Audiophiles Once Loved Direct Drive, Now They Seem To Hate It

I've always wondered what's being referred to when people say a turntable is noisy. Is it the mechanicals, as in you can hear the table in operation from your listening position, or is it the noise floor you can hear in the recording through the speakers?

 

- .004% you're right for the low side

(.008% x 33.333)+ 33.333 = 33.3356 (high side)


My math for the high side ( .01% x 33.333)+33.333 = 33.3363 (high side)

Pick one.

M~

 

Tim, VPI has a rumble meter, the Clasaic Direct rumble was below what the meter can measure. They estimated it to be -100dB.

 

VOT, You missed the single most important fact about modern high performance DD TT's. They now use modern non-cogging motors that eliminate the worse feature of old DD TT's. It isn't just the accuracy of the speed, bu tthe stability of the speed. Motor bearings have improved to the point were rumble figures of below -90 dB are possible.

 

Most of the better vintage direct drive tables also used coreless motors, starting with Dual in the early 70s. I think by the early 80s, it was just Technics and Denon among the majors that stayed with the iron core motors.

That said, some of the new breed of coreless motors are really nice, including the ones from Technics, and your VPI motor from ThinGap in California with the cylindrical composite stator is really cool. A lot nice ideas out there now, I especially like some of the low torque, low feedback designs from Brinkmann and STST. They're all pretty expensive, for sure, but great to see some of the neat designs out there now.

 

People who know what DD can do, still appreciate these tables. I bought a Rega Planar 8 last year but am using my good old Dual 704 all the same. Throughout the decades I owned several DD an belt driven tables and both can be crap or great.

 

Ortofon, Neutrik, Grado and Technics for example.
Many AT and Yamaha products aren’t made in China either.

 

Regardless of whether someone is designing a direct drive or a belt drive table, one of the major challenges is dealing with having to drive the turntable with a motor that's fixed and coupled to the platter and the cogging torque ripple and the motor vibration and the possibility of mechanical breakthrough of that vibration, in relation to not just speed stability but also noise and cleanness of tracking too. The motors themselves, how they're driven, schemes maybe for decoupling them from other elements of the system -- the choices designers make about these things go a long way to defining how one turntable or another will sound, regardless of whether or not the platter is being driven directly or by via a belt.

 

Keep in mind that the part of the turntable that vibrates are the bearings.
At a DD turntable, the motor spindle bearing is the platter spindle bearing, while a BD turntable has both a platter spindle and motor spindle that produces vibrations.
So if you fully isolate the motor from the rest of the BD turntable, that same source of rumble as in a DD turntable - the spindle bearings - is still there.
I've never heard cogging, but that's part of the past... some manufacturers still use those old designs though, like Hanpin and Thorens.
Modern DD turntable designs use a coreless motor that has no cogging issues.
The ripple from "hunting" around the correct speed is an issue with badly designed controllers that have a stepped output, where one step represents just too much speed and one step less is just too slow.
Similar to a stepped volume control, where for example 30 is too loud and 29 is too quiet and there's no step in between.
Proper motor controllers don't work with those steps but are variable instead (like an analogue volume knob), so once they achieved the constant speed, they maintain it unless the motor is too weak to overcome the irregular friction of the stylus.

 

this is the correct way to calculate deviation:
high side 33.336
(33.336-33 1/3)/(33 1/3) x 100 =
0.0080%

Either way, moot

Has this ever been tested using external measurements? The reason I ask I'm familiar with process control and that kind of accuracy is not readily seen, especially considering the measurement feedback sampling = the controlled variable target or set point.

That would be the equivalent of controlling a 1800 RPM motor to +/- 0.15 RPM or based on 60 Hz, +/- 0.005 Hz, which is outside grid tolerance iirc.

Is the VPI motor AC or DC?

 

After reading this I get it now. That is a long term weighted average sampled every 1.8 sec (@ 33.33 rpm). The unit adjusts the 60 Hz (or 50) freq going to the synchronous motor.

Based on that those low numbers make sense.
The SL1500C is 0.003% based on that. 0.1 Hz average deviation on a 3150 tone.


The procedure entailed measurements during 15 minutes of turntable warm-up and an additional 5 minutes of rotation after the LP was placed on the record, complete measurements of beginnings and endings of songs on an entire LP side, and a few minutes of measurements of the LP runout groove. These steps were followed by an additional set of 2-minute measurements with the stylus removed from the record and then with the LP removed from the platter.

Using this test procedure, I tested 13 different turntables (Acoustic Signature, Basis Audio (3), Grand Prix Audio, Kronos Audio, Nottingham Analogue, SME, TW Acustic (2), VPI, Walker Audio, and Wave Kinetics) with the RoadRunner tachometer alone. All but two direct-drive units and one closed-loop (speed read and adjust) belt-drive unit were subject to system tolerances, warm-up times, and drag forces to one degree or another.

