Technics SP10 artistic blue

This is my SP10 page on ICL's Eject server: Feb 22nd, 2010
You are viewer number  
When Jean's Radio & T.V Spares closed forever at the end of 2006, I acquired among a huge pile of components that she gave to me, a new-old-stock motor for the legendary Technics SP10 turntable.
To become part of a turntable, the motor needs a 3-phase drive system, and a quartz phase-locked-loop servo control system. This page describes the design and assembly of such a drive system, and a web-forum here is dedicated to questions and answers concerning it.

Technics SP10 motor armature
September 2009: I have now received from Shaun O. in Cape Town, an SP10 chassis and platter - thank you Shaun. This will allow me to design the control system specifically for the original platter mass and inertia. While many will choose to locate their updated control circuits, outside the SP10 chassis, I intend to make the circuit boards so they will physically fit the original mounting structures, for those who want to keep them inside the SP10.

Because of its high grade of engineering and the fact that it incorporates the main bearing, the MJX-12A motor assembly used in the SP10 makes an ideal starting point for a home-made turntable of high quality. On downside, being of the direct-drive principle, an elaborate control circuit is required.

The control circuit of the original SP10ii comprises 3 main circuit blocks:

  • a power stage to energize motor coils
  • a frequency divider to generate a reference clock
  • a phase-locked-loop servo control system
  • The motor itself contains several independent windings, not only the 3 main drive coils. The 3 phase drive signal for the motor is in fact generated within the motor itself - it is both a motor and a 3-phase alternator all in one. The alternator signals are amplified and returned to motor as its 3-phase power supply. The various windings in the motor are thus 3 main drive coils, one for each phase, 3 generator windings, an exciter winding to energize the 3 generator windings, and finally an FG (frequency generator) winding by which the speed of rotation is measured. Because of the fixed physical relationship between drive coils and generator coils, the phase relationship of 3 drive signals is always correctly maintained, no matter what the rotational speed. As far as designing a control system is concerned, the motor can be treated as an 'ordinary' commutated DC motor.

    Power supply and drive circuit.

    The drive circuit consists of 3 identical power amplifiers. L165V PentawattTM power op-amps are a convenient solution, in a small package. A discrete transistor version - using something like TIP31/32 transistors would work just as well. The L165 saves PC board space and eliminates a few design steps, and costs no more than the equivalent disscrete transistors and their associated components. A TDA2040 'chip-amp' and its siblings would be drop-in substitutes for the L165V, at a higher cost, with no real advantage.

    The required bandwidth of this stage is no more than a few Hz - actual drive waveforms for 33.3 rpm being 5.5 Hz, rising to 12.9Hz at 78 rpm. Being almost pure sinusoids (more on that later) they have no appreciable harmonic content, and amplifier stages require no remarkable performance characteristics.

    The power amplifier section was to be built on the same printed circuit board as the linear power supply to eliminate any PCB interconnects involving heavier current wiring. But after I obtained a real chassis, I discovered there was not enough space insude to fit all the circuitry in. So the original Matsushita arrangement of an outboard power supply box will be used. The motor draws only a few tens of milliamps during normal running, and several hundred milliamps briefly at startup.

    Linear voltage regulation is by ordinary LM317t/LM337t positive and negative regulators. 78/79 fixed regulators would be adequate, if slightly noisier, but the 3x7 series was chosen because they offer more versatile remote shutdown and current limiting. Any center tapped transformer giving something around 15-0-15 Vrms at 20VA or so is suitable. Some purists may even try running the system from four 12V lead acid accumulators.

    CAUTIONS & DISCLAIMERS: ANY wiring concerning the AC mains - 220v or 110v - is potentially hazardous. If you have no experience of such, seek help & advice, or make use of a commercial DC power supply. The low voltage DC side of this project is electrically safe, but since I cannot personally check what other people build, I will accept no responsibilty for any damage you might do to your turntable components through using any part of whole of my experimental circuitry.

    3-phase drive

    All so-called "brushless DC motors" are in fact electronically commutated AC motors. A simple and common approach is to use Hall sensors to detect rotational phase of the rotor and switch a fixed-level DC voltage to the appropriate armature winding.

    To achieve a smoother, pulse free drive, Matsushita derive a 3 phase sinusoidal drive for the motor by incorporating a 3 phase alternator within the motor itself. At the very low rotational speeds involved, only very low voltages would be generated by conventional induction designs. Moreover, the amplitude of the generated waves would be speed dependant. The Matsushita motor incorporates a set of three variable reluctance pickup heads - similar to tape recorder playback heads - that read the magnetic reluctance of a sinusoidally shaped mask, mounted on the rotor shaft. Being reluctance-sensitive pickups, they need some form of excitation, and this is provided by three additional coils, energised by a fixed frequency, fixed amplitude oscillator running at 50kHz. The 3 reluctance heads each pick up a 50kHz tone, amplitude modulated by a sine wave that exactly matches the rotational phase of motor. Since there are three pickups, equally spaced around rotor, a triad of constant-amplitude sine waves, 120° apart, can be derived by simple envelope detection. The phase of these is always correct in relation to drive coils, no matter what rotational speed.

    The pickup coil signals are amplified and put through a full-wave envelope detector. The low frequency sine components are passed through a variable gain cell - a PWM modulator - and thence become the input signal to the three power amplifer stages.

    Since the electrical phases of all three drive coils are always correctly maintained by the physical positioning of the pickup coils, the motor always starts, and always runs in the same direction. To allow for reverse rotation, the phase of any two armature winding can be interchanged (relatively shifted by 120°), which is easy to do electrically by FET switches.

    SP10 diect drive block diagram
    Block diagram: click for bigger.

    Although the motor itself is structurally a 3-phase synchronous design, when coupled with the power amplifier, which generates 3 x 120° phase-shifted waveforms of 3 phases, it behaves exactly as a conventional DC motor. The amplifier drives the motor at a speed dependant only on the magnitude of the drive voltage - there is no 'frequency' component whatsoever used in controlling the motor speed.

    A variable gain cell is used to control the amplitude of the 3-phase AC waveforms (the drive voltage), thus allowing speed control with a single DC control voltage. The gain cell can be implemented in many ways: as a variable transconductance amplifier (Akito Kaneta approach), a FET configured as voltage dependant resistor, an optical gain block, a multiplying DAC, or a digital multiplier (pulse-width-modulated). I chose the PWM method as it is cheap, needs no esoteric components, is straightforward to control with a microprocessor, and gives very good linearity.

    There is not a great deal of accuracy even in this section of drive system - the drive voltage does not determine the speed precisely, but only closely (as with any DC motor). If the control voltage were held absolutely constant, loading of the motor mechanically would cause it to slow; this is reason the why a servo is included. A dual servo system is used - a velocity feedback loop, and a phase feedback loop (PLL)