Quad 909 Technical details.

[Hodges]

I wonder if anyone can answer my question.

I have a 405 which has been and still is superb. However, I recently bought a pristine Sugden Bijou Headmaster, which is "allegedly" a Pure Class A Headphone amplifier and preamp. This was bought to replace my 33 which is now 51 years old and deserves to be retired. But I have some Stax SR 3 headphones which will get more use as we are moving to an apartment - downsizing.

The 405 uses a very high quality class A amplifier in it's design, but is not sufficient to drive my Stax before transiting over to Current Dumping.

So, I bought a spanking Quad 909 - brand new really - in the hope that it utilized Class A up to a higher output. It is not the sole reason for buying the 909, as I have long owned Quad gear, however, anecdotally it is said that it does run in class A at a higher level - is this true and if so, what power level would that be.

My Yamaha/Sugden/909 combo, driving either the Stax or my Spendor BC1's, provide superb listening, so my question is really academic, given how pleased things are working out - perhaps the addition of a pair of headphones to use directly with the Sugden in future, however, it would be interesting to know a little more about the 909.

 

[Arkless Electronics]

Forget class A from Quad current dumping amps! It's only a few mW, 18mW in fact in the case of the 405 and prob maybe 30mW ish for the 909.

 

[Hodges]

Forget class A from Quad current dumping amps! It's only a few mW, 18mW in fact in the case of the 405 and prob maybe 30mW ish for the 909.

The quiescent temperature of my 909 seems to indicate, more than just a few milliwatts. Not only that, at the levels i am playing - very low, the temperature of the amplifier drops. How does one explain that?

 

[Arkless Electronics]

I have nothing to add to my previous post. The 405 figure is accurate and the 909 one will be in the right ball park.

 

[Hodges]

I have nothing to add to my previous post. The 405 figure is accurate and the 909 one will be in the right ball park.

That's fine. However Quad confirmed to me the 405 is circa 32mW. Still not enough for what I was hoping for.

However, it is the 909 I am talking about and the heat sink temperature does drop from the quiescent value when playing mood music at the kind of level that I want. It does seem a bit odd however, that further development over 40 - ish years, would see Quad mimicking their original design?

Tightening up on component tolerances alone would not be sufficient - I don't think - to keep this design in vogue. There has to be more to it than that.

 

[Hodges]

Further to the above, is this attached thread. As can be seen from it, there has been quite some work done by well informed individuals. A figure for the 405 remaining in class A is given as 1 watt into 8 ohms. This seemingly was also proven to be correct with the dumpers disconnected.

The question remains however, what power level will the 909 remain in class A, before the Dumpers switch on?

My observation about the amplifier temperature dropping below quiescent when driven by low level music to me is relevant given the 909 is a direct coupled amplifier and I believe this in effect causes the less power to be dissipated by the heatsink - transferred to the load (loudspeakers). Hence the tangible lowering of amplifier temperature, whilst the the Current Dumpers remain switched off.

https://www.diyaudio.com/community/threads/quad-current-dumping-class-a-output-power.406673/

 

[Arg]

The quiescent temperature of my 909 seems to indicate, more than just a few milliwatts. Not only that, at the levels i am playing - very low, the temperature of the amplifier drops. How does one explain that?

That is how Class A works. The amp is running at full power all the time. At idle it is all heat. As the music goes louder the power is shared between music power and heat power. Hence less heat.

 

[Hodges]

That is how Class A works. The amp is running at full power all the time. At idle it is all heat. As the music goes louder the power is shared between music power and heat power. Hence less heat.

''

That is how Class A works. The amp is running at full power all the time.''

A Class A amplifier does not run at full power all the time, when quiescent. The current through the output stages is ''Half Peak Current' as the stage is biased midway along the 'Transfer Characteristic'. When a signal is present, it swings above and below the bias point. Given it is sinusoidal, it's mean value is zero - positive going it adds to the bias current, negative going it subtracts, so net effect is no change in heat sink dissipation. However, the signal is transferred to the loudspeakers.

My point with regard to the 405 and all current dumping amplifiers is; a) these amplifiers are direct coupled, b) if a sinusoidal signal is applied and its mean value is below the quiescent current value, so low that the peaks of the signal do not exceed the standing bias current, then the temperature of the heat sink will drop, as the current through the output stages is reduced.

