I had the chance to spend some time with the Wave Kinetics NVS table this weekend to do some measurements and analysis. I was most impressed by the metal work and overall mechanical design as well as the platter and bearing, which appear to be top notch. That said, the electronics and the actual performance was rather shocking, especially at this price point. Looking at their website and the review is even more shocking, given the specs they flout and how far off they are from actual measurements.
My first surprise was when I opened the lower cover that houses the motor. The motor is an Anaheim Automation BLY172D-24V-1000 ($102) with some minor modifications from the stock offering (BLY172D-24V-2000). The stock motor has 8 poles, the modified version in the NVS has 24 poles. The motor is packaged in an standard NEMA 17 footprint (42mm square) and is commonly used for industrial purposes. They refer to this motor as "laboratory-grade" in the PF review (I can't believe how naive reviewers can be). The motor also has an iron core for the stator windings which means it will have cogging. I don't think I've ever seen a DD table that didn't use a coreless motor, and cogging is the primary culprit with iron core motors and why they shouldn't be used. The motor uses an optical encoder for feedback and speed control, but it only has 2500 PPR resolution, similar to the VPI HW40 which produces equally bad results.
I measured the speed accuracy and stabilty directly off the encoder which will not have any record eccentricity or other defects to skew the results. At 33 1/3 RPM, the encoder should produce 1388.88888Hz, but ran slightly high at 1389.2Hz (+0.0224%), a long way from 1PPM.
The speed stability was quite poor actually, measuring ~0.1625% as shown on the polar plot:
View attachment 118990
There are 24 distinct "lobes" where the magnetic rotor passes each pole of the stator which causes cogging and speed instability. 1PPM?
The controller uses a Motorola MC56F80 16 bit fixed point DSP which was designed specifically for motor control, but it's limited resolution and math capabilities does not make it suited for high end audio IMHO. The display uses an Arduino Uno to drive a TFT color LCD shield which are commonly used by hobbyists. The interconnects between the controller and the table use a pair of CAT5 cables with RJ45 connectors; sensible enough for the control lines but woefully undersized for the motor as the windings are just over one Ohm and require considerable current. It leads me to believe the controller was designed for other purposes, possibly as a development kit or demo PCB.
The metal work, bearing and platter are most impressive and I could understand where that could take a lot of time and effort to achieve. But assembling a team of aerospace engineers etc. to do this kind of motor drive seems a bit incredulous. The 1PPM spec really strikes me as odd for a number of reasons: It obviously does not measure anywhere close to that and secondly, what would they have used to produce such a number in the first place? I've not seen any instrument or software that could measure W&F or speed accuracy to 6 places right of the decimal point. One of my criticisms of mfrs is that they sometimes appear to pull numbers out of the air for specifications and are never able to back them up with actual measurements or data; unless someone can post some rationale for these specs, I'd have to conclude that WK are doing the same thing here.
