'Loudness' Control - fourth tap on volume control potentiometers

[David G4EBT]

This topic often seems to crop up when someone restoring a radio is puzzled to find a fourth tap on a volume control and wonders how to find a replacement pot. (I believe that Blore Edwards can probably make replacements, but it isn't something I've explored). It's said that the purpose of the fixed tap is to accentuate the bass at low volume levels so that the radio doesn't sound 'tinny'. I'm not sure whether there is any discernable benefit, but can't help thinking that if it was such a great idea, why didn't it feature as almost a standard item on may more radios, given the very small additional cost?

I came across the problem when restoring a Stella ST105U some time ago. The pot worked fine, but the switch was faulty. In the end, I was able to repair the switch by dismantling that section of the pot, cleaning and lubricating the contacts and reassembling it. At the time, I did wonder whether or not the 'fourth tap' could be simulated with a couple of external resistors across a standard 3 terminal pot. I was told 'no - it wouldn't work' and didn't bother following it up, but for illustrative purposes, I've attached a sketch of what I had in mind. Although the circuit is from the Stella 105U, the same principle could apply to any set with a 'loudness' tap.

As will be seen, the 105U uses a 500k log pot, tapped 10% down from the 'hot' end, at 50K. Why not then, use a 1 meg pot, with a 100k and 900k resistor across it (910k would in practice be the nearest a 'preferred' value), to form a tap? Certainly on a 3 terminal pot, if a 500k one is required and not to hand, a 1 Meg with a fixed 1 meg resistor in parallel to simulate a 500k pot works fine.

I thought it was a topic worth airing as it seems to crop up quite often and unless I've missed something, I don't recall seeing a definitive solution. Personally, if I come across a set with this problem again, I'd probably just disconnect the loudness components and fit a 3 terminal pot, because I'm inclined to the view that this flummery was a precursor to audio tomfoolery, but I know that many restorers will wish to preserve the original design concept as best as can be achieved. That said, the 105U is the same era as the Bush DAC90A, with a similar valve line up, though the 105U has a higher component count, (Stella: 37 Cs & 24 Rs, 1 thermistor - Bush: 22 Cs & 17 Rs), and to my ears, is a much nicer sounding radio than the DAC90A. I don't think that can be just down to the loudness control though.

Lots of new production stereo pots with loudness control taps on e-bay so it is a concept that features on audio equipment today.

There's a paper here about the topic - most other debates I've seen on internet seem to generate more heat than light.

http://www.extron.com/company/articl...nesscontrol_ts

Hope that's of interest.

[ms660]

It does indeed, there's a current thread on that subject here:

https://www.vintage-radio.net/forum/...d.php?t=127024

My conclusion is that a tapped pot is tapped, and one that doesn't have a tap isn't.

Lawrence.

[RF Burn]

Simple analysis says that your simulated pot will not work in the same way as the tapped pot, if at all. It seems that from the point of view of the circuit node '3' (top of pot) any effect on frequency response will be very reliant on the output impedance of the detector, as the actual frequency dependent node '2' is buffered from it by the 100K fixed resistor. It would be interesting to see what actually happens if someone can run your circuit through 'Spice' or similar ciruit simulator.

[mhennessy]

Hi David,

The idea of a loudness tap isn't audiophoolery, but it's not quite as it's usually described...

The usual explanation makes reference to equal loudness curves from Fletcher, Munson, et al. And yes, our hearing does become less sensitive to bass and treble as the volume goes down, so wouldn't it be nice if our radios could somehow automatically compensate for that?

The trouble is, the curve changes with level on a continuous basis, whereas a volume control with a single loudness tap does not. At all settings below the tap, the curve is fixed. As you move clockwise past the tap, then the effect does fall away (quite quickly in practice). But what are the odds of this small section of progression coinciding with your hearing and the level of ambient noise in the room? How well controlled are the audio levels coming from the radio station, and how do those levels compare with how they were when that radio was designed?

Today we have Optimods (other processors are available!) at the output of each radio station, and as a result, average levels are much higher (despite the fact the peak levels haven't changed). So some sets have problems today - I'll come back to that.

So, the science is a convenient "hook" to hang this on, but it's not the real reason. If it were, the loudness control would be a potentiometer in addition to the volume control (I have one on a Denon DRA-455 hi-fi receiver, and jolly useful it is too!).

The real reason is this: most loudspeakers in most radios require some bass and treble lift. Without it, they would sound "thin" and "forward".

However, bass and treble lift is expensive. I don't mean in terms of component cost as such - although that is part of the bigger picture - but in terms of power. A 3dB lift - which is audible but not dramatic - requires a doubling of power. Typical values needed might be 9 or 12dB, so several doublings. When you've only got a watt or two to play with, this doesn't leave much in reserve.

Suppose you didn't use a loudness tap, and suppose instead you just fitted this bass/treble lift into the amplifier as a full-time characteristic. Now, it'll sound great a low to medium volumes, but what happens when you need to turn up the volume?

What happens in this scenario is that the amplifier will be driven into clipping with these large amounts of bass/treble boost that actually, as we know from Fletcher-Munson curves, aren't actually needed any more. So if there are separate bass/treble controls, we could turn them down to allow the midrange to come up in level proportionally. But if not, then you have a very real limit on the maximum sound pressure level available from the radio.

So if you are incorporating a loudness tap, you have a juggling act. Ideally, you'll use it to extract the most amount of bass/treble boost you can while narrowly avoiding clipping the amplifier. Meaning that when the wiper is level with the tap, the bass/treble boost will be only a few dBs away from clipping, but hopefully as you wind up the volume, the midrange can come up in the mix.

