circuitry that provides an engaging listen than an innovative design that reduces music to Muzak.

Which is what a stock 44/405 does.

Can't comment on the 33/303.

The problem with Quad was that they had great ideas, but as you said and I also have to agree with you they penny pinched in certain areas (although there products are mechanically well built – not to Japanese standards, but by British standards very good IMO) - or maybe there ethos was the problem - designing with "technology" and not prepared to really listen to the designs (PW was very much against the idea to entertain the possibility that competently designed amplifiers can sound different - when a few moments of listening will clearly highlight differences).

The Current Damping Bridge is very clever, but the 405 / 606 designs failed to show its full potential - IME due to the design ethos at Quad (Component quality etc).

Likewise the ESL63 was bettered by certain Martin Logan designs and others... Quad got itself into a rutt and kept spinning new designs based on repackaged earlier circuits (and ESL panels etc).

The 606, 306, 707, 909, and now QSX etc... If you look at todays QSX then your "basically" looking at the 606 design ... likewise with the ESL63 and all later "variants are just a repackage of the ESL63... While time has moved on...

Its such a shame to see the downfall of Quad like this...

 

[JohnW]

Some excellent posts there John! Largely agree.
We are indeed spoilt by the ease of CAD using SPICE etc but as you will be only TOO aware it speaks with forked tongue and it all HAS to be done for real.... Then we can do the iterative process, comparing measured with simulated results and tempering both with the results we get from those wiggly bits of skin on the side of the head

I hear you I'd certainly like to believe I'm a better designer due to the power of simulators - now when I have a crazy idea in the bath or bed, I can jump up to the simulator and and workout the design.. make crazy ideas real... or not as the case maybe

 

[Tony L]

The problem is the components were not "sonically fine", as anyone who's upgraded a 405 knows only too well.

You need to place things in their correct historical context, e.g. remember a Quad 303 is a late 60s design and the 405 dates from the mid 70s. The quality of components are fine, e.g. the 303 used the biggest caps you could physically get in the (large) case along with a massive frame transformer, the 44 & 34 were using high-quality bespoke stepped attenuators etc long before tweaky trendy hair-shirt / boutique audio companies jumped on that particular bandwagon.

As for "sonically fine" I'm prepared to bet that a huge amount of your favourite music was created via a 303 or 405 - these were the staple amps in UK recording studios from the late 60s right through to the mid-80s and beyond. Just about every studio I've ever set foot in had it's Tannoy or JBL main monitors driven with a Quad amp. They are sonically fine - a very large percentage of what you put in at one end comes out the other, and that's exactly what you need in an environment as critical as a recording studio.

Can they be improved upon with current technology? Maybe so, but that goes for just about anything doesn't it?! A market has certainly built up around doing just that - Quad kit was so well built and reliable much is still around today so little companies like Net Audio, Dada Electronics etc can make a living upgrading and tweaking them for those that fancy it.

 

[ian r]

Until I had my esl57's 'rebuilt ( due to an accident) to modern standards' I would have always tended towards the notion of modern boutique components being manifestly a 'good thing' .... sonically the end product of modernised 57's is a huge disappointment let alone the 1200£. The complete absence of magic is blindingly obvious even after a 2 month run in. TBH a very slight improvements in treble imaging etc are manifesting ... with a slight diminution in their (imo modern) in your face delivery

Im sorry to drone on about that again but appropos of this Quad amp thread I an a doubting Thomas now and wonder if some sort of Quad amp bake off can occur so we could listen to the modern tweakery available.

I VERY much admire the enthusiasm of our electrical engineers boldly going and am excited to hear of the thinking and 'common sense' but the baby has been thrown out with the bathwater in regard to 'restored' quads as far as I have experienced it.

PJW may have been too geeky for his own firms good but he clearly had wonderful ears as regard the 57's

Martin Logans better - not heard that before

 

[Arkless Electronics]

I think that current dumping and feed forward certainly "has legs". Back when it came out it certainly helped enormously to ameliorate the inadequacies of the available output transistors (not to mention the lack of output mosfets etc) but today, with much better devices available, it can shine brighter than before IMHO.
There are various DIY options already available.... for a cheap and easy one google "death of gainclone".
The 303 was also quite innovative in its day, also mainly in getting around output device limitations, this time by the use of triples.
A 405 with just a few mods can sound VERY good IMHO and will in fact show a clean pair of heels to many a £1K plus amp
The 303 far less so. It tends to loose out in ultimate transparency etc.

I always found Quads pre amps rather disappointing. The 22 and 33 are pretty awful by modern standards and the likes of the 44 used too much modern technology before it was evolved enough to do the job (analog switch IC's and inadequate op amps of the time).

To my shame I well recall, as a spotty 16 year old, "lecturing" Peter Walker as to what was wrong with the 44! He took it in good humour.....

 

[Robert]

You need to place things in their correct historical context, e.g. remember a Quad 303 is a late 60s design and the 405 dates from the mid 70s. The quality of components are fine, e.g. the 303 used the biggest caps you could physically get in the (large) case along with a massive frame transformer, the 44 & 34 were using high-quality bespoke stepped attenuators etc long before tweaky trendy hair-shirt / boutique audio companies jumped on that particular bandwagon.