In effect, a signal with peaks which do not exceed the quiescent current value will ''Shift" the bias point of the output stages below it's quiescent value. It is acting to a degree, in the manner of dynamically modifying the amplifiers 'Transfer Characteristic'. Essentially, with the 405 and more particularly the 909, the quiescent temperature is that which is generated - not entirely - by the high quality Class A Section, as the Current Dumpers are off. Providing I drive the amplifier and it's peak value does not exceed the switch - on point, the output will remain in Class A and the temperature of the amplifier will fall below its quiescent value.

I have yet to take some readings of the actual temperature change, but it is quite substantial - I use the term substantial; in a relative context, as my main point is my listening level is now very much lower than hitherto.

This is of course equally true of any Class A amplifier that is directly coupled, providing peak excursions; +/- do not exceed the quiescent current value. I specify 'Direct Coupled' as the nominal output voltage is but a few milli volts above zero, when the amplifier is quiescent.

 

[Arkless Electronics]

Where to start with correcting most of that..... nah lifes too short but I'll just say that it has nothing to do with being direct coupled or not.

 

[Hodges]

Where to start with correcting most of that..... nah lifes too short but I'll just say that it has nothing to do with being direct coupled or not.

You could start by explaining yourself rather than simply being dismissive. I too could go further into my reasoning/logic, but, in my case, life really is too short. However, the point I am making about direct coupling is, that the DC conditions that exist within the amplifier, are not separated from one stage to the next, by isolating capacitors. Therefore, dynamically, the amplifiers operating point will be modified by the signal drive. Under drive - the standing bias on the output stages is no longer fixed, but alternating in sympathy with the signal drive. I am referring to signal levels below the point at which the Current Dumpers are brought into play.

High Power Broadcast (2 Mega watt), TV Broadcast Transmitters, Tropo-scatter, Earth Station HPA & Receiver Systems, UHF Trackside Data Communications Systems, Ship Borne Naval Communications Systems, Government HF Communications Systems, Gas Exploration Rig HF Data communication systems, etc, etc, I could go on - Engineer, Retired.

 

[Arkless Electronics]

Other than in long term DC conditions ie over many seconds and even minutes, there is no difference between capacitor coupled and direct coupled amplifier in terms of operating point being dynamically changed. More often than not even an amplifier with a capacitor at the output, such as the Quad 303 or Armstrong 600 series etc etc, is direct coupled, much as a dual rail amplifier would be but the DC feedback keeps the output (pre output coupling capacitor) at the mid rail voltage (to allow max voltage swing) in exactly the same way that it keeps a dual rail amps output at nominally 0V DC for no signal input.

When loop feedback is used over an amp that has coupling capacitors it is essential to minimise the number of capacitors within the loop otherwise the cumulative phase shift will annoy Mr Barkhausen at some very low frequency and we can get the bass driver moving from one excursion limit to the other at say 0.5Hz....

Emitter/cathode bypass capacitors are often found even in overall DC coupled amplifiers and must of course have a long time constant compared to the lowest frequency of interest.

I dabble with RF but no doubt you could teach me a thing or three on that front!

 

[Arg]

@Hodges please fix your post so that it is not saying "Arg said" and then a whole lot of text that I never said.

 

[Arg]

A Class A amplifier does not run at full power all the time, when quiescent. The current through the output stages is ''Half Peak Current' as the stage is biased midway along the 'Transfer Characteristic'. When a signal is present, it swings above and below the bias point. Given it is sinusoidal, it's mean value is zero - positive going it adds to the bias current, negative going it subtracts, so net effect is no change in heat sink dissipation. However, the signal is transferred to the loudspeakers....

It does run at full power all the time. I am talking about power consumption, not power output. I'm obviously not talking about power output, that's the music!

If we put a power meter on the wall socket where a true Class A amp is drawing its power, we will observe that the power consumption is constant. Whether the amp is idling, or running full power output, or anything in between, it consumes the same power. It runs at full power all the time. Nature of the beast.

Now, first law of thermodynamics, all that power running into the amp has to go somewhere. It's either heat or music. When it's idling, there is no music, so it's all going into heat. But when the music starts, some of the (constant) power consumption is going into music, so less of it is going into heat.