For me, the "state of the art" is the often-discussed Grundig Yacht Boy N210. This has two loudness tappings, and an extraordinary amount of bass boost. And I reckon that in the early '70s, it was OK. But today, with highly compressed bass-heavy programme material and radio stations using Optimods at their output to squash things down even more, we find that the N210 is able to clip as you advance the volume control towards the tappings. It might be that adding ~10dB of attenuation ahead of the volume (or reducing the gain of the power amp by the same) will fix it, and one day I'll try it...

So, a loudness tap in a radio is not about Fletcher-Munson as such - despite what Marketing tell us - it's about getting the required bass/treble lift to help the small loudspeaker in a non-ideal cabinet give some semblance of quality.

The loudness tap is obviously very popular, but a nightmare for restorers. Fortunately for me, Hacker never used them, preferring instead to use generous loudspeakers and providing bass/treble controls that are actually "lift" controls (in most cases). With them set in the middle, you've got roughly enough lift to give a smooth response, but you can turn them down if you need to play loud.

It's worth noting that all loudspeakers in a cabinet - whether transistor radios or "proper" hi-fi loudspeakers - require bass lift at a frequency that is proportional to the width of the cabinet. This is built in to the crossover in decent loudspeakers. Some manufacturers omit it, forcing the user to put the speaker against a rear wall to recover the bass, but this positioning affects other parts of the frequency spectrum and the stereo image.

So back to the question. Yes, I'm afraid that your proposed modification won't do anything because the wiper of the potentiometer has no way to "see" the R25/C37 pair. It won't hurt, but it's a waste of effort (sorry ). For that reason, your proposal to just fit a standard pot and ignore the loudness components is fine, and arguably the most sensitive as someone in the future might be willing to shell out for a custom tapped pot.

As a result, the repaired set will sound different to how it did before. Is that a problem? That's a personal judgement. If you never heard that set working before, you might be delighted with it as it sounds, so no problem. However, you might really miss the bass boost (no treble boost with your example) - especially if it's a stereotypical "German" set, where there was lots and lots of boost built in. In which case, you'd have to modify the set elsewhere to achieve it. You could build a "loss network" ahead of the volume control, or you could add a series "R and C" across the feedback resistor somehow (with care, naturally!). But be aware that the boost will be there all the time, and it will restrict the maximum level you can play at before LF clipping sets in. Of course, if the set also features a bass control, then that might give you enough of a bass boost to overcome the loss of bass boost that was previously being supplied by the loudness tap.

Does this help to clear things up?

All the best,

Mark

[stevehertz]

Just out of interest, my Yamaha CR-2020 receiver has a continuously variable loudness control that works like this. You turn the loudness control to the fully clockwise position, at which it does 'nothing' ie has no effect on tone control settings at any volume, it is flat. You then set the volume to your normal (maximum?) listening level and adjust your tone controls to suit. Thence onwards, any reductions to the desired volume level are done using the loudness control, and the more you turn it 'down', the more the 'loudness curve' is accentuated to reflect the needs of our ears as identified by the Fletcher Munson curves. It works well since you have identified and set a true datum with your original normal listening settings that are 'correct' for you, and the loudness control then works with that datum, not at some arbitrary point.

[Andrew B]

I just want to chuck this morsel in that I remember from school. That the human ear has a logarithmic response to volume and to frequency, based on natural logs.

[David G4EBT]

Thanks everyone for reading the thread and for your considered input.

I guess that the only way I'd know if the loudness tap makes any real discernable difference to my Stella ST105U would be to disconnect it for a comparison. Certainly many nice sounding sets lack the loudness tap, albeit they do have tone controls. One of my favourite radios, both in terms of sound and appearance, is an HMV 1122 'woodie'. An expensive set in its day (1950), which looks like it was built up to a standard rather than down to a price. It has a large elliptical speaker, and the absence of a loudness control doesn't seem to detract from the sound quality.

The HMV 1122 cost £21 plus PT in 1950, which equates to about £900 in todays terms, so had HMV thought a loudness control would have made a material difference to sound quality, I'd have thought they'd have had ample headroom to fit it, for the small additional cost. Seems odd that the Stella 105U - costing half the price - has a loudness tap, so it seems not to have been a cost consideration, but a technical judgement by the designers. My Bush AC41, (1954) - another 'woodie with a large speaker, and also with a tone control but no loudness tap - would be a close second to the HMV in terms of sound quality and attractiveness, though such opinions are inevitably subjective rather than objective.

Oh well, the question has been answered - no, it isn't possible to simulate a loudness tap by fitting external resistors!

[Alistair D]

An alternative solution from a recent thread on here.

https://www.vintage-radio.net/forum/...d.php?t=126515

Al

[David G4EBT]

Thanks for pointing that thread out Al - I'd missed it completely.

A very nifty solution!

[Herald1360]

Originally Posted by David G4EBT

As will be seen, the 105U uses a 500k log pot, tapped 10% down from the 'hot' end, at 50K. Why not then, use a 1 meg pot, with a 100k and 900k resistor across it (910k would in practice be the nearest a 'preferred' value), to form a tap? Certainly on a 3 terminal pot, if a 500k one is required and not to hand, a 1 Meg with a fixed 1 meg resistor in parallel to simulate a 500k pot works fine.

The parallel network has no first order effect at all on loudness, it's just a parallel load to the volume control on the detector. If the detector is fairly high impedance it will provide a slight but fixed change in the frequency response.

Couple of plots attached for the tapped pot and parallel circuit cases.

Both pots at same relative settings- you can see level response on the parallel network case at same as LF attenuation of tapped pot. The tapped pot shows increasing levels of relative treble cut as it is turned down.