Tone Controls

[Radio Wrangler]

I suppose balanced signals on XLRs led them into it, but their impedances are wrong, anyway.

We always used dBm for the milliwatt, and this was making test gear for US/UK/Global telephone companies. I think dBu was recording studio terminology.

I think dBrnCO may have been the most convoluted decibel variant.

I really really don't like the looks of that Quad preamp.

A fiddly little touch surface where your finger obscures what you're trying to set and a lot of adjustable parameters that have to be buried in deep menu structures!

A very high price to pay for clean looks.

And they had it so right in the 44.

David

[Ted Kendall]

Originally Posted by Radio Wrangler

And they had it so right in the 44.

Hear, hear - bought mine new in 1983...

[Craig Sawyers]

I think it is too early to judge regarding the Artera. Many high quality audio companies are going (or have already gone) for minimalist appearance, with some of them needing a iPad, phone or similar to drive all their features via a virtual front panel.

I'll keep my powder dry until there is a sensible product review, or at least a downloadable product manual on the Quad website.

[Synchrodyne]

Originally Posted by Ted Kendall

Hear, hear - bought mine new in 1983...

I’ll “third” that – I bought my Quad 44 new in 1986. I also have a Quad 66, bought new in 1990, so I guess that’s a “fourth”, or quad of recommendations for the Quad tone control system. I change between the 44 and 66 arbitrarily, right now the 44 being in use. An interesting application for the tilt control is with some modern TV dramas, where the dialogue is perhaps more “verissimo” than it was in times of old, when it was more in theatrical declamation style. (E.g. go way back to the real Forsyte Saga series, the Lotus Eaters, or in the 1980s/1990s series like Telford’s Change, A Kind of Loving, The Camomile Lawn, etc.) Anyway, SWMBO finds some of the current drama dialogue hard to understand (The Bay comes to mind as a recent example), and simply turning up the volume doesn’t help much. But a step or two of uptilt on the Quad does the trick. Then in addition to that, depending upon the programme, I find just a touch of filter (10k on the 44 with a smidgeon of slope to suit) sometimes helps.


Cheers,

[Synchrodyne]

Originally Posted by Edward Huggins

Continuing this informative Thread, I have always been somewhat curious about the use of their familiar, much used, description as "continuously variable tone controls".
Essentially, all potentiometers are continuously variable, including their application as volume controls as well as tone controls.

Some tone controls were/are operated stepwise by switches with a finite number of positions rather than by potentiometers.

Before c.1939-40, if potentiometers were used to obtain continuous variability, then with the circuits of the time, it was necessary to switch between the lift and cut functions. In other words, treble and bass controls of this type were continuously variable over their respective lift and cut ranges, but discontinuous between the lift and cut functions. The Volkoff and Voigt, and following similar passive circuits enabled continuous variation over the full range, from full lift to full cut. The Baxandall did the same in a more elegant way.

I don’t think that there are any hard and fast rules, but reserving the “continuously variable” descriptor for those tone controls that use potentiometers that sweep from full lift to full cut without interruption strikes me as reasonable. That would encompass the majority of examples from the early-to-mid 1950s onwards. Switching between lift and cut was used to some extent in the late 1940s, e.g. in the Williamson control unit. Fully switched tone controls, where say rotary switches went from full lift to full cut in a number of steps, retained a minority following. (Later on at least, sometimes they were in the form of continuous potentiometers constrained to move in steps rather than continuously.


Cheers,

[Synchrodyne]

Originally Posted by Synchrodyne

J.E. Sugden also included a similar “Quiet” control in its C51 control unit of the late 1960s. (I think it was also included on the preceding C41 which was sold as a Richard Allan product.)

Sugden’s rationale for its “Quiet Control” was given as:

“Attenuates mid frequencies by 16dB but lower frequencies to a less extent in accordance with the subjective equal loudness curves of the ear. This facility where normal volume is set and then "quiet" selected is to be preferred to the so called "loudness" compensation of a volume control which has no relevance to the absolute value of the incoming signal level.”