All of which, as per my post #7, explains why your 909 gets cooler when low-volume music starts playing.

Of course, when cranking the volume further up, the current dumping circuit kicks in, and the amp's temperature will increase as that circuit draws its own (fluctuating) power from the wall, some of which is converted into heat.

cheers

 

[Hodges]

@Hodges please fix your post so that it is not saying "Arg said" and then a whole lot of text that I never said.

Sorry, if that is the implication as it was not intentional. However, I am enjoying this. It is really great to keep the old - grey matter churning and I am still curious about the actual function of the Current Dumper Amplifiers of Quad. My observations could have been better explained, but in essence, I am specifically referring to the output stages of the power amplifier - in the case of the Quad 405/909, or the output stages of amplifiers such as Sugden's A21. Delving back decades, the output stages are 25% efficient. The Standing Bias on the output transistors puts the Quiescent current as 50% of peak. However, this is the no signal conditions, but the issue I am wrestling with is when the amplifier is handling a signal, the peak of which does not invoke the Current Dumpers. If the peak signal which I have chosen, is up too, but not greater than when the Dumpers are invoked, if my theory is correct, dissipation in the heat sinks will drop as the value becomes only 0.7071 of the Standing Bias Current. This is my understanding of why the 909 runs cooler than when quiescent. When the Current Dumpers are invoked, by higher signal drive, then power dissipation increases and the heat sink temperature will rise as with a Class A/B amplifier. So I am struggling to see that power consumption is constant. When the current dumpers operate, although dissipation increases, efficiency of the output stages improves; when not, it returns, at best to the 25% figure for the Class A section.

 

[Arkless Electronics]

There is rather more to current dumping than it at first appears. If you visit the worldradiohistory.com site you can find all the past issues of wireless world including the original papers on the subject by Walker and Albinson. There is then years of debate and further analysis including from Baxandall, Ellis, and Vanderkooy and Lipschitz. There is quite some disagreement between even such luminaries as to how current dumping even works!

 

[Hodges]

There is rather more to current dumping than it at first appears. If you visit the worldradiohistory.com site you can find all the past issues of wireless world including the original papers on the subject by Walker and Albinson. There is then years of debate and further analysis including from Baxandall, Ellis, and Vanderkooy and Lipschitz. There is quite some disagreement between even such luminaries as to how current dumping even works!

Thanks, I'll have a look - see!

 

[Arg]

Sorry, if that is the implication as it was not intentional. However, I am enjoying this. It is really great to keep the old - grey matter churning and I am still curious about the actual function of the Current Dumper Amplifiers of Quad. My observations could have been better explained, but in essence, I am specifically referring to the output stages of the power amplifier - in the case of the Quad 405/909, or the output stages of amplifiers such as Sugden's A21. Delving back decades, the output stages are 25% efficient. The Standing Bias on the output transistors puts the Quiescent current as 50% of peak. However, this is the no signal conditions, but the issue I am wrestling with is when the amplifier is handling a signal, the peak of which does not invoke the Current Dumpers. If the peak signal which I have chosen, is up too, but not greater than when the Dumpers are invoked, if my theory is correct, dissipation in the heat sinks will drop as the value becomes only 0.7071 of the Standing Bias Current. This is my understanding of why the 909 runs cooler than when quiescent. When the Current Dumpers are invoked, by higher signal drive, then power dissipation increases and the heat sink temperature will rise as with a Class A/B amplifier. So I am struggling to see that power consumption is constant. When the current dumpers operate, although dissipation increases, efficiency of the output stages improves; when not, it returns, at best to the 25% figure for the Class A section.

I think you are confusing yourself by thinking of the Sugden A21 and Quad amps as both Class A. The Sugden is, but the Quads are not. The output stage of the Quad amps operates in Class B.

Therefore, it is a mistake to use Class A logic to explain the heat sink temperatures of the Quad amps when you are "specifically referring to the output stages of the power amplifier". The output stage of the Quad amps is Class B, its efficiency is as per Class B, and the heatsink temps will vary as per a Class B amp, at least when we are, as you desire, "specifically referring to the output stages of the power amplifier".

