Preface

This book aims to give a comprehensive account of the electronics needed to extract the best possible signal from grooves in vinyl. This is timely, as what is called the “vinyl revival” is still in full swing. Vinyl reproduction presents some fascinating technical challenges; the signal levels from moving-magnet cartridges are low, and those from moving-coil cartridges lower still, so a good deal of low-noise amplification is required. In the moving-magnet case this is much complicated by the high inductance of the cartridge. RIAA equalisation is required for a flat response, and achieving this both accurately and economically is a major study in itself. Vinyl produces large amounts of subsonic noise, and this has to be filtered out both effectively and unobtrusively to prevent bad things happening in amplifiers and loudspeakers.

Some of the basic material was published in my Small Signal Audio Design (2nd edition). When the time came to consider updating the book, there was the problem that its length was already pushing the limits for practical publishing. I have heard it said you should never write a book you cannot pick up with one hand. This issue was solved by splitting off everything related to vinyl reproduction and adding a great deal of new material to make a new and separate book focused on vinyl electronics. Here it is, containing more than three times as much text and more than three times as many illustrations; most of the content is new material. There are new chapters on phono amplifier architectures and ultrasonic and scratch filtering. The new chapter on subsonic filtering is, I think I can say without fear of successful contradiction, the most comprehensive account of it ever published. I have tried to make it easy to dip into this book—not everyone will want to read it straight through. A few important concepts are therefore explained more than once, when they are particularly relevant. I hope you will not accuse me of hesitation, repetition, or deviation.

The focus is on the analogue domain, where the processing is done with opamps or discrete transistors, usually working at a nominal level of a volt or less. There are good reasons for this. While you can do almost anything in the digital domain, you first have to get the signal into the digital domain. Since the nominal signal level from a moving magnet is only 5 mV rms and the output from a moving-coil cartridge generally much lower than that, clearly you are going to have to quite a lot of analogue amplification before you can apply it to the input of an A-to-D converter. Most of them require between 3 and 6 V peak to peak for maximum digital output, usually called FSD. Also, the subsonic disturbances that come from disc warps and irregularities can occur at frighteningly high levels, barely 20 dB below maximum, and it is an excellent idea to remove them before they erode headroom in the ADC. Only 18 months ago I was involved in a telemetry project where some quite sophisticated analogue processing, including allpass filters for time-compensation, was required before the signal could be accurately digitised and transmitted.

In the pursuit of high quality at low cost, there are certain principles that pervade this book. Low-impedance design reduces the effects of Johnson noise and current noise without making voltage noise worse; the only downside is that a low impedance requires an opamp capable of driving it effectively, and sometimes more than one. The most ambitious application of this approach so far has been in the ultra-low noise Elektor 2012 Preamplifier.

Another principle is that of using multiple components to reduce the effects of random noise. This may be electrical noise; the outputs of several amplifiers are averaged (very simply with a few resistors), and the noise from them partially cancels. Multiple amplifiers are also very useful for driving the low impedances just mentioned. Alternatively it may be numerical noise, such as tolerances in a component value; making up the required value with multiple parts in series or parallel also makes errors partially cancel. This technique has its limits because of the square-root way it works; four amplifiers or components are required to half the errors, sixteen to reduce them to a quarter, and so on.

RIAA equalisation is a prime example of a requirement for non-standard component values (crossover design and filter design in general being the others in the audio business). As a general rule, in a series-feedback RIAA network only one component can be a preferred value from the widely used E24 resistor series. While resistors are freely available in the E96 series, if you are faced with a required value that is effectively random, you will do much better by using two E24 values in series or, preferably, in parallel. The nominal value will on average be three times more accurate than a single E96. With three E24 values combined the accuracy of the nominal value is further improved by ten times compared with a single E96. This may appear profligate with components, but resistors are cheap. There is another advantage to the multiple component approach that is less obvious; the effective tolerance is reduced. By how much depends on how much the values vary; the maximum improvement with 2xE24 is √2 times and with 3xE24 is √3 times. The problem is more difficult with capacitors as they not usually available in as many values as the E24 series. Finding the best 3xE24 values is a non-trivial problem, and I have been much helped in this by the work of Gert Willmann, which he has generously shared with me. You will find much on his contribution in the body of this book.

There is also the Principle of Optimisation, which may sound imposing but just means that each circuit block is closely scrutinised to see if it is possible to improve it by doing a bit more thinking rather than a bit more spending. One example is the optimisation of RIAA equalisation networks. There are four ways to connect resistors and capacitors to make an RIAA network, and I have shown that one of them requires significantly smaller values of expensive precision capacitors than the others. This new finding is presented in detail in this book, along with related techniques of optimising resistor values to get convenient capacitor values.

And now what there is not. I have no time for faith-based audio, so there will be no discussion of esoteric components with insulting price-tags that actually achieve nothing. There will be no truck with the anti-science of cables that are alleged to know which way the information is supposed to be flowing. Passive RIAA equalisation is not recommended; it is thoroughly deprecated with good reason. I have spent more time than I care to contemplate in double-blind listening tests—properly conducted ones, with rigorous statistical analysis—and every time the answer was that if you couldn’t measure it you couldn’t hear it. Very often if you could measure it you still couldn’t hear it

You may be surprised to find that there is not a great deal in this book about balanced phono preamplifiers. This is because a magnetic cartridge is a floating winding—it does not have a centre-tap or a ground reference until you connect it to an amplifier. There also should be no unwanted currents flowing in the ground conductor unless something is miswired. A balanced input therefore gains you nothing—a view that was substantiated in long discussions on the DIYaudio forum—and loses you some signal-to-noise ratio because of the extra electronics required for balancing.

So much has been added to make this book that it is difficult to summarise it, but the new material includes:

• • More on resistor and capacitor selection for awkward values
• • Preamplifier architecture and interfacing with phono amplifiers
• • Optimal use of capacitors in RIAA equalisation stages
• • Switched-gain MM amplifiers that retain accurate RIAA
• • Hybrid phono amplifiers
• • Noise in balanced MM inputs
• • Noise weighting: A-weighting and ITU-R weighting, practical filter designs
• • More on discrete-transistor circuitry for phono amplifiers
• • Butterworth subsonic filters from 1st to 6th order
• • Elliptical subsonic filters from 3rd to 6th order
• • Elliptical subsonic filter optimisation
• • Subsonic filtering by VLF crossfeed: The Devinyliser
• • Butterworth ultrasonic filters from 2nd to 6th order
• • Combining subsonic and ultrasonic filters in one stage
• • A wholly new chapter giving six practical phono amplifier projects

If you have made the decision to use vinyl as your music-delivery medium, I believe this book will help you get the best possible audio off your discs. Whatever you think of vinyl, there is no doubt that it presents some fascinating technical challenges. The circuitry presented here is not merely good enough for home construction but sound enough for manufacture; I have done a lot of that. All measurements were performed with an Audio Precision SYS-2702.

A good deal of thought and experiment has gone into this book, and I dare to hope that I have moved analogue audio design a bit further forward. I hope you find it both useful and enjoyable.

To the best of my knowledge no supernatural assistance was received in the making of this book.

All suggestions for its improvement that do not involve its combustion will be gratefully received. You can find my email address on the front page of my website at douglas-self.com.

Further information and PCBs, kits and built circuit boards of some of the designs described here can be found at: www.signaltransfer.freeuk.com

Douglas Self
London, December 2016