The DD motor has noise of 0.004, I believe measured a the headshell. VPI has made TTs with belts but nothing as quite as this and the speed is super stable.
I think you are referring to W&F; 0.004% is what HW has claimed in numerous videos, but their official specs have it at 0.008%. Both those numbers are dubious and need some explanation from VPI (i.e. what method they are using to measure it). HW says you can’t use a test record you have to use the software that is supplied by Elmo Motion Control who make the controller. I have the software, and the table measures much worse than 0.008%, more like 0.04~0.07%, which isn’t bad, but not much better than a well tuned BD table. Using a test record measures nearly the same. If you look at the output of the magnetic encoder ring they use for feedback, it looks even worse. The spec from the mfr of the ring encoder (RLS in Slovenia) is ~0.02% (system error=magnetic encoding error + sub-division error + eccentricity at 50µm), so even if the motor/controller were perfect, it’s difficult to see how W&F could drop below that.
As far as speed accuracy, the encoder they use produces 2560 pulses per rev; the controller locks the speed to a count taken on 1 second intervals. For 33.333 RPM, that equates to 1422.222 counts; the controller can only work with integer counts, so they round it up to 1423 which yields a platter speed of 33.351. Not bad, but not as good as I get with on a DIY belt drive using the RR tach and a Condor PSU (33.333 ±0.005 RPM).
I believe the reason people question the design choices and even the published specs for VPI products, is because in the past, they have made misleading or inaccurate statements, some of them so bad, they are outright laughable.
FYI, very little of the mfring is done by VPI. Most if not all of the machining is done by MDI mfring in Lakewood NJ; VPI assembles the components.
