Stylus Drag

My Final Audio TT has a 19kg platter and very sophisticated motor drive, basically a huge AC motor driven by oscillator preamp and power amplifier which reconstructs sine and cosine waves for the motor and provides precise adjustability for both speed and torque applied to the motor.
As Halcro alluded to above when setting speed, I have to do that with the stylus in the groove.
Once set though it is very stable. In recent tests with the Timeline and KAB there was no speed change from heavily modulated passages both inner groove and outer groove.
Arm and cartridge were Naim Aro/Koetsu Black tracking a 1.8g

Aoliviero -
I can only surmise that the power supply and feedback mechanism of the Raven AC is inadequate if it cannot hold the speed. I certainly wouldn't be happy spending that sort of money on a TT and it not being able to hold the speed accurately whilst playing. Have you asked the manufacturer ?

Tonywinsc,
The closed loop open loop is not as simple as implied. They are inextricably linked to the motor technology used, AC or DC.
DC motors require a feedback or speed control system to actually run. They will not run without it unless you use a computator type DC motor.
Now it is my understanding that AC motors only require the frequency to be locked in order to run at the correct speed.
The fundamental difference as I understand it is that when changes in load, eg stylus drag variations, occur, the DC motor does not respond and the speed will vary - hence the need for a feedback or speed correction system.
However when there is a change in load, eg stylus drag variation, on a synchronous AC motor, the AC motor will self correct as the motor speed is determined by frequency, not load.
Now most, but not all direct drives are using DC motors, the big exception being the Denon DP100. There appear to be 2 camps on how the speed is maintained. I believe the L07D & Halcro's Victor have soft recovery speed control, whereas others such as the Technics have a faster recovery. I think this explains some of the sonic differences between these DD's. In my view the LO7D and Victor 101 are simulating an AC motor in how they respond to load variations.
You cannot compare DC and AC motor driven turntables as such because all DC motor driven TT's have speed controllers. Unfortunately in my view most TT's that use AC motors have little or no speed control and rely on incoming mains frequency being stable which is simply not true. Examples of this are the VPI Classic series, which even at the US$6000 price point, they only have 1 capacitor between the mains and the motor to split the phase. There is no provision for mains frequency fluctuation.
The only conclusion that I can see is that mass in the platter will assist in maintaining constant speed under load variation in both AC & DC motor applications.
I note that in a recent post Atmasphere checked a cutter lathe for speed accuracy whilst cutting. The formula here was massive motor, way in excess of domestic TT's coupled with 'direct drive' via driveshafts not connected to the bearing and by implication a massive gearing down from motor speed to platter speed.
Having overhauled and upgraded my crossovers recently I have been re-listening to a lot of piano music lately and it is clear that, in my particular TT/system, there is no instability induced by stylus drag, but there is on eccentric records, this now seems to be the larger issue, as you have suggested. Hope your needle file is still doing the business.