Like what you see?
Subscribe to the Korf Blog!
- Be notified when new articles are posted
- Access members-only secret preview sales
- Get deals and discounts on Korf Audio products
Expect 1-4 emails a month. We will never ever spam you. You can always unsubscribe—no hard feelings!
The Korf Blog
The inside story: our research,
development and opinions
development and opinions
8 January 2019
Turntable Main Bearing Vibration, Part III
A couple of months ago, we started exploring turntable main bearings. Their contribution to turntable performance is undisputable, but measuring their influence have proved difficult.
In the Part I, we have successfully demonstrated that vibrations generated at playback do excite the bearing. The resulting acceleration was very small but detectable.
In the second part, we tried measuring the noise generated by the bearing itself. We could detect some small irregularities around the expected rotation frequency of 0.5Hz, but all the other measurements came up empty. Most likely, we used the wrong tool to measure the noise.
For the next attempt, we have decided to use microphones. As it's impossible to fit a calibrated microphone of the required sensitivity inside the turntable, we've gone for the standard electret ones.
In the Part I, we have successfully demonstrated that vibrations generated at playback do excite the bearing. The resulting acceleration was very small but detectable.
In the second part, we tried measuring the noise generated by the bearing itself. We could detect some small irregularities around the expected rotation frequency of 0.5Hz, but all the other measurements came up empty. Most likely, we used the wrong tool to measure the noise.
For the next attempt, we have decided to use microphones. As it's impossible to fit a calibrated microphone of the required sensitivity inside the turntable, we've gone for the standard electret ones.
Here's our very precise, reliably repeatable and completely scientific way to attach a measurement mic to a Telefunken S-600 bearing (sorry, couldn't resist the joke).
Consumer electret microphones present 3 significant problems. First, the data they return below 50Hz makes no sense. For the task at hand, we certainly can live with truncated LF data.
Second, they are remarkably sensitive. At maximum gain, the mic attached to the turntable main bearing easily picked up conversations 3 rooms away. A door slammed at the other side of a building drove a DAQ into clipping. Car passing by on the street filled the data with noise. Because of that, all the measurements had to be done late at night, and we didn't do as many as we wanted to.
Third, the microphone and preamp turned out to be quite noisy. Fortunately, this noise is easy to subtract from the signal. Here's what the microphone, preamp and DAQ noise taken at 3 am looks like:
Consumer electret microphones present 3 significant problems. First, the data they return below 50Hz makes no sense. For the task at hand, we certainly can live with truncated LF data.
Second, they are remarkably sensitive. At maximum gain, the mic attached to the turntable main bearing easily picked up conversations 3 rooms away. A door slammed at the other side of a building drove a DAQ into clipping. Car passing by on the street filled the data with noise. Because of that, all the measurements had to be done late at night, and we didn't do as many as we wanted to.
Third, the microphone and preamp turned out to be quite noisy. Fortunately, this noise is easy to subtract from the signal. Here's what the microphone, preamp and DAQ noise taken at 3 am looks like:
Next step is measuring the noise generated during normal operation.
Ouch. That's a lot of noise.
"Motor running" is taken with platter running under motor power. "Running by inertia" is a measurement taken with motor switched off, belt taken off, and platter rotated by hand to reach about 33 RPM.
Here's an interesting thing: the bearing noise by itself is almost nonexistent. I believe that the LF component we captured in Part II is the biggest contribution of the bearing into the overall noise picture. Everything else looks insignificant. And please keep in mind that this bearing is very far from state of the art.
"Motor running" is taken with platter running under motor power. "Running by inertia" is a measurement taken with motor switched off, belt taken off, and platter rotated by hand to reach about 33 RPM.
Here's an interesting thing: the bearing noise by itself is almost nonexistent. I believe that the LF component we captured in Part II is the biggest contribution of the bearing into the overall noise picture. Everything else looks insignificant. And please keep in mind that this bearing is very far from state of the art.
And the last chart for today. Here's the recording of a 0-20kHz sweep from a test LP:
Energy generated during playback absolutely dominates the picture here. Even the motor noise looks not very intrusive in comparison.
I'll refrain from the detailed analysis of the chart for now, because the microphones we are using are not calibrated, and I have no idea what magnitude of harmonic distortion they introduce. But it does look like bearing's own resonance lies around 1kHz mark.
I'll refrain from the detailed analysis of the chart for now, because the microphones we are using are not calibrated, and I have no idea what magnitude of harmonic distortion they introduce. But it does look like bearing's own resonance lies around 1kHz mark.
For your amusement, and to give you an idea of how much energy travels through the bearing, here's a recording from an "Electric Dreams" soundtrack LP. It's taken from our measurement microphone, and no editing of any sort has been applied except for gain.
The main bearing's own contribution into the overall noise profile of turntable is insignificant. Bearing noise is a solved problem
I think that with the measurements presented today, we finally have enough information to answer the question in Part I: are plain bearings really that quiet?
It looks like yes, they are extremely quiet on their own. All the data guides us to a conclusion: the main bearing's own contribution into the overall noise profile of turntable is insignificant. Bearing noise is a solved problem.
That leaves us with a second question: why is there such a pronounced difference in subjective performance between various bearing designs? A part of it certainly lies in bearing's acoustical performance, in the way it acts as a conduit of vibration from the LP into the chassis. We'll test this hypothesis with measurements of more advanced bearings.
It looks like yes, they are extremely quiet on their own. All the data guides us to a conclusion: the main bearing's own contribution into the overall noise profile of turntable is insignificant. Bearing noise is a solved problem.
That leaves us with a second question: why is there such a pronounced difference in subjective performance between various bearing designs? A part of it certainly lies in bearing's acoustical performance, in the way it acts as a conduit of vibration from the LP into the chassis. We'll test this hypothesis with measurements of more advanced bearings.
Please subscribe to receive blog updates in your inbox!