"Ey up chuck!"

[ef37a]

Fascinating article about the KEF monitors but, sorry Phil, the Quad 405 amplifiers were not a "class AB" design. They were "current dumpers" and the one of the main aims of the design was to eliminate the high standing current of the then state of art amplifiers. The other design goal was to remove all presets from the circuitry especially for OP stage current adjustment.

Of course, if you are running an amplifier at 200-300 watts continuously it will get hot, can't beat The Laws of TD but at least Cdumpers get some respite when the music stops!

I would also assume the amplifiers used '50Hz' supplies? Those transformers would have run pretty hot as well but I guess at the time SMPSUs were not a reliable option?

Dave.

[Hugh Robjohns]

Fascinating article about the KEF monitors but, sorry Phil, the Quad 405 amplifiers were not a "class AB" design. They were "current dumpers"...

The 'current dumper' design essentially uses a class-A amplifier for low level signals and error-correction, with a class-B stage to chuck in wallops of current for the louder bits. So no, it's not a 'class-AB' in the classic sense... but I'd argue that it could fairly be described as a 'class-A/B' -- the oblique being significant here -- which is what he did!

The real point, of course, being that these amps weren't modern Class-Ds, so they weren't very power efficient and therefore generated a lot of heat -- which it KM1 certainly did!

I would also assume the amplifiers used '50Hz' supplies? Those transformers would have run pretty hot as well but I guess at the time SMPSUs were not a reliable option?

Yes, they were linear supplies. You can just see the PCB across the top of the transformers, and the reservoir caps in some of those pics, sitting below the vertical amp cards in the base of the chassis.

[James Perrett]

I would also assume the amplifiers used '50Hz' supplies? Those transformers would have run pretty hot as well but I guess at the time SMPSUs were not a reliable option?

Yes, they were linear supplies. You can just see the PCB across the top of the transformers, and the reservoir caps in some of those pics, sitting below the vertical amp cards in the base of the chassis.

From the pictures it looks like they used the same encapsulated transformer as was used on the standard 405. I would guess that they must have used four of them to power the 8 amplifier boards.

[N i g e l]

yeah the Quad 405s are called current dumpers, the schematics are on line.
I always wondered what "current dumper" meant; my present understanding is that the output is in two parts: a large coarse drive to do the donkey work and a smaller drive with more finesse, to correct for the errors in the coarse driver ? Is that right ??

[ef37a]

Fascinating article about the KEF monitors but, sorry Phil, the Quad 405 amplifiers were not a "class AB" design. They were "current dumpers"...

The 'current dumper' design essentially uses a class-A amplifier for low level signals and error-correction, with a class-B stage to chuck in wallops of current for the louder bits. So no, it's not a 'class-AB' in the classic sense... but I'd argue that it could fairly be described as a 'class-A/B' -- the oblique being significant here -- which is what he did!

The real point, of course, being that these amps weren't modern Class-Ds, so they weren't very power efficient and therefore generated a lot of heat -- which it KM1 certainly did!

I would also assume the amplifiers used '50Hz' supplies? Those transformers would have run pretty hot as well but I guess at the time SMPSUs were not a reliable option?

Yes, they were linear supplies. You can just see the PCB across the top of the transformers, and the reservoir caps in some of those pics, sitting below the vertical amp cards in the base of the chassis.

Ok Hugh, well I just about recall the Wireless World article about the current dumping amp and distinctly remember that the design would be for 'worse case zero current in the output devices' until they need to dump that is. I cannot see any way to classify that OP stage by the usual nomenclature? Everyone is surely going to take "A/B" to mean a significant standing current? Could be 150mA or more.

Bit 'o' History! The late Clive Sinclair's company produced a handy little amp, the Z50 and that used a zero current OP stage which only turned on for signals above about 1V pk-pk. The design was called the "Edwin amplifier" I used two in bridge mode to drive two woofers for a primitive sub.

Dave.

