Graham Slee Revelation Technology

Paralleled op-amps
This is a very old technique in which two (or more) op-amps are connected in parallel, either to reduce circuit noise by 3dB, or to increase the output current by a factor of 2. In this case, it also increases the quiescent or no-signal standing current in the output stage to 2 op-amps worth (obviously). This means that for more of the signal, the output is operating within a "class A" envelope - especially the subtle signal information that is responsible for the "live" feel and the accuracy of the stereo image - which means it is not affected by output stage cross-over distortion.

Frequency Tracking Matched Time-Constant By-pass Capacitors
The trouble with so many audio designs is that:-

1] although great attention is paid to high frequency stability (with some exceptions...), there is little or no understanding about low frequency stability. Looking back at some early (pre-renaissance) valve circuits, it can be seen that the time constants between local power supply decoupling and the pass frequencies of the stage, are matched. If they were not, the power supply for that section will gyrate in sympathy with the signal up to the point the decoupling becomes effective. Therefore that gyration feeds back into the signal and distorts it, and in the worst case will lead to self-oscillation. These frequencies are in the sub-bass, and if not Time-Constant matched, exhibit themselves as a "muddy" sound. (all our products have matched time-constant de-coupling)

2] It has become common practice to by-pass large capacitors with smaller ones because, very often, large capacitors have poor performance characteristics at mid to high frequencies. However, it would seem a mystery as to how the by-pass capacitor value is chosen? Many articles on the subject simply use a rule of thumb like "all electrolytics are bypassed with 100n polyester capacitors" (or similar). However, it is so easy to see using conventional engineering math, that the bypass capacitor can be chosen to start at a particular frequency from the impedance of that part of the circuit! Now, by using the argument in 1] above, the decoupling capacitor can also be bypassed such that it shares the same time-constant. Without this, the circuit is "reinforced" above a particular frequency, but the power supply to that section is not, so that the power supply becomes "loose" enabling it to again feed back (at the "rule of thumb"??? frequency). The result without is the listener is aware of "seams" between particular frequency ranges. The result with Frequency Tracking Matched Time-Constant By-pass Capacitors is a totally seamless performance from both subjective and objective observations.

Accuracy in Grounding
I know it will be hard for some subscribers to the hi-fi press to conceive that in many design offices circuit grounding is not fully understood, but if one poses the question "why is it done in that way" the answer never seems very tangible. One manufacturer will ground a circuit at one point, whereas another manufacturer will be adamant that another point is the correct place. Looking back again to pre-renaissance valve designs you will see some disagreement, but not as "worlds-apart" as it seems to be today.

Pseudo-Differential Ground Sensing Inputs
When is a ground not a ground? Answer: When it has impedance!

If you take a journey from your (MM or MI) cartridge to the phono stage input by the earthy (ground) conductor, you will see on your travels:

1] the coil wire is attached by solder to the inside of the cartridge pin.

2] On the outside the cartridge pin has a rolled spring contact gripping onto it.

3] this contact is soldered to the thin arm wire which travels along the arm tube, loops out to pass the arm bearing, and then down the middle of the arm mount where it is attached by

4] solder to the arm plug pin.

5] The arm plug pin is in contact with the arm socket and

6] to the arm socket is soldered the screen of the arm cable.

7] at the phono stage end the screen is soldered to the body of the phono plug and 8] the phono plug body grips the phono socket.

There are 8 contact points. 3 are pressure contacts and 5 are soldered joints. All 8 exhibit some impedance, and also the metal of the wire/cable exhibits a very tiny impedance. You may think otherwise about a soldered joint but it is a joint of disimilar metals, and so is a thermocouple. A thermocouple develops a voltage and for a potential difference to exist there must be resistance, and the AC form of resistance is impedance.

Therefore there IS a voltage drop between cartridge and phono stage input which allows the voltage at the lower end of the cartridge to float above ground - there is ground current.

Those who don't understand these finer points favour balanced wiring, but unless that is done to an exacting standard, it simply won't work. It is also very inconvenient and obviously expensive in having this work done with only a suggestion of a promise that you just maybe get an increase in performance - if you're lucky. And you will be stuck with a balanced input phono stage unless you revert back to conventional turntable wiring.

The Revelation features Pseudo-Differential Ground Sensing Inputs which require no modifications to conventional turntable wiring whatsoever.

What a Pseudo-Differential input does is to float the "ground" input above ground by just more than the lumped impedances described above. A conventional grounded input does not have much of a common mode noise rejection ratio (CMRR). A balanced input has infinite CMRR, but only provided that the wiring balance is perfect. A Pseudo-Differential input has CMRR = Gain, which at hum frequencies is 58.5dB (due to equalisation), or nearly 1,000:1.

It does not mean that you can get away with bad or corroded contacts along the ground path as that could exceed the expected lumped impedance, and hence hum would result as always. Also, the arm "earth" is still required as that is the only form of screening for the cartridge, and arm wires contained within.

However, it does mean that the smaller artefacts of induced mains frequency in a properly conventional-wired turntable is ignored by the phono stage, meaning that the signal is no longer modulated by it and the resultant cleaned-up signal reveals far more musical information heard in the form of greatly improved timbre. Also, as one channel could exhibit a greater ground impedance to the other, that is now ignored and therefore the stereo image is much improved.

At high frequencies there is less gain (due to equalisation), but CMRR is still assisted by approximately 22dB. Therefore, although it will not cancel large radio signals which a badly screened input could pick up, any tiny radio frequencies present in a well screened input will be ignored.

High Quality Audio Electrolytics
Electrolytic capacitors are two layers of aluminium foil rolled-up with an oil-soaked paper between them as dielectric, with an oxide film anodised onto the foil as polarised insulator. You may have observed the behaviour of the acid electrolyte in a car battery. The behaviour inside an electrolytic capacitor is similar but on a vastly smaller scale. There is movement inside the dielectric due to the flow of AC current (the signal). The paper/oil composition within an audio electrolytic is formulated so that these movements do not introduce distortions of their own - whereas with a general purpose electrolytic the behaviour is not predictable. Generally bass frequencies are more prone to distortion because of the greater energy mass contained in them. As a phono stage has 10 times more gain (due to equalisation) at bass frequencies it is particularly important to place the emphasis on accuracy here.

Fast Pulse Polypropylene Capacitors
Polypropylene is a very inert plastic. It prevents almost anything sticking to it. Hence the dielectric film cannot make intimate contact with the foil and therefore there is less dielectric absorption to slow the propagation of an AC signal through them. The latest generation of Polypropylene film capacitors are made to be very fast indeed - the equivalent in amplifier terms of 1,000 volts per microsecond! However, they are not available in the values to suit the impedance of most modern phono stage equalisation networks. The Revelation uses our high impedance EQ network (featured in all our phono stages) which we specially developed to exploit this property.

RoHS Ready!
This design has been optimised to use RoHS compatible components and production units already feature RoHS compatible op-amps, many other RoHS approved components, an RoHS compatible board and are assembled using Lead-free solder.

<< Back To Graham Slee Main Page - Table Of Contents