An Interview with Thom Mackris of Galibier Design

LB:      Keeping with key architectural decisions, tell us about your preference for unsuspended designs.

TM:    At the end of the day, it’s about speed stability.  The combination of our motor drive, lack of suspension, and rigid belt (along with bearing tolerance and lubrication specification) gives Galibiers a direct-drive like sense of immediacy and pace.

I view both suspension and compliant (rubber) belts as an opportunity to introduce chaotic speed behavior into the system, and I avoid it at all costs.

A suspended design implies using a stretchy belt, and the combination of suspension “wiggle” (it doesn’t take much) and the stretch of the belt introduces speed instability which, while it may be below the threshold of audible wow and flutter, is nevertheless destructive to playback fidelity.

In a rant on my forum, I discuss this “micro” level of speed stability in the context of a turntable I hold in high regard – the Rockport Sirius.

Basically, this next (more subtle) level of speed instability produces IM distortion which is particularly nefarious.  If your readers are interested in this topic, they can read more about it here:

http://www.galibierdesign.com/phpbb/viewtopic.php?f=29&t=137

We’ve all heard how stylus drag modulates platter speed, depending on the amplitude and dynamics of the musical passage cut into the record groove.  With a rubber belt, this drag results in a continuous stretch and release of the belt, and suspensions work the same way, by “winding up” and releasing.  Bear in mind, that on a suspended ‘table, the motor is almost always separated from the platter by the springs, so you have two unstable, chaotic systems between the motor and the platter.

The problems produced by suspensions and rubber belts are the key reason why people have been gravitating to direct drive and rim drive.  We’ve solved this problem differently, by removing the suspension and using rigid belts.

With each improvement we made to our drive system, we heard better top end extension, smoother highs (not a paradox – this is the reduction in IM distortion), tighter, and yet more harmonically dense bass. This last attribute may seem a bit odd, but take this in conjunction with the extended highs, and it will make sense to you.  Bass harmonics are after all, upper frequencies overlaying the fundamentals.

LB:      Let’s turn now to what many consider the “heart” of a turntable, it’s bearing.  What kind of bearing(s) do you use, and why?

TM:    I agree with you about it being the heart of the turntable, and to date, the original design has been the most difficult part of our turntables to improve on.

Our bearings are a simple stainless spindle and hardened thrust ball, housed in a brass body.  They’re overbuilt, and made to industry-leading tolerances which require extremely thin oil.  The tolerances are so precise that the shaft must be dry during assembly (no lubrication).  Otherwise, the air captured in the bearing body cannot be displaced as the shaft attempts to seat into the oil reservoir.

Our bearings are non-inverted.  The argument for inverted bearings is that, due to the inherent wobble, an inverted design with the pivot point high (the thrust surface), and the platter mass slung below it confers dynamic stability in the same way that a tightrope walker’s bar helps stabilize him or her.

The theory is good, but it once again presumes its necessity because of poor manufacturing tolerances.  We’ve changed belt heights and performed other experiments and have never been able to induce rocking.

We like lubricants, and a non-inverted bearing frees us to (1) use them, and (2) have additional choices for the thrust surface (other than jewels).

I honestly can’t say much more about our bearings, other than that their attributes are a combination of simplicity and execution to the highest possible tolerances.  We pay a lot of attention to this.

We have a new design on the drawing board but it’s too early to tell if this will yield any benefits.  Should it prove itself, it will of course be reverse-compatible with all Galibier turntables ever produced.  I’m not at liberty to say more at the moment.

LB:      An important aspect of turntable design is that of controlling resonance.  While the plinth and platter should prevent resonance from reaching the stylus, there are different views as to how this should be accomplished.  What was your goal, and how did you approach it?

TM:    The nature of higher mass materials is that they tend to be resonant to a greater or lesser degree – the “stiffie” philosophy I mentioned earlier.

In the resonance control hierarchy, the platter gives you a far bigger return for your manufacturing dollar than the base does.  As we developed our product line, our focus was on retaining as much technology in the platter in each model as the budget would permit.

Our Gavia and Stelvio platters utilize aluminum carriers whose resonant characteristics are addressed by damping internally with oil and lead, capturing vibrations from both directions – the stylus as well as the turntable stand.  The business end (the mat) is a ¼” layer of carbon fiber.

Let’s trace the vibrations beginning with the chatter generated by the stylus tracing the record groove.  Any platter will fail miserably if the stylus chatter reflects back to the LP from the platter.

Assuming you’re trying to drain vibrations from the LP into the platter, you need to match the speed of sound between these two adjacent materials.  If there’s too much dissimilarity, the stylus chatter will reflect off the surface of the platter and back to the LP instead of being absorbed or transmitted.

There’s a balancing act involved however, because you also have to transition from the mat to the body of the platter as well.  Taking this into account, three ideal materials for the LP interface (the mat) are PVC, graphite, and carbon fiber.  For reasons I still can’t explain, materials like Sorbothene mats are suboptimal.

To give you an idea of how delicate this balancing act is, we performed an experiment, inserting a sheet of paper (.003” thick) between our platter top surface (graphite, at the time) and the layer below it.  Transient response fell apart.

The above experiment points to the reason we use mechanical fasteners rather than chemical bonding in our platter construction – to exercise manufacturing control over the material interfaces.  You never know when a manufacturer will decide to change the formulation of a bonding agent, essentially changing your design without your knowledge.

Once vibrations have made it to the platter body, you’re home free.  Energy transmission through the bearing and into the base is fairly straight-forward if you’re using a metal base. The material similarities between these parts encourage energy transmission.

Taking things in reverse (vibrations coming from the shelf, etc.), the lead and oil damping in the platter serves as an effective barrier.  In the Stelvio-II base, the added damping provides an additional barrier to stand-borne vibrations.