The Korf Blog

The inside story: our research,
development and opinions

17 October 2019
Building a High-Performance Headshell, Part I
In the coming series of posts, we will describe some of the design process steps we use at Korf Audio, from material selection to the use of 3D printing in prototyping. I can't promise to show you everything, but I will try to make this little excursion into the Korf Audio lab interesting.

This is a first article in a series, and the topic today is material selection.

99% of headshells on the market are made of light metal alloys. Most manufacturers use aluminium alloys, though some prefer magnesium. Wood is another common choice for the more expensive kind. Carbon fiber is not unheard of. There's some exotica that's made out of titanium. JVC and SAEC used zirconia ceramics in their top of the line headshells for a brief time in 1970s.

But how do we tell what works and what doesn't?

We could choose the same path we walked with the tonearm arm wands: design a simple prototype and build it out of all available materials. With tonearms, we had no choice—there's no substitute for actual physical measurement of something that is as complex as the tonearm.

Fortunately, the headshells are a lot simpler. We don't need to build them all. A simple computer simulation will do just fine.

We need to make a 3D model of a headshell first. For that, I use and heartily recommend Onshape. It's a full-featured parametric CAD system that runs in your browser. There's nothing to install, and you don't need a typical high-powered CAD workstation to run it—all the heavy lifting is done "in the cloud" by Onshape's powerful servers.

Another great thing about Onshape is the relatively shallow learning curve. If you know the 3D parametric CAD basics, you can watch a few videos and be productive within a day or two. And if you don't mind others seeing your designs, it's completely free.
Finite Element Analysis
We live in a truly wonderful era. The progress in computing, and in particular the advances of "cloud computing" mean that the tools that required many millions in investment only 15-20 years ago are now available for free (or almost for free) to anyone who can sign up online. One of such amazing tools is Finite Element Analyses.

I would not go into depths of what it can do, but for our application it can simulate the vibrational behaviour of our headshell model.
There are dozens of FEA (Finite Element Analyses) systems on the market, and choosing one is a daunting task if you're just starting out. Fortunately, there's a reasonably simple one that is well-integrated with Onshape. You can import Onshape models directly without jumping through fiery hoops. It's called Simscale.
When you first open it, the interface might seem a bit overwhelming. Fortunately, there are very well made videos and tutorials that explain in great detail how to make your first simulation. Again, if you don't mind others looking over your shoulder, 3000 hours per year of simulation are free. It's a lot, and doing what I do, I cannot imagine ever consuming them all.

For simulations to run quicker, we've designed the simplest possible one-piece headshell. You can find the document here, and you're welcome to use it for your own experiments.

FEA systems produce the prettiest graphics ever. Here are the vibration modes of our simple headshell. The warmer the colour, the more the deflection. Grey mushroom-like arrows show the exact vectors of said deflection.
But, pictures aside, we are doing the simulation to understand what to make our headshell out of. How do we decide?

One of the outputs of a FEA simulation is a chart of modal frequencies. They are not quite the same modes we are exploring with our accelerometers. Rather, the software finds distinct ways our part can bend, and tries to guess at which frequencies it would do so. The animation above actually illustrates all these modes.

We have started with aluminium alloy and have done dozens of simulation for other alloys and light metals. After a while, this became very boring. The differences with normal engineering-grade aluminium alloy were nonexistent to minimal.

So we decided to be a bit more adventurous and include the ceramics. There must be a reason why SAEC and JVC used them, right?
Alumina ceramic is the real winner
Now we're talking! Zirconium dioxide ceramic comfortably outperforms the aluminium alloy. But the real winner is alumina ceramic. The predicted natural frequency of our simple headshell in alumina is twice that of metal. This is an amazing improvement—if it works in real life!

Can we build a headshell out of alumina ceramics? What configuration shall we use if we do? Stay tuned, and you'll find out in our next blog post!
Please subscribe to receive blog updates in your inbox!
comments powered by HyperComments