LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Hi-Fi discussion
Post Reply
hifiaf
Posts: 1526
Joined: Tue Apr 28, 2020 12:49 am
Contact:

LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by hifiaf »

In my pile of LP12 things to try out — and try to understand — I've left the bearing-related stuff mostly for last. This is partly because I know how critical they are to performance, so I want to get everything else sounding really good so that I can really hear the differences. It's also because I don't really understand the differences between them.

With the arrival of the Karousel in the last few months, my confusion about bearings has only increased. The threads I've found on the Karousel seem to very quickly turn away from what it is and how it works to other topics. Even the ones filled with people claiming it's transformed their system don't seem to offer much in the way of explanations of how it works.

Since the discussions in this forum are so refreshingly rational and constructive, I thought this would be the ideal place to learn more about bearings. But if I'm asking obvious questions that have already been adequately covered elsewhere, please let me know!


1) CIRKUS + TRANQUILITY

I currently have a Cirkus bearing. I had a pre-Cirkus until a few years ago, and I vaguely remember it sounding better when I swapped it for the Cirkus, but I don't think my setup at that time was good enough to show much of a difference. Until recently, probably because of my background in bikes, "bearing" meant balls or needles rolling against races, so I thought that must be what a turntable bearing was. So I was surprised when I learned that the Cirkus bearing is just some PEEK-lined steel with a polished steel thrust pad at the bottom (if I'm not mistaken!), against which the spindle of the subplatter sits. (As I also learned, the "single point" -- an idea never actually realized in practice -- is what is depicted in the Linn logo. And there were lawsuits about it, and lots of water under the bridge, and some lingering hurt feelings based on the passion this all evokes in forums.)

I currently have a Tranquility on its way to me in the mail. Although the details surely elude me, I think I understand the basics: it uses magnets to reduce the weight of the spindle on the thrust plate by 90%. Less weight, less friction, less noise.


2) MOBER

I have one of the Mober bearings in my basement, which Edmund sent me to try. Now, Edmund's bearing actually does have ball bearings in it. He has told me that these ball bearings are NOS 1990s Fafnir bearings that he considers of extremely high quality, and that with lesser ball bearings his design would not work well. I know that he believes the Cirkus is too tight and doesn't spin freely enough, that the pre-Cirkus is too loose and wobbles, and that his bearing is an ideal compromise between stability and free spinning. I also know that Simon and Mark have had their hands on the bearing and consider the design flawed, theoretically as well as in practice. But I don't really understand the reasoning for or against, and I haven't heard a turntable with a Mober bearing — though at least one person on pfm with a Mober bearing uses it with a Tranquility and is very happy with the combination.

Here is a diagram that Edmund sent me of the bearing:

Image


3) KAROUSEL

I don't have a Karousel and don't currently plan on getting one, but some idle Googling tonight reveals that lots of people seem to be buying and enjoying them, so I want to understand better what they are.

Beyond the fact that the body of the bearing is threaded, the official Linn product page (https://www.linn.co.uk/sources/turntabl ... s/karousel) is light on details about what is new or different about the Karousel versus the Cirkus. Better machined, better designed, stronger, more rigid — but not a lot on how or why. At one point, this product page had something about a "diamond-like" substance on the thrust pad, I believe, but that language is now gone. I found a post on the Lenjonklou forum has an image that suggests that in the process of production, the design of the Karousel has changed, and it now has a standard Cirkus-style mirror-finish steel thrust plate. https://www.lejonklou.com/forum/viewtop ... &start=225

Image

So I don't feel I understand the Karousel. But I do imagine that it is incompatible with the Tranquility because of its threaded body-and-nut style.


Any thoughts and clarifications are very much appreciated!
Nobeone
Posts: 977
Joined: Sun Mar 29, 2020 8:32 am

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by Nobeone »

Rolling element bearings are more complex mechanical systems, tend to be noisier, and have less easily predicted life times. When plain bearings wear you get a general broadband rise in noise, with rolling element bearings you get ringing and knocking at frequencies related to the size, speed, and number of rolling elements as they rotate under the point of load. There is a whole industry out there (that I used to work in) trying to monitor bearings and predict maintenance requirements.

What rolling element bearings are good at is low rolling resistance I believe. So on a rotating horizontal shaft with the load weighing down vertically then I think they will typically offer lower rolling resistance than a plain bearing. But here the shaft is vertical, and the load is mainly vertical too. The horizontal loading would be relatively small I assume. I am not sure what the benefits are in this application. It may be "nice" to see your platter whizz round and take ages to stop, but does that improve replay? Here my understanding runs out too ...
User avatar
YNWaN
Site Admin
Posts: 6797
Joined: Wed Apr 27, 2011 2:02 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by YNWaN »

There is a great deal to write about all this so forgive me if it comes out a bit piecemeal.

