Understanding
Turntable motors: why 90° is not enough
The motors on Transrotor turntables are synchronous motors with two windings. To make them run smoothly, they are fed two sine voltages of the same frequency, shifted by 90° relative to each other. That is the rule you find everywhere. It is correct, but it assumes a perfect motor, and there is no such thing.
Why two windings?
A synchronous motor needs a rotating magnetic field: the rotor locks onto it and follows at the same speed. A single winding on alternating current does not create one. It creates a field that pulses in place and reverses every half cycle, which amounts to two rotating fields of equal size going in opposite directions. The rotor locks onto one, the other drags on it. The motor starts badly, when it starts at all.
The second winding fixes this. It sits 90° from the first around the rotor (90 electrical degrees; on a motor with several pole pairs the mechanical angle is smaller, but the reasoning is the same). Feed it a current shifted by a quarter of a cycle and the two pulsating fields combine into a single one of constant amplitude, rotating. 90° in space, 90° in time: that is where the number comes from.
On a motor connected to the mains, the shift usually comes from a capacitor in series with one winding. With a two-channel electronic supply it is calculated: a sine wave, and the same sine wave delayed by a quarter of a cycle.
When the angle is not exact
If the shift is not exactly 90°, or if the two currents differ in amplitude, the field is no longer a circle but an ellipse. An ellipse breaks down into a circular field rotating the right way, plus a weaker one rotating backwards. The first drives the rotor. The second passes it at twice synchronous speed and produces a torque that oscillates at twice the supply frequency.
That alternating torque does not move the platter forward. It shakes it slightly, and the current that produced it has heated the motor for nothing. A heavy platter filters out much of it, though not necessarily all: some of it may be transmitted through the belt and the bearing.
The theoretical 90° is not necessarily the motor's reality
The calculation assumes two identical windings, perfectly orthogonal, in a uniform air gap. A real motor departs from that. The number of turns, the resistance and the inductance are not exactly the same from one winding to the other, the windings are not wound or placed to the micron, the air gap is not perfectly constant. Each tolerance is tiny. Together they shift the angle that would give a perfectly round field.
The ideal angle is therefore a little below or a little above 90°, by a few degrees. It is not the same for two motors of the same model, and for a given motor it can move with load and temperature. A supply whose phase is fixed at 90° suits an average motor. It cannot suit every motor equally well.
Not to be confused with the offset between two motors, covered in the article on synchronisation: here we are talking about the phase between the two windings of a single motor. On a dual-motor turntable, each motor has its own optimal angle.
How to find the right angle
It cannot be calculated, and the motor's nameplate does not give it. It has to be searched for: vary the phase and watch an indicator. The most accessible one is the current drawn. At equal load, the motor draws the least current when its field is roundest, because no share of it goes into building the stray field. The current minimum marks the angle you are after.
The difficulty is measuring it cleanly. Between two neighbouring angles, the current changes by only a few tenths of a percent. If the measurement does not cover a whole number of cycles, the truncation error can exceed the effect being looked for.
One caveat: this is an electrical indicator, not a listening test. It says the motor is working cleanly, not what the ear will make of it.
How Isostase does it
Isostase generates the two voltages of each motor separately, with an adjustable phase. It starts at 90°, tries a slightly larger angle and then a slightly smaller one, measures the motor current over a whole number of cycles and moves towards the side where it drops. The step, 2° at first, goes down to 0.7° and then 0.25° as it closes in on the minimum; once the point found is bracketed by higher values, Isostase fits a parabola through the three measurements to aim at the bottom. As a safety limit, the deviation from 90° is capped at 15° either way.
On a dual-motor turntable, the search is done motor by motor, with the other one switched off during the measurement; the two are then adjusted relative to each other. The search can be restarted from the screen, for instance after replacing a motor.
Pre-series units of the board are being measured and validated: current-versus-phase curves will be published as soon as the first tests are done.
Related videos: Motors and Belt