Shortening the path

December 20, 2016
 by Paul McGowan

Over these last few days, I've been writing a very long series on the path to understanding and appreciating volume controls in preamplifiers.

Without hesitation I would maintain their design and implementation remains one of the bigger challenges in high-end audio when sound quality is the primary concern.

To date we have covered many ways to control level: pots, stepped attenuators, light dependent resistors, and transformers. All have one thing in common. They require extra parts in the signal path.

The best volume control would be no volume control.

As I have written over the years, we designers can rarely make things better—though Power Plants would refute that. Generally, we do our best to cause less harm and the ubiquitous volume control is a great example. No volume control element of any design can improve the signal. Best you can hope for is as little damage as possible. But what if we could eliminate the extra circuitry in the signal path altogether?

Control the volume without an additional element?

This is where an idea I came up with years ago comes into play. Instead of adding a pot, or attenuator in series with the signal, would it be possible to simply change the gain of the amplifier within the preamp? In fact, it is, but it turned out that's not an easy thing to do well.

All active preamplifiers have an amplifier at their heart. Generally low gain, these amplifiers are the core of any device.

Remember back to the first post in this series? There I described simply the internals of a preamp.

  • Input selector
  • Volume/balance control
  • Gain stage

The question I had asked myself was simple. Instead of a three element design might it be possible to eliminate the middle and create a two element design?

  • Input selector
  • Gain stage

In so doing we obviate the problems associated with the volume balance control. Because if you don't have one to muck up the sound, you've scored!

I'll delve deeper starting tomorrow.

Subscribe to Paul's Posts

11 comments on “Shortening the path”

  1. "To date we have covered many ways to control level: pots, stepped attenuators, light dependent resistors, and transformers ....."

    About transformers used as volume control, the only thing you have said is that they are not to your liking, and nothing else.

      Unless I missed another post of you about the subject, which I do not think so.

      1. Actually, the information on the VC transformers given yesterday I'm sure many of us know, what I'm not clear is that if these, introduce colorations or other aberrations in the signal, but on the other hand transformers that handle low voltage signals in audio , Such as transformers to deliver balanced digital signals on transports from well-known manufacturers, suggests that the losses are not formidable.

        I understand that some DAC manufacturers convert the digital coaxial signal into a balanced one using transformers, so apparently the losses are not significant.

  2. So what about high quality lossless digital volume control with an FPGA chip? It is the single most popular solution in the UK since it went mass-market about 2 years ago.

      1. The only need for analogue volume control is with analogue sources, which is probably less than 0.1% of the audio population. Anyone into classical (like me) dropped vinyl for classical 30 years ago.
        So it follows that the vast majority of volume control is in the digital domain, as I did with my PSA DAC and now do (sometimes) with another DAC.
        To avoid the problems and cost of analogue volume control (and Audiomano points out it is more problems than answers), you can apply ADC to your turntable, as Linn do with their digital system or get a separate ADC converter.
        Some DAC/pre manufacturers are now including a line input or even a phono input on the back of their DAC, and either applying ADC or having a separate analogue volume control.
        The fact that you can now get a near perfect lossless volume control in a DAC/pre headphone amp using FPGA that goes in your trouser pocket and costs $599 is (and I've never used the word before) a real gamechanger.
        I have a nice collection of records, but the advances in digital (and PSA have taken a lead in some areas) make them feel an increasing burden to just sitting down to enjoy listening to music.

        1. @stevensegal

          You hit the nail on the head!
          For me too, if I understand your statement correct, any volume control other than done with a FPGA is archaic.
          As I mentioned a couple of months ago I've build a phono pre using a FPGA and do everything in the digital domain. This phono pre is connected to the DACs I've in use at my home and works better than everything else I have had in the chain before.
          My guess is that there are a lot of vinyl lovers and that they avoid any kind of digital like the devil avoids the holy water.
          A manufacturer who wants to adress these people too, might have some problems selling their products to them.
          Maybe you remember my post from Sunday Dec. 18th when I spoke about the PASS LABS XP-30 preamp and I said, that if ever I would integrate a preamp in my system the XP-30 would be one of my first choices. Well I still found no preamp that improved the performance of my stereo not even the XP-30.
          The next thing I am really waiting for is a truly veritable digital power amp. I do not have those Class D designs in mind, when I say veritable digital power amp. I bet this will come in the following years and it will be - I use your words - a real game changer.

          Regards

          1. My second - pure analog (vinyl source only today after having removed my Revox B760 FM-tuner because of unacceptable sound quality of the radio stations today - and the tape deck of course) - stereo system features a transformer based passive preamp (silver wire) since ten years now. I haven't yet found a preamp surpassing this tiny box. And I enjoyed a lot of preamps for some decades starting with a Burmester 808 MK 3.

            1. I firmly believe that the only correct path for VC is based on transformers, for me it is frustrating that having bought preamplifiers of well-known brands that cost a lot of money (of valves or SS) use Pot. VC that soon show a malfunction either by noise or Inexact volume at different gain points, this is not only a matter of cleaning but the concept, I also have preamplifiers with equally imprecise electronic VC. My goal is to have a pre-amp with VC that uses transformers. I hope that my worries about this matter will end.

  3. Looking more carefully at the Gilbert cell, two things struck me. One is that there are a lot of additional parts. So much for less is more. Second, the notion that the signal path does not include resisters or potentiometers my be an illusion. This is because since the gain is controlled by the bias voltages, the potentiometer must be in the emitter to collector circuit to control it instead of the emitter to base circuit (source to drain versus gate to drain for FETs. ) The only way around this I can see is to adjust bias voltage with a Variac and a well filtered power supply output.

    It should be noted that some of the most sensitive scientific instruments in the world in the past and probably in the present use potentiometers. So does military hardware. Those are built to much higher standards in critical applications than you'd see in even high end consumer audio.

Leave a Reply

© 2022 PS Audio, Inc.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram