Direct Drive

Actually its not cogging.

The essence of the main difficulties with DD is derived from the fact that a direct drive requires a motor to run at 33rpm - the gearing is 1 to 1. In the old days ( 70's ) it was very difficult to control motors accurately at such a low speed.

The problem is pulsing.

If you run an 8 pole motor @ 33rpm there are effectively 8 poles per revolution ( of the record ).

For a non direct drive, say using the same 8 pole motor at 250rpm ( with gearing either via belt/thread plus pulley or other ) then there are 60 poles per revolution.

If you translate this to a merry-go-round - imagine 8 people pushing it versus 60 people pushing it - which model is smoother in operation. Especially if the "pushers" are uncoordinated, statistically the 60 push model would be much smoother by some margin.

The other issue is error correction.

If you think you cant hear error correction then look no further than the Grand Prix Audio DD turntables.They released a DD with state of the art error correction software using latest computer chips only a few years ago  - "unparalleled speed accuracy " they said.

Well, several upgrades later including some based solely on new updated software algorithms, the sound has miraculous been improved considerably - as confirmed by reviews and owners.

Conversely the main issue with most belt drives is weasel motors, and elastic rubber belts. Replace the rubber with silk thread, which has virtually zero elasticity, a decent motor and inertia in the platter and you get excellent speed stability.

I note the new Kuzma R & M now use rigid polymer belts with a motor that has been designed expressly for this purpose. 

Idler drives can achieve good speed stability, but only if they use decent size motors. My view is the "drive" of idlers" is due not only to the removal of the stretchy rubber belt, but more importantly most of them use decent size AC motors with bags of torque compared to the usual cotton reel sized motors used in most belt drives.

I just knew I should not have hit that bee hive with a stick😊

Correct, it is not cogging, as Lew and I have already explained.
“ the problem is pulsing”  let’s get this out of the way..,,
An 8 pole motor would be 2 phase. This is not a linear torque design, better DD TTs use a 3 phase motor, designed for sinusoidal back emf. These synchronous motors are inherently linear torque. They do not pulse 
Any deviation from linear torque due to manufacturing tolerances is called torque ripple. In good three phase motors, this is vanishingly low, but not zero. Nothing is perfect. Further, today’s sophisticated controllers can now “look” at this ripple and mitigate it in real time. Thus calibrate the controller and motor pair. Further this ripple for say a quality 18 pole motor running at 33.33 is sub sonic, well below 20 Hz

Consider what happens when you replace a compliant belt with a material that does not stretch. Since the motor is running much faster, you are now injecting the motors torque ripple into the platter. It’s frequency right in the audio band. 

Monaco improving their error correction is audible not because it is improving the motors no load speed accuracy but because it is improving how the drive reacts to platter speed changes due to the dynamically induced load caused by stylus drag. 
The fact that this can clearly be heard should concern some here who champion alternate drive architectures. Such drive systems are “blind” to these real time platter speed changes. 

As I said earlier, making a drive speed stable under dynamic load conditions is insanely difficult.

Making a drive speed stable under constant load is relatively easy.

cheers 

@richardkrebs 

better DD TTs use a 3 phase motor,

Can you provide a list of any DD TT's that use a 3 phase motor.

The platter will slow down and a belt, thread, tape or rim drive, would slingshot the paltter up to momentary over speed once the high stylus drag reduces.

Thats funny.

My TT platter is 26kg, that you would think that immediately after Horowitz has banged out a C1 fortissimo on his Steinway, that the 26kg platter would suddenly take off into overshoot like your direct drive software going wonky is fanciful.

A heavy platter like that used on the Final Audio, will lower the rates of deceleration, acceleration. High inertia reduces the magnitude of speed changes, it does not eliminate them. 

The connotations of the words "immediately and "suddenly" aren't appropriate and I did not use them.

DD TTs using three phase motors. The new Technics, the SAT, VPI. I'm not sure about the NVS and the Monarch, but would be surprised if they weren't three phase, given their quality. Then of course there is the motor I build and install into my K3 design