I have not yet found a schematic for the Sugden C51, and the available data do not include a curve showing the effect of the quiet control, just those for the tone and filter controls. Whilst Sugden was aligned with Radford in respect of the quiet control, it evidently differed to some extent when it came to low-pass filters. The C51 had the same turnover frequencies as used by Radford in its SC22 and SC24, namely 4, 7 and 10 kHz, but offered two slopes for each, namely 6 and 18 dB/8ve, the latter steeper than Radford thought advisable.

With their quiet listening controls, both Radford and Sugden sought to address one of the several criticisms, both theoretical and practical, made in respect of simple loudness control systems, which was that their frequency bending action was related simply to volume control position and not in any way to sound pressure level. That is, when switched in, they simply changed the volume control from a frequency insensitive device to a frequency contoured device – at least at lower levels - with a rather arbitrary transition point.

I have since found the schematic for the Sugden C51, but not the “quiet” control curve. Here is a derived block schematic.



The tone and filter control section was based upon a three-stage circuit, namely a first emitter follower stage following the volume control, then a second emitter follower, then a single transistor gain stage (without a bootstrapped output load).

The low-pass filter was wrapped around the second emitter follower. I haven’t unravelled it, but what it might be is switchable a 2nd order (12 dB/8ve) Sallen-Key filter with three switchable turnover points, preceded by a switchable passive 1st order (6 dB/8ve) filter, again with the same three switchable turnover points. On the other hand, such an arrangement would appear to allow for three slopes, namely 6, 12 and 18 dB/8ve, whereas the C51 has only two available, namely 6 and 18 dB/8ve. So perhaps the circuitry is something different.

The quiet control was a switchable passive network following the second emitter follower stage, providing both attenuation and bass contouring. When switched in, it formed a divider of the type R over [R plus (R//C)]. That I think would provide a step-curve with bass lift.

The treble and bass controls were essentially of the Baxandall type, wrapped around the gain stage transistor and with the feedback arm fed from the transistor’s collector via a resistive divider, 1k over 4k7. Assuming that the Baxandall network was fed from the unity gain point, that implies that the stage provided around 15 dB gain from input to output. The nominal gain from high-level input (150 mV) to output (600 mV) was 12 dB, which allowed room for some losses in the filter circuitry. Unusually, the Baxandall circuit included bass and treble trimmer potentiometers, in parallel with the respective main potentiometers, to allow trimming for an exact flat response with the controls at their centre positions.

The C51 had a switchable high-pass (rumble) filter, but this was in the disc input/equalization stage (of the Bailey three-transistor type), so operated only on that input. The filter circuitry (R-C) was in the DC feedback loop between the output emitter follower and the first stage emitter. (The RIAA (or other) equalization was in the corresponding AC feedback loop.)

The Sugden C51 design dates back to the time when transistors were relatively costly, and so had to be sued quite sparingly, hence the “compact” circuitry, with only 6 transistors per channel. I suspect that it used substantially the same circuit as the Richard Allan C41 (designed and probably made by Sugden), and that goes back at least to the 1968 April London Audio Fair, and perhaps back into 1967. The Quad 33, from late 1967, also had just 6 transistors per channel. The much more complex Radford SC24 had 50 transistors total, two of which were in the power supply, leaving 24 in each audio channel. In part the much higher transistor count reflected that it was aimed at a much higher price point (around twice that of the 33 and C51), but almost certainly the reduction in unit transistor costs that had occurred over 3 or 4 years helped to allow an apparent design approach of “let the device count lie where it falls”. Hence the spread-out nature and complexity of the SC24 tone controls.



Cheers,

[G8HQP Dave]

It would be easy to add mechanical detents to a potentiometer so that although electrically continuously variable it can only rest at certain positions. I seem to recall that some 1970s audio did just this. It gives the user impression of a switched attenuator, while actually being much cheaper to implement.

[Synchrodyne]

As mentioned in post #126, Sudgen used an RC feedback type of low-pass filter in its C51 control unit.

Linsley Hood did similarly in his 1969 Modular Pre-amplifier, Wireless World 1969 July p.306ff. Here is his actual circuit, based upon two-transistor gain stage:



This is shown (on the left-hand side of the image) for a single value of the capacity Cx. This would be varied according to the desired turnover frequency. Not shown is the switching that would be required (3-gang) to obtain a choice of turnover frequencies.

Linsley Hood had also presented (on the right-hand side) the circuit options available to achieve the desired result.