So, when I say "power consumption of a true Class A amp is constant", I am talking about the Sugden but not the Quads. And I am correct.

The Quad amps do have a peculiarity for a power amp, though: the output stage only turns on once the power output reaches a certain (very low) level. Until then, the prior stage, ie the voltage gain stage, provides both voltage and current, and that stage is Class A. This is unlike a Class AB amp, where the output stage is always in operation, but there is enough bias current for Class A at low power, and the output stage itself is sometimes in Class A and sometimes in Class B. For the Quad amps, the output stage is plain old Class B, albeit a clever one.

That is why my initial answer is correct: at idle and at very low output power, the Quad is only operating its Class A circuit, which has constant power consumption like any true Class A amp, which is all heat when idling, and diverts some of the heat to music as the music volume is slowly increased at low levels, causing the heatsinks to cool. But once the output stage with the current dumpers turns on, that is Class B and the heatsinks get hotter the louder the music is played…as per any Class B amp.

Cheers

 

[Hodges]

Thanks Arg.

However, Quad themselves state the 405, has a very high quality Class A amplifier, which is always in control of the output current dumpers. Investigation has demonstrated, the Class A section of the 405 will produce up to 1 watt output power - current dumpers were disconnected to determine its power rating. I have not seen any articles relating to the 909's Class A section.

However, at no time have Quad indicated the Current Dumping action is Class B and indeed it would be strange to be given patent rights for an amplifier, which has already been established for many decades. My understanding is, the dumpers are switched on, at a point well away from the non - linear region of the output device junctions, so that would suggest it is beyond ''Projected cut-off'', even twice. In doing so, ensures the dumpers start to conduct well into the linear part of their transfer characteristic.

The attached PDF shows the feedback, but the text indicates it is an error signal, meaning - I presume - that which is derived from the error generated between the class A section and that of the Current Dumpers when they switch on. For me, the only error that would exist, assuming the dumpers do turn on well away from the none linear region, would be a phase error.

The output voltage of the dumpers; when they turn on, is virtually the same as that of the Class A section as their voltage gain is approximately 1. So the error is the difference between the Class A Voltage drive to the dumpers, and post dumper output voltage. If the dumpers are, as I believe, switched on, well away from the non - linear region of the devices which handle the +/- signal excursions, and they switch on extremely fast, then the Class A section is left to deal with the region where cross - over distortion would normally occur in a Class B amplifier, yet at the same time, dictating exactly when the dumpers are brought into service.

I acknowledge your point regarding transfer of energy/heat from the amplifier heat sinks to the loudspeakers in the form of Audio Power.

 

[Arg]

If you are not going to admit that the dumpers are a Class B amplifier (they are), then we can't move on.

Peter Walker's own 1975 paper describes the principle: "...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." That is a description of the current dumper stage.

The amp is a low power Class A stage, followed by a high power Class B stage with unity voltage gain (ie current amplification only) that only turns on when needed, and the latter has this very clever configuration that corrects distortion, self-maintains bias (at exactly 0 mA, ie Class B) and auto-tracks for thermal drift.

A bit of background reading on current dumping amps might help to clarify that they are invariably Class B...in fact, if one wants to be really picky, one could argue they are Class C because the conduction cycle of the output transistors is not quite 180 degrees. But if Walker is calling them Class B, then let's go with that, since the practical use of the term Class C is for amps with much less than 180 degrees of conduction cycle.

Walker's 1975 paper in Wireless World, link.

A follow up 1981 paper in Wireless World by Malvar, expanding on Walker's early paper and leaving no doubt that the output stage is Class B, link.

cheers

 

[Arkless Electronics]

The dumpers are actually class A/B in practice. A small amount of forward bias is used (it was increased in later 405's) to reduce the distortion contribution from the dumpers.

The Crown DC-300A could be classed as a current dumper of a crude form but without the "magic" of the LCR bridge and feedforward etc of the Quad. The drivers are connected to the output and forward biased. They directly drive the speakers until enough voltage drop is present to turn on the actual output transistors. It's not unique in this and in fact a PE mag design used the same system. It is though a very clever and well designed amplifier, especially bearing in mind it's age. I was surprised and impressed to measure 0.003% THD from a DC-300A at 150W!