The Korf Blog

The inside story: our research,
development and opinions

9 August 2022
Turntable Drives. Part II, Belt Drive Problems
In our previous post, we have introduced the turntable belt drive and explored its function as a lowpass filter, rejecting the motor's torque ripple. Today, we'll go through the key problems associated with the belt drive—and some of the solutions.

Belt resonance: string-like and in tension

A belt is basically an elastic string stretched between the pulleys. And we all know what a stretched elastic string does—it sings (resonates).

With some poorly designed belt drives, one can hear the hum-like higher harmonics of the belt's low frequency "song" with a stethoscope. And, of course, this resonance is easily measureable with an accelerometer.

String Resonance
Other than being noisy, a resonating belt modulates the platter rotation with its own frequency. This degradation is immediately visible on any wow and flutter measurement, usually falling within the limits of "perceptual weighing" most meters use.

Stretch Resonance
Besides the transverse ("string") resonance, there's another mode that happens arguably more often. It's a resonance in tension (stretch and compression) rather than deflection. Typically lower in frequency than the string resonance, it's even more noticeable. This resonance is responsible for terrible measured performance of drives with soft belts.
Both resonances can happen even on well designed turntables with wrong, worn or stretched belts. All elastic belts are perishable. In daily use, they seldom last more than a year. When you replace a belt, It's a good idea to put a date sticker on or inside a turntable. Time flies, and it's easy to think the belt was "recently replaced" when in fact it's a few years old!

It always pays to use an original belt, or a high quality reproduction if the original is unavailable. With a cheap generic belt, your turntable will not perform anywhere close to specifications.
Belt Sticking and Slipping
Just having a resonant mode in a system can be benign. For it to activate, a provocation is needed. Something needs to feed in the energy that would excite the resonance.

With the belt resonances, these energy sources are the motor's torque ripple and the belt sticking to a (usually larger) pulley.

Belt sticking to a larger pulley
Belts do not release smoothly for a lot of reasons. Some simply have a sticky rubbery surface; some create static electricity that makes them cling to the pulley; some pulleys are poorly designed and cling to the belt. A belt-shaped groove on a pulley can capture the belt and prevent its orderly departure. Dirty surfaces stick a lot more than clean ones. Oily fingerprints on the belt or on the pulley create sticking points. Worn disintegrating belts stick to pulleys like glue.

To prevent belt stiction, engineers specify precision ground belts made with non-sticky conductive compounds. They roughen the pulley surfaces a bit. And they pay a lot of attention to static electricity paths.
What can you as a user do to prevent the belt sticking?

Use a new one made to manufacturer's specifications. Do not touch the belt with your bare hands. Clean the pulleys with alcohol every time you change the belt. Make sure your turntable is adequatly grounded to prevent the build-up of static electricity.
The belt seldom sticks to a smaller (driving) pulley. But it can slip instead. Paradoxically, the reasons for this slipping are pretty much the same as for sticking—dirt, fingerprints, thoughtless pulley design etc. Belt slipping on a motor pulley ruins torque ripple rejection and can also excite the stretch resonance.
Pulley Coaxiality, Belt Scraping and Wandering
Let's look at a typical belt drive from the side. What happens when the vertical axes of the driving pulley A and the driven pulley B are not exactly parallel?
In our illustration, the belt C would try to move up the pulley A. As there usually is something to prevent it from leaving A completely, it would bounce up and down on it. The belt is scraping the driving pulley A.

This vertical movement of the belt would cause it to wander vertically on the driven pulley B too. It is easy to see the belt's gentle oscillating movement when this happens.

Belt scraping and wandering is a bane of suspended turntables. It is virtually impossible to have any sort of a compliant suspension between the A and B, and consistently enforce the coaxiality. Is there no hope?
One straightforward solution is making the driving pulley A in the shape of a barrel, as shown above. This presents the same effective diameter to the belt no matter what the angle is, preventing the scraping. Unfortunately, it introduces some other problems that we will have to leave outside the scope of today's post. You can google "Crowned Pulley" for a more thorough treatment of the subject.
What can you do to stop the belt scraping and wandering?

Use a small round bubble level to make the upper flange of the motor as horizontal as your platter. Some turntables have an adjustment screw to tip the motor axis in the right direction a bit.
Side Load on the Bearing
An explanation of why the turntable bearings are the way they are is a long and sad story, worthy of a separate article. For now, we just need to know that they are hydrodynamic plain bearings that have to operate far below their minimum design speed. At 33 1/3 RPM, no consistent oil film forms between the sleeve and the shaft, and the bearing's performance is thus severely degraded.

Belt's tension creates a side load on the bearing
If this alone wasn't bad enough, the belt drive makes the problem worse. The belt's tension pulls the shaft to one side of the sleeve. An already inadequate lubrication is thus completely extinguished on the side facing the motor. This is clearly visible in the wear patterns. And the stick-and-slap performance of the turntable's main bearing goes from bad to worse.

A dry polymer sleeve, needing no lubrication, might be the answer. Sadly, these sleeves are often quite soft, and are not suitable for diametric loading.
Another possible solution is shown above. An addition of an idler on the opposite side solves the diametric loading issue. But the resulting doubling of all the other belt-related problems calls for an in-depth cost/benefit review in each specific design.
What can you do to minimize the side load on your turntable's bearing?

Do not tighten the belt more than specified. There's no benefit in having it highly strung. Make sure your turntable's main bearing is not running dry—there often is a felt pad on top of it that should be soaked in oil.
Tipping Force on the Bearing
This illustration shows a design common to inexpensive belt driven turntables. The belt's tension is applied above the main bearing, creating a tipping force. As the bearing, especially built to a budget, has some play, its axis is deflected from the vertical.

This makes all the problems we have described in (3) and (4) a lot worse. Hello, belt wander. Hello, dry spots inside the main bearing. Goodbye, good measured and subjective performance.

Poor driven pulley design creates a tipping force
This is a rare kind of a problem that can be completely and cost-effectively solved. Apply the belt's force in the middle of the bearing sleeve, and the tipping force is gone for good. That 9 out of 10 new belt driven turntables are built the other way tells us a lot about the quality of engineering going into them.
Pulley Circularity
This is another problem that would have not deserved a mention were it not so depressingly common.

For cost reasons, many driven pulleys even on fairly expensive turntables are injecton moulded from thermoplastics. Many of those polymers warp during cooling. The prototypes are made slowly and carefully, and the threat of warping goes unnoticed.

Polymer driven pulleys are often not very round
Then, in mass production, dozens of pulleys are moulded every minute, without any regard to ambient temperature, orientation during cooling etc. Some still run true, but most are not round at all. And the performance you get from the turntable you bought with your money is very far removed from a prototype in the hands of a magazine reviewer.

One well-known manufacturer deserves a special mention for building the driven platters out of MDF, a notoriously dimensionally unstable material. Fortunately, it looks like they have stopped this practice.
An upgrade to a metal driven pulley, if available, is often a cost effective way to significantly elevate your turntable's performance.

Belt DriveSummary
That's it for today. Have we covered all the belt drive features? Of course not! We haven't even mentioned the whole issue of non-elastic belts, or the role belt deformation plays, or why some are round while the other are flat, or the platter mass influence, or...

Still we can quickly outline the key strengths and weaknesses of a belt driven turntable:
Can be very cost effective; no additional moving parts
Good motor torque ripple rejection
Good vibration isolation
Fair measured performance
Good subjective performance is achievable
Poor reaction to load variation
Hard to design a bearing for
Belts do not last long, can be inconsistent
No longer cost effective when top performance is desired
In our next post, we will examine the idler drive. There's some superficial similarity with a belt drive, but a close examination will reveal critical differences.

Please subscribe to receive blog updates in your inbox!
comments powered by HyperComments