Circuit (b) was attributed to Baxandall, whom Linsley Hood said was the first to use an RC feedback filter in an audio amplifier. Circuit (d) was said to be the Sallen and Key rearrangement to use a unity gain amplifier. Circuit (c) was chosen for the Modular Pre-amplifier because of its suitability for use with a simple two-transistor amplifier.

The Baxandall reference was to his 1955 Gramophone and Microphone Pre-amplifier, WW 1955 January and February. Therein Baxandall went into reasonable detail to show that RC feedback filters could be made equivalent to LCR prototypes. He included the comment:

“It is sometimes said that filters using resistors and capacitors only, in suitable feedback circuits, give better transient response than can be obtained with passive filters which include inductors. In general, however, this notion is quite incorrect, and any filter employing feedback principles may, in fact, be shown to be equivalent, in both frequency response and transient response, to a particular passive filter using inductors. The feedback filters employed in the present equipment are equivalent to, or "simulate," simple constant-k filters" with one π (or T) section and resistive terminations, the rate of cut-off tending to 18 db/octave.”





Baxandall did not claim originality for these RC feedback filters, but rather referred back to an RCA article from the 1950 January issue of RCA Review, namely “Design of High-Pass, Low-Pass and Band--Pass Filters Using R-C Networks and D.C. Amplifiers with Feedback”, p.534ff. Baxandall would have developed and published his circuit a little ahead of Sallen’s and Key’s work, which I think was during 1955.

Baxandall used an EF80 high-slope pentode for the filter/output stage, which had a gain of about 20 dB. Apparently the valve gain was preferably quite high in order not to unduly affect the calculated filter characteristics. Circuit variations were also offered for the EF91 (CV138) and SP61 (VR65, CV1065). (This speaks to the concurrent EF91 vs. EF80 discussion at: https://www.vintage-radio.net/forum/...d.php?t=157399.) The (Mazda) SP61 was the valve used in Baxandall’s original tone control circuit. Its Mazda successors were the 6F13 and then the 6F1, with the EF91 and EF80 being in broadly the same class. Nonetheless, all other stages of the 1955 pre-amplifier used the EF86, including the tone control, thus:

“Tone-control Stage.—The tone-control circuit is almost exactly as previously published, but an EF86 valve is used in place of the high-slope valve originally specified, in order to secure reliable freedom from microphony and hum. The signal output from the tone-control valve, for a final output from the pre¬amplifier of 4 V, is 400 mV; under these conditions, the non-linearity distortion introduced by the tone-control stage is much less than 0.1 per cent despite the low-slope valve employed.”

An observation is that in both the Linsley Hood and Baxandall cases, the 18 dB/octave low-pass filter was effectively a combination of 6dB/octave and 12 dB/octave units, with the latter having a slight “hump” at its turnover frequency that was smoothed out by the 6 dB/octave unit, which had a “clean” turnover.

That “hump” might (or might not) explain why the Sugden C51 filter could be switched to operate at 6 or 18 dB/octave, but not at 12 dB/octave. The 6 dB/octave filter could be used standalone, but the 12 dB/octave filter was less suited to standalone use because of its hump. Possibly Sugden saw no need for other than the two steps. On the other hand, 6, 12 and 18 dB/octave seems to be a reasonable sequence of steps in a situation when some of the competition of the era were offering continuously variable slopes over that range or more. But it would appear that adding a 12 dB/octave step (with a suitably smooth turnover) would have involved more extra complexity than was warranted.

The above-mentioned Wireless World and RCA Review articles are available at the excellent American Radio history site, https://americanradiohistory.com/Wir...d_Magazine.htm, and https://americanradiohistory.com/RCA..._Issue_Key.htm.


Cheers,

[Craig Sawyers]

The switched frequency variable slope LCR filter that Quad adopted in the 22 and 33 preamp originally appeared in November 1956 WW, by Leakey "Inexpensive variable slope filter"

I think that the simpler version of Leakey (that was actually adopted in the 22 and 33) was due to Baxandall. I'll try to find the reference to that.

Craig

[Synchrodyne]

Interesting!

I was under the impression (not necessarily correct) that there was direct lineage from the Quad QCII, through the Quad 22 and Quad 33 to the Quad 44. And also some linkage back from the QCII to the original Quad (Quad I).

