it is very funny, the DD-drive concept divides the users in two seperate fields, lovers and deniers, as sometimes happens with the idler drive as well. From my experience with DD drives I am convinced that a DD motor needs a perfect environment (sorry Halcro). There are excellent examples of "very musical drives", e.g. the Technics MKIII in a fine plinth, the Nakamichi (Micro Seiki motor) or the Brinkmann Oasis.
I found an interesting description at Brinkmann maybe with relevance to the speed/time issue:
The drive mechanism is probably a turntables single most important component. This significance results from the fact that music consists of sounds organized in time. A turntable therefore has to play back a record at precisely the same speed at which it was cut, normally 33 1/3 or 45 rpm. Any deviation, no matter how small, from the correct speed will ultimately change pitch and tempo and result in music reproduction that is not true to the recording.
Our hearing is much more sensitive to short term speed variations, as opposed to long term ones. The onomatopoetic term wow & flutter correctly identifies the underlying issues wow describes continued, longer lasting deviations, caused for example by eccentric records, whilst flutter denotes short-term irregularities best demonstrated by piano tones that fade away with a slightly howling quality.
Let us consider the following example: a 1 kHz tone (1000 cycles per second) is cut into a lacquer at precisely 33 1/3 revolutions per minute. If the pitch is to be accurate, the turntable has to play back this tone at exactly the same 33 1/3 revolutions per minute. A seemingly insignificant variation of 1% would result in a playback speed of either 33 or 33@d revolutions per minute with pitch being off by ±10 cycles. Even people with perfect pitch would have a hard time correctly identifying a 990, 1000 or 1010 Hz tone without a reference yet even brass ears would easily hear the differences in a direct comparison.
Speed precision is perhaps even more important for the musics tempo. For our perception of music to be real, the tempotiming and paceis more crucial than the pitch. This is where the 1% difference would make itself heard far more easily. Considering that a typical record has a playing time of around 20 minutes per side, a ±1% difference would amount to no less than 24 seconds. This seemingly irrelevant deviation would result in either a slightly restrained and darker sound, or a livelier, brighter reproduction of the original event.
In order to meet the claims of High Fidelityno more, no less than music reproduction true to the original the cutting and playback speed have to be absolutely identical. Due to reasons being outlined in the following pages, this theoretical ideal is practically not feasibleturntables are hence (and in the best case) the closest possible approximation to the theoretical optimum.
Two elements determine the rotational speed of an AC based electric motor: the power line frequency and the number of magnetic poles. The formula used to describe this is as follows: rotational speed = line frequency x 120/number of magnetic poles. In countries with a 60 Hz power grid, a two-pole electric motor therefore runs at 3600 rpm (60 x 120/2 = 3600). Each doubling of magnetic poles cuts the rotational speed in half: 4 poles equal 1800 rpm; 8 equal 900 rpm; 16 equal 450 rpm. Theoretically this could be continued even further, however, practically, there are only so many poles one can fit inside a motor case.
Further, each additional pole also increases the motors torque, which in turn increases the cogging effect. DC electric motors dont fare any better either and most are offered in standard 1500 rpm configurations. This showcases the ensuing problem quite easily: on one hand we have an electric motor spinning at 1500 rpm, on the other we have a platter which has to spin at 33 1/3 rpm. The ratio of motor to platter speed is therefore 45:1. Logic dictates that the drive pulley has to be 45 times smaller than the platter.
If we take a platter diameter of 30 cm as an example, this would equate to a pulley of only 6.6 mm in diameter. The contact area for a belt is therefore very small, which in turn increases slippage; which in turn influences speed stability negatively. The problem is further exacerbated by the fact that the cogging is transformed into a frequency range in which human hearing is very sensitive
The power line frequency (and thus the motors rotational speed) can be reduced with frequency converters; however, the cogging problem still persists.
The platter of a turntable can either be driven directly or indirectly. In case of an indirect drive mechanism, platter and motor are two separate components whereby the power from the motor is transferred to the platter through intermediary means. In case of a direct drive turntable, platter and motor are assembled into one and the same unit.
Motor and platter are two separate entities; some sort of medium is necessary to couple the motor to the platter. Typical applications include the following:
> Belt (round or flat; geared or v-belt type; various threads made of synthetic or natural fibers)
> Idler wheel
> Gear wheel drive
The choice of transfer medium is critical since it not only transmits the motors torque to the platter but also all motor-based interferences, among them the motors cogging, which compromises speed stability since the platter is not driven continuously but somewhat jerkily instead. The more direct the connection between platter and drive, the more directly these interferences will be transferred. An idler wheel for example couples the motor very directly to the platter. As a result, the cogging effects are more pronounced; in the best case they manifest themselves as a slight flurryness of the sound; in the worst case through clearly audible distortions.
Because of the reasons stated above, most turntable designers therefore prefer a softer coupling between the diameters of the motor pulley and the platter. Even tiny discrepancies in the belts uniformity will have a serious impact on the platters speed stability. Consider the following example: say a DC motor running at 1500 rpm drives a 20 cm sub-platter via a 1 mm thick flat belt. A discrepancy in tolerance of only 1/100 mm will have a net effect of 0.9% wow & flutter on the platter!
Countering these effects demands substantial efforts in the design of a belt driven turntable. Heavy platters (12 kg or more) and low rpm motors tend to mitigate these side effects to the point of not being significantly audible.
Studios (radio stations in particular) demand quick start-up times turntables typically have to reach their correct speed within half a revolution. For LPs this means 0-33 1/3 rpm within 0.9 seconds. Such acceleration figures can only be achieved through use of high-torque motors and correspondingly tight coupling between the drive and platter. It isnt a surprise then that for decades idler wheel drive designs were the de facto standard in studio applications.
But idler wheel turntables also had seriously high maintenance costs in order to be up and running 24/7 and to avoid rumble and other sound degrading issues caused by worn out idler wheels to affect the sound negatively.
Thus out of necessity, in the late 1960s manufacturers of studio turntables began to look for
low(er) cost maintenance alternatives. They came up with direct drive, whereby the platter was placed directly on the motors shaft, ie the stator was mounted around the bushing and the shaft was used as the rotor and voila, the goal was achieved; at least in theory.
But start-up times of less than 1 second necessitated high torque motors, which designers achieved byusing motors with 32 and more poles. The penalty they paid were heavy cogging effects accompanied by high wow & flutter numbers. The cure was found in quartz locked motors and phase locked regulators; which corrected for any deviations from their preset with an iron fist.
On paper at least, these corrected direct drive turntables boasted hitherto unimaginable low wow & flutter numbers down to 0,001%. But the rigorous iron fist regulation prevented the platter from spinning smoothly; instead, the regulation caused the platter to oscillate continuously between speeding up and slowing down.
These start/stop motions translated into an unpleasantly rough and hard sound; odd as wow & flutter numbers in the 0,001% range are deemed inaudible. Once the direct drive technology had gained a foothold in pro audio applications, the benefits of mass production (ie. trickle down effect) made sure that very soon even $100 turntables were equipped with direct drive and advertised as having less than 0.01% wow & flutter. This is precisely where direct drive got its bad rap sheet.
Under closer scrutiny however, this assumption were based on some misunderstandings. For one, in home audio application use, turntables are not really required to reach 33 1/3 rpm in less than a second, thus 32 pole motors and phase locked regulators are not really necessary either.
Finally, one more relatively unknown factoid: while it is common knowledge that direct drive turntables became popular in the late 1960s, their actual invention dates back to a small Swiss manufacturer in 1929. The name? Thorens, who enjoyed and even today is looked upon with great respect for their idler wheel and belt driven classics.