[Hugh Robjohns]

I always wondered what "current dumper" meant; my present understanding is that the output is in two parts: a large coarse drive to do the donkey work and a smaller drive with more finesse, to correct for the errors in the coarse driver ? Is that right ??

Pretty much.

The small amplifier is actually capable of the full voltage swing required at the output, but with a restricted current capability. In the Quad 405 it can delver around 5W of power and at low signal levels, it provides the entire output.

However, for louder signals a pretty crude class-B stage is used to 'dump' the necessary current into the output to provide the required power. The amount of current needed is determined via a clever load sensing scheme which also leads to the generation of additional current from the small amplifier which exactly cancels out any distortion products from the current dumping transistors -- this error signal being delivered straight to the output in a 'feed forward' arrangement.

Peter Walker's original description (in Wireless World, Dec 1975) of the current dumping concept is available on the DADA website here: https://www.dadaelectronics.eu/uploads/ ... r-1975.pdf

[Hugh Robjohns]

Ok Hugh, well I just about recall the Wireless World article...

I've linked to it in the post above. It's a fascinating read...

In the introduction he says (my emphasis):

Nevertheless, whereas feedback reduces distortion to a small and no doubt negligible amount, feed-forward carries the promise of reducing to zero the distortion of that part of the amplifier over which it is applied. If this is the class B stage, then not only does the distortion itself disappear but all the paraphernalia of quiescent current adjustment and thermal tracking disappears with it.

So PW himself recognised that what he was doing in his 'current dumping' approach was to feed forward an error-correcting signal for a Class B output stage... It's just that in his refined design, the class B is effectively biased so that it doesn't even switch on for quiet signals.

There is no specific class description specifically for a current dumping amp, and it really is an amalgam of Class A and Class B... so in think in fairness we should cut Phil a little slack for his short-hand description, and especially so when the amp topology is of only minimal significance to the point he was trying to make... which was that the KM1 got bloomin' hot and a major part of its design was about dealing with that!

[ef37a]

Fair enough Hugh (sorry Phil!)
I guess Mr Walker dare not use the term "Class C" which is appropriate for a device with zero current! There is a paradox in the sense that he refers to biasing and thermal tracking which is of course the mark of a class B stage as it is usually configured.

Dave.

[Folderol]

Being somewhat pedantic, neither classification is correct. The closest I can think of would be class A || C
It's a class A amp in parallel with plus and minis two halves each of which conducts for less than a half cycle

[ef37a]

Being somewhat pedantic, neither classification is correct. The closest I can think of would be class A || C
It's a class A amp in parallel with plus and minis two halves each of which conducts for less than a half cycle

I am pretty sure Duggy Self defines it as " A+C" but I shall look it up shortly.

IIRC he defines class B as that current condition which gives the least crossover distortion and eschews class AB as it actually produces more distortion than the previously defined class B.

Dave.

[Hugh Robjohns]

KM1s did run very hot though....

[Nazard]

I am pretty sure Duggy Self defines it as " A+C" but I shall look it up shortly.

Doug designed the 'Trimodal Power Amplifier', whose PCBs, inc parts list, design notes, etc., are available via The Signal Transfer Company. It is switchable between Class A, AB and B: "there are no compromises of any kind in either mode. " So if you build one, then you can form your own judgements. Don't think I haven't been tempted!

http://www.signaltransfer.freeuk.com/trimodal.htm

There were some issues with the Signal Transfer Company, a year or so ago, which Doug told me about, and very kindly resolved: I shall not go into the details. But the take home message is that Doug will supply the boards directly via the website.

P.S. One of the things I love about the 405, is that the heatsink is a design feature, rather than hidden around the back, as per the norm.

[N i g e l]

One of the things I love about the 405, is that the heatsink is a design feature, rather than hidden around the back, as per the norm.

Definitely, the design still looks well cool !

Peter Walker's original description (in Wireless World, Dec 1975) of the current dumping concept is available on the DADA website here: https://www.dadaelectronics.eu/uploads/ ... r-1975.pdf

Ive bookmarked that for later. Your explanation has satisfied my curiosity for now