I'll need to map out the filter circuits to see them in isolation. For the QCII and 22 it is hard to tell what is what given that their filter circuits are entwined with the tone controls.


Cheers,

[kalee20]

There's an article in May 1971 Radio Constructor which describes a rather elegant tone control which I've not seen in practice. With ideal devices, it's theoretically capable of flat response with potentiometers mid-position, and set at extremes, the response heads to zero or infinity.

With bass and treble combined, there would be interaction between them. But two circuits in cascade, one for treble one for bass, there wouldn't.

Photo of the article below.

[Craig Sawyers]

The QCII used a different circuit topology to generate the filter curves, and was introduced in 1953. So that precedes the Leakey paper by 3 years.

The Quad 22, introduced in 1957 adopts a simpler circuit, of the same topology as the 33, but using smaller capacitance values (and hence larger values of inductance).

In the 44, the same topology as the 22 and 33 is used, but with component values again changed. The capacitors again increase, needing yet smaller values of inductance.

With the 34, the filters are active, provide only two frequencies and a steep/gradual selection.

[Synchrodyne]

It’s a pity that the author of that Radio Constructor article did not provide a representative set of curves. Still the bass and treble combination shown does embrace earlier thought about symmetry:

Originally Posted by kalee20

I've always felt vaguely unhappy about 'classic' tone controls, such as the Baxandall one, that they are basically non-symmetric for bass and treble, if one uses R and C, then the other should use R and L. Some day, I may try and make one.

At one time, the use of LCR tone control circuits seemed to be unexceptional. Sturley included some examples in Part 2 of his book “Radio Receiver Design” (1945). But post-WWII, there seems to have been a general move to R-C tone controls, and perhaps to a slightly lesser extent, to R-C filters.


Cheers,

[Synchrodyne]

More on the chronology of the Quad LCR low-pass filters. (I had written this before I saw Craig’s post #132, so it overlaps somewhat.)

To start with here is the simplified circuit for the original Quad (Quad I) control unit of 1951, the first from Acoustical to be fitted with a variable slope filter:



This had a pair of mutually coupled coils in series. The turnover frequency, 6 or 8 kHz, was determined by three sets of switched capacitors, respectively at the input and output ends of the coil pair and at the bottom end of the slope potentiometer. The input end and slider of the latter shunted the second coil. The slope range was 10 to 100 dB/octave.

Next came to Quad QCII control unit in 1953:



This retained the mutually coupled coil pair, this time separated by a 15k resistor. Turnover frequency, 5, 7 or 10 kHz, was determined by a switched series capacitor network (3 capacitors) between the input end of the second coil and the top end of the slope potentiometer. The bottom end went into the feedback network around the ECC83 double triode gain stage, joining feedbacks from both the treble and bass tone controls. The slider went to ground. So it looks as if the filter (like the bass control) was partly in the output arm of the amplifier stage, and partly in the feedback arm, the proportion depending upon the position of the slope control potentiometer slider. On that basis, this was the only Quad LCR filter that had a feedback element, the others all being passive. The slope range was 0 to 50 dB/octave. Having the slope control go right back to zero was a change from the Quad I, where it went back only to 10 dB/octave.

Chronologically, the Leakey circuit was next, in 1956:



This used a single coil, with switched RC networks (triple-gang switch) to set the turnover frequencies of 5, 7 and 10 kHz. Here the slope control potentiometer was configured as a rheostat across the coil, the slope range being 6 to 40 dB/octave.

The Quad 22 stereo control unit arrived in 1959:



Here a change was made to a single coil with tappings for the 5, 7 and 10 kHz turnover frequencies. The 10 kHz tapping (fewest turns) was shunted by a capacitor. The active part of the coil was shunted by the fraction of the slope control potentiometer between its top end and its slider. The bottom end went to ground (via a capacitor), and the potentiometer itself was shunted by a capacitor. The slope control range was 0 to 25 dB/octave. Note that in the cancel position only, the slope control became a pathway to the feedback network. At first glance I thought that the filter circuit had a feedback element, but closer inspection showed that this was not the case.

