Platter Speed

Olydyne Isostase. What we measured. These measurements were made on a Transrotor Zet 3, driven by a pre-series Isostase, first with one motor, then with two. We compared several Isostase configurations with two original Transrotor power supplies, called A and B, and kept anonymous. The original supplies were measured exactly as found, with no adjustment. Same turntable, same conditions, throughout. The test signal is the three thousand one hundred and fifty hertz tone on The Ultimate Analogue Test L P. A QuantAsylum Q A four-oh-three records it for forty seconds, about twenty-two revolutions, four times per configuration. The instantaneous frequency of the tone gives the speed. After an A E S six type weighting, it also gives wow, from half a hertz to six, and flutter, from six to two hundred. This is a weighted R M S measurement, not an A E S six, two thousand and eight measurement, so it should not be compared directly with a datasheet figure. First, speed. With regulation, the platter stays within zero point zero one three to zero point zero two five percent of nominal. The original supplies, measured as found, run zero point six three percent fast for supply A, and one point four seven percent fast for supply B when cold, zero point nine seven once warm. One and a half percent is about a quarter of a semitone. Without regulation, one motor at a fixed frequency runs zero point four eight percent fast. The fine frequency adjustment can bring that error back to zero, but a fixed setting does not follow drift, such as supply B's change with temperature. That is what the regulation is for. The one surprise: at exactly fifty hertz, this turntable runs about zero point four percent too fast. The optical sensor reads thirty-three point three three R P M at about forty-nine point eight hertz. Here, forty-nine point eight zero three on a Keithley D M M sixty-five hundred. That is a static mechanical ratio, pulley, belt and slip, specific to this turntable. A supply with an exact frequency does not guarantee the right speed. Only measuring the platter does. Wow and flutter tell a different story. All weighted totals lie between zero point zero four five and zero point zero five one percent. Isostase does not measurably degrade wow and flutter compared with the original supplies. It does not improve them either: the differences are too small to interpret. That is expected. Regulation corrects the average speed, while wow and flutter come mostly from the record, the cartridge, the arm and the platter. The output frequency is derived from a crystal reference, calibrated on each unit. On this unit, a Keithley D M M sixty-five hundred reads one point zero zero zero zero two one kilohertz before calibration, twenty-one parts per million high, and exactly one kilohertz after. One caveat: the multimeter's own uncertainty is one hundred parts per million. The two readings show the shift, but do not prove one part per million absolute accuracy. On the oscilloscope, the two motor phases are regular sine waves, about ninety degrees apart. Phase is adjusted automatically, motor by motor. Fifty-two point seven volts peak to peak at full level, thirty-seven point eight at seventy percent: about seventy-two percent. With the motors connected, the tops of the waves are slightly thick. Replace the motors with one hundred and fifty ohm resistors, and amplitude and frequency are unchanged, but the trace is thin. The disturbance comes from the motors, not from the output stage. Isostase is not an audio amplifier, so distortion is used here as an indicator of the purity of the output voltage. Into one hundred and fifty ohm resistors at thirty-three and a third, T H D plus noise is zero point zero four seven percent, with a signal to noise ratio of ninety-four decibels. With the motors connected, it rises to zero point six three percent, and the noise floor rises by thirty to forty decibels. Only the load has changed, so this excess comes from the motors. Start-up and stop, with a twelve kilogram platter. At thirty-three and a third, nominal speed within half a percent is reached in seven seconds with one motor, five seconds with two. At forty-five, eight and a half and five point eight seconds. At seventy-eight, twenty-six seconds with one motor, fourteen point six with two. Optical regulation engages at the end of the ramp and settles in three seconds. What these measurements do not show. One turntable. One unit of each original supply, measured as found. Absolute wow and flutter reflect the whole chain, and no independent instrument was used. Speed was cross-checked with the optical sensor and a one kilohertz track, but not with a quartz stroboscope. And nothing here says what the ear hears. With Isostase, speed stays between zero point zero one three and zero point zero two five percent of nominal. The original supplies, measured as found, are off by zero point six to one point five percent. And there is no measurable degradation of wow and flutter. Methods, figures and limits are published on olydyne dot com.