As for "sonically fine" I'm prepared to bet that a huge amount of your favourite music was created via a 303 or 405 - these were the staple amps in UK recording studios from the late 60s right through to the mid-80s and beyond. Just about every studio I've ever set foot in had it's Tannoy or JBL main monitors driven with a Quad amp. They are sonically fine - a very large percentage of what you put in at one end comes out the other, and that's exactly what you need in an environment as critical as a recording studio.

Can they be improved upon with current technology? Maybe so, but that goes for just about anything doesn't it?! A market has certainly built up around doing just that - Quad kit was so well built and reliable much is still around today so little companies like Net Audio, Dada Electronics etc can make a living upgrading and tweaking them for those that fancy it.

I would agree, and there is often a world of difference between what shows up on a graph and what's audible. Quad are different to almost all audio companies today in properly researching what was actually audible, then ensuring they built to satisfy that spec with a little headroom.

There weren't an audiophile company in the sense we understand the term today, and Walker views toward the audiophiles of his day are well documented. He couldn't understand the fixation on minutia being assessed without controls or references and he was right IMO.

The proof of the pudding is to connect a standard 405 to a pair of 63s and see if you get 'muzak'. In 2013 I've had three visitors poping over specifically to hear ESLs. Two have bought 63s (within weeks) and one intends to. Ok, its not a standard 405-2 but it only has slightly larger PSU caps and a quieter input. The one memorable comment from all is that the 63s don't sound as they expected, specifically they do have low bass, do go loud, aren't soft and do have dynamics.

Less talk and more demonstration is always the best policy.

Im sorry to drone on about that again but appropos of this Quad amp thread I an a doubting Thomas now and wonder if some sort of Quad amp bake off can occur so we could listen to the modern tweakery available.

Yes, that's the only way, with some controls in place.

I always found Quads pre amps rather disappointing. The 22 and 33 are pretty awful by modern standards and the likes of the 44 used too much modern technology before it was evolved enough to do the job (analog switch IC's and inadequate op amps of the time).

To my shame I well recall, as a spotty 16 year old, "lecturing" Peter Walker as to what was wrong with the 44! He took it in good humour.....

You can make an analogue switching IC or ancient jfet op amp look crap or wonderful, depending how they are used. Distortion from both is very sensitive to load impedance. The 44 and 34 is designed with this in mind and both types of device are used optimally.
Both pre amps use many such devices in series yet you'll typically see 0.008% THD from these inadequate devices. The 66 pre amp continued to use them into the 90s, precisely because they work well enough. I've pulled and replaced every switch and op amp on a 34 with technically better modern versions, with appropriate attention to local decoupling to ensure they all work correctly. Doesn't improve the sound. The phono stage can benefit from something al little better than was available in 1980 so that's worth fitting but don't go mad - OPA134 is ample.

 

[Arkless Electronics]

I would agree, and there is often a world of difference between what shows up on a graph and what's audible. Quad are different to almost all audio companies today in properly researching what was actually audible, then ensuring they built to satisfy that spec with a little headroom.

There weren't an audiophile company in the sense we understand the term today, and Walker views toward the audiophiles of his day are well documented. He couldn't understand the fixation on minutia being assessed without controls or references and he was right IMO.

The proof of the pudding is to connect a standard 405 to a pair of 63s and see if you get 'muzak'. In 2013 I've had three visitors poping over specifically to hear ESLs. Two have bought 63s (within weeks) and one intends to. Ok, its not a standard 405-2 but it only has slightly larger PSU caps and a quieter input. The one memorable comment from all is that the 63s don't sound as they expected, specifically they do have low bass, do go loud, aren't soft and do have dynamics.

Less talk and more demonstration is always the best policy.



Yes, that's the only way, with some controls in place.



You can make an analogue switching IC or ancient jfet op amp look crap or wonderful, depending how they are used. Distortion from both is very sensitive to load impedance. The 44 and 34 is designed with this in mind and both types of device are used optimally.
Both pre amps use many such devices in series yet you'll typically see 0.008% THD from these inadequate devices. The 66 pre amp continued to use them into the 90s, precisely because they work well enough. I've pulled and replaced every switch and op amp on a 34 with technically better modern versions, with appropriate attention to local decoupling to ensure they all work correctly. Doesn't improve the sound. The phono stage can benefit from something al little better than was available in 1980 so that's worth fitting but don't go mad - OPA134 is ample.

Whilst I agree that the impedances are crucial in this type of circuitry I'm afraid I disagree with the rest.... Much better performance can be derived from more conventional circuitry IMO.

 

[Tony L]

I always found Quads pre amps rather disappointing. The 22 and 33 are pretty awful by modern standards and the likes of the 44 used too much modern technology before it was evolved enough to do the job (analog switch IC's and inadequate op amps of the time).

I've never used a 22 and I'd agree with you regarding the 33, it's just a preamp too far away from it's context now needing all kinds of gain reduction etc to even make it usable.

The 34 has however really surprised me. After having mine fully serviced by Rob, adjusting the gain a little to make the volume control usable with efficient speakers etc, I can barely tell it apart from my high-end Audio Synthesis passive attenuator. It's maybe just a tiny bit fuller sounding and less three-dimensional, but that may well be poor volume matching (comparing two stepped attenuators, it just isn't going to happen!) - certainly no difference I care about in any way. The phono stage surprised me even more once it was setup properly for my 2M Black (i.e. getting the load and capacitance absolutely bang-on), IIRC Rob also reduced the rumble filter a little too. Put it this way: I no longer own a £1300 EAR 834 as I far prefer the 34, and I *really* wasn't expecting that!