The Quad 33 control unit of 1967 had, component values aside, very similar circuitry to that of the Quad 22, albeit with apparently simpler switching from the push-button array as compared with the rotary switching on all preceding models. The filter circuitry was essentially standalone, completely following the tone control stage and not entwined with it.



Within that sequence, I should find it difficult to determine who informed whom. Leakey’s main concern was to show that the LCR filter with variable slope was, when properly implemented, an acceptable alternative to the by-then mainstream twin-T R-C type in a feedback amplifier. Nonetheless, it was surprising that he made no mention of the prior art from Quad, who by 1956 was well-known in British hi-fi circles. If anything, Leakey’s circuit had more complex switching for turnover frequency than that of the Quad QCII or Quad I.

Quad certainly made significant changes in the Quad 22 filter as compared with that of the QCII. Perhaps the use of a single coil was inspired by Leakey’s work, but on the other hand, it may also have been facilitated by the fact that the required maximum slope was 25 dB/octave as compared with the 50 dB/octave of the QCII. Leakey’s choice of a single, untapped coil was at least in part in deference to the home constructor, in terms of both economics and simplicity. Quad may well have seen the tapped coil, with simpler switching as a better choice from a production viewpoint. I think we’ll need to await the Baxandall commentary on the origin of the Quad 22 and 33 filter circuits to ascertain the actual development pathway.


Cheers,

[Synchrodyne]

To complete the sequence, here is the Quad 44 filter circuit:



It was a standalone item, placed in the middle of the whole control unit circuit, between the volume control and the tone control sections. Component values aside, it followed the Quad 33 pattern except that capacitor shunting the 10 kHz section of the coil was omitted. Turnover frequencies were 5, 7 and 10 kHz and the slope was variable from 0 to 25 dB/octave, the same as had applied to the Quad 22 and 33.


Cheers,

[Craig Sawyers]

I could not find an image of the filter in the QCII. But this is the inductor in the 22. Both left and right channel tapped inductors (L1 and L2) are in the same large housing

Next the 33, which used larger capacitors and smaller and separate inductors. This one has been messed with, with modern resistors and capacitors in evidence.

And finally the 44, with again larger capacitors and very small pot-core inductors.

[Craig Sawyers]

Found an image of the (mono) QCII. Note the size of the (single) inductor, on the left side of the chassis.

[Edward Huggins]

This learned thread continues apace. However, it could be said that true Hi Fi amplifiers of the "straight wire" type (i.e. no tone controls at all) do not need such devices and that they are far better suited to, er, lesser equipment, wherein some enhancement is necessary to compensate for their non-linearity....

[Radio Wrangler]

Back in the 80's I for my own preamp, I synthesised networks to give the same shapes as the Quad tilt control which I rather liked, and cloned the 33 filter. I used RM8 pot cores in Siemens N28 material... I think Al=400 gapped.

The filter works fine, though I've hardly used it. I use the tilt control a lot.

With the 44, Quad had a product very suitable to its use. One glance and you could see the status of everything. One control actuation made any change you wanted.

Menus are a big step backwards and should only be used when forced by the sheer number of parameters. Going to them just to save money on switches is a cop-out. Going to them to give a minimalist appearance, ditto. Sure, it looks pretty, but each and every time you do anything with it, it inconveniences you.

Nouvelle cuisine leaves you still hungry. Someone's lost sight of the primary function with the new one.

[Craig Sawyers]

I agree that a software driven preamp might sound like a daft idea - but we have been using smart TV's, CD players, NAS drives, and streamers that are all processor intensive and are menu driven, in the case of CD players for nearly four decades.

Things like the Aurender units https://aurender.com/ (these cost a small fortune, and incorporate a hard drive for things like ripped CD's, a streamer, DAC etc etc) can *only* be driven via an iPad (though they say they now support Android). A well heeled friend bought one a couple of years ago, unboxed it, then had to drive out and buy an iPad before he could listen to it.

It is all going this way, for better or worse.

Craig

PS I have a story about nouvelle cuisine, on a business trip to Italy about 20 years ago when it was in fashion. After five courses that were basically small stains on large plates (This trip was with a Russian guy who kept saying "but where are the potatoes?"), we went out and found a McDonalds to fill up. That is the most delicious McDonalds I have ever eaten.