Phoenix Engineering Eagle PSU and RoadRunner Tachometer - The Absolute Sound

 

I don't care to engage in belt vs. direct drive. People will like what they like, and there are plenty of options out there.

But I am surprised that people deny that anti-direct-drive sentiment is a thing. When I was shopping for a turntable, my local hifi dealer turned up his nose at the idea of a direct drive turntable. "That's for DJs," he said. As in, not for audiophiles. I considered several options before settling on my Technics SL-1200GR, which I brought into the dealer to show him. He was impressed and looked into becoming a Technics dealer right there on the spot. Unfortunately, they wanted him to carry their full range of products, including amps, but he only wanted the turntables, so it didn't work out. But that anti-direct-drive sentiment is still out there, for sure.

 

I wonder if with abx heartest you can hear differences between turntables (including tonearms of course).

 

The motor itself can vibrate as it moves, and those vibrations can be transmitted to whatever the motor is coupled to -- which can include not just the platter but the tonearm as well,. And there's cogging torque and torque ripple with any motor with magnetic poles. So, how designers address that -- like using a coreless motor, or reducing voltage to a belt drive AC synchronous motor after start up to lower torque and ripple amplitude, etc, all that kind of stuff you're talking about, are the things designers do to make one turntable better and quieter and more stable than another. That's all I'm saying. Motors are a challenge in turntable design. And how designers address the issues that the need for a motor raises make a big difference whether you're designing a direct drive deck or a belt drive deck.

 

Except for Garrard you mean?

 

And Lenco (Lenco 75). And Thorens (TD124). And EMT.

 

The spindle is the motor itself.

Without iron cores, there's no cogging.
A few years ago, Technics released a turntable that has a cogging free motor, a few sources:
Technics Unveils A New Direct-Drive Turntable At A More Affordable Price
The Resurgence of Direct Drive Turntables
Grand Class SL-1200 Series SL-1200G - Technics Canada

However, this is not new and there have been more turntables in the past that also had these coreless axial motors to eliminate cogging.
For example this Sony PS-X50:

In a variable motor control, there is no torque ripple as it runs at a constant speed. Torque is constant and inversely proportional to the acceleration of the platter.

 

VPI uses a 60Hz sync motor. Why they'll also sell you a SDS or ADS speed controller with a speed knob for $1250. Helps to get the inconsistent belt stretch, belt width, rubber band friction, pulley groove diameter problems temporarily in line.


Funny, the calculated 33 RPM speed above is just what I obtain on my coreless direct-drive non-quartz Technics SL6 - but it stays constant (and took a lot of trim pot tweaking):



Deviation from speed within one revolution instead would be in the wow-and-flutter specification - a specification that would also measure "cogging" and the separate "torque ripple".

 

T = Power/speed
Power is very closely regulated, constant, since load is constant. This can be measured.
Speed we know is constant, or dang close.
Therefore T fluctuation is constant, or very close.

But this applies to belt drive too
The belt oscillates under tension

 

Exactly. The load is constant, so there's no reason for a feedback loop to create a ripple, unless it permanently oscillates around the setpoint - which is only the case with a "stepped" speed regulation when the setpoint is in between two steps (or badly damped system of course).

 

Without opening a can of worms the Technics DD motor is an AC synchronous motor (they call it a brushless DC). Regardless of the control method if it is given a constant frequency it will turn a constant speed.
Load is constant within a very narrow range: coef x tracking x cos (offset) = drag, < 1/2 gm.
This may vary by 10% or 0.05 gm.
A very small change in load will not impact speed, only draw a bit more power.

My GR
Starting T ~ 2200 gm-cm
If running T is only 1/3 or 700 gm-cm
At 6" thrust = 46 gm >> 0.05 gm

Moot

 

Yeah... that's just a matter of definitions...
Many DC servo motors for example are actually an AC motor with a controller built in that takes DC as an input.
In those cases the "DC" describes the motor as an assembly, not the specific part with the rotor and stator.

Yep, the only fluctuations in load are the drag of the stylus (various amounts of groove modulation) and irregularities of spindle bearing friction, which is "overruled" by a powerful motor and a heavy platter. Until the load becomes too high, like touching the platter and the motor/platter is actually slowed down.

 

In summary:
Intrinsically based on design type
DD > belt or idler FALSE
Belt of idler > DD FALSE

DD = belt = idler TRUE

all based on design engineering and quality of implementation

 

You forgot time and money constraints.

 

I consider that as a factor of implementation, budget (basically time & materials)