As usual with Linn there is quite a bit of obfuscation and marketing nonsense surrounding the design of the Karousel. In reality, one of the key design features of the Karousel is one that isn’t discussed and that is the way it attaches to the chassis. Now to describe Linn’s mechanical development as incremental hardly does the slowness of it justice. All previous bearings attached to the chassis by three machine screws. The head type, but not the thread or diameter, of these screws changed over the years. Linn also recommend doing these fasteners up very tight (I see they now give torque values). The result is a very high point load at the head of the screw. This was fine when the chassis was steel but now it is made from aluminium, which is very much softer, I suspect a number of installations have been ruined by over tightening these screws - remember, the Keel is only about one millimetre thick at this point and there is very little material. The Karousel now uses one large load spreading nut and flange and the three holes previously used for attachment are now just locating lugs to prevent rotation whilst tightening the main nut. Of course, all of that could have been achieved simply by making a steel collar/washer that fits over the top of the existing bearing and spreads the load....

Linn seem to have already dropped the ‘diamond like coating’ and I suspect for this reason. Such coatings are intended to protect sliding surfaces and that’s fine because that describes almost all bearings...... but not a turntable bearing. The kind of bearing the LP12 uses is actually two bearings of different types - one that is a sleeve bearing and carries the side load and one that is a point bearing and carries the weight of the platter. The DLC was applied to the thrust pad to carry the vertical load. I suspect Linn found that the black finish of the DLC quickly wore from the extremely high, and extremely small size, point load of the main bearing tip which resulted in a silver dot being visible if looking into a clean bearing and this is clearly a problem for even non engineers ;). In other words, the DLC coating didn’t work in this application.

Okay, I’m having a break, tons more to write though - in a bit.
Mark
ChrisOH1
Posts: 505
Joined: Tue Jan 02, 2018 9:58 am

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by ChrisOH1 »

Look forward to this thread, I've been looking at alternative bearing options for a while now.
Nobeone
Posts: 977
Joined: Sun Mar 29, 2020 8:32 am

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by Nobeone »

Ah I had not spotted that Mark, "Linn tight" on a Keel is an expensive mistake perhaps?
User avatar
YNWaN
Site Admin
Posts: 6797
Joined: Wed Apr 27, 2011 2:02 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by YNWaN »

The black colour of the DLC that Linn was finishing their thrust pads in suggests silicon nitride and likely vapour deposited - to what thickness we don’t know and it’s possible that the cost to benefit ratio was just deemed too poor.

A lot of play is made of increased stiffness in the Karousel but, in reality, this will be one of the minor elements - if indeed there is any increased stiffness at all! I say this because the key features that effect the stiffness of the structure are unchanged from the Cirkus (or even the pre-Cirkus) bearing. Those aspects are 1/ the diameter of the flange used to mount the bearing and 2/ the diameter and length (you could argue this makes three aspects) of the bearing shaft. Other elements do make a difference but these are tiny compared to the fundamental ones I’ve just listed.

The use of stainless steel is also made play of but this isn’t actually inherently better for this application.

————-

What does make a difference in a turntable bearing is,1/ the size of the gap allowed between the bearing shaft and the liners, 2/ what material those liners are made of, 3/ what side load is applied to the bearing, 4/ what viscosity lubricant is used and 5/ how tight the allowed tolerances are. So, let’s try to deconstruct those points.

1 and 4 are inherently linked and pretty much always overlooked by those discussing the correct oil to use in a turntable bearing. In essence, the smaller the gap the thinner the oil or, the larger the gap the thicker the oil. Significantly, altering the viscosity of the lubricant significantly alters the drag of the bearing but also its stiffness. Altering the drag of the bearing, if applied via the lubrication, does not increase noise but it does alter the way the motor is loaded. Our experiments show that applying a constant load to an AC motor increases applied torque and reduces cogging. DC motors are another matter though and certainly, if quick changes in speed need to be applied, it is likely to make this very difficult (but then so does a platter acting as a flywheel which is why many DD decks use low mass platters). As far as I’m aware Linn are using the same oil as previously which suggests the the bearing gap is largely unchanged.

2 is self evident I think? Different materials are more, or less, compliant and have different friction. They also differ significantly in how accurately they can be machined.

3 the side load, in this case, is defined by four aspects. 1/ how square/perpendicular the thrust pad is fitted to the bearing, 2/ how perpendicular the bearing is mounted relative to the chassis, 3 how level the platter is and 4, how well balanced the platter is.

5 in an ideal world components should have zero tolerance. However, in the real world, a tolerance is always required - a range of measurement within which a component is said to conform to the base specification. The closer, or tighter, this specified tolerance the harder and therefore more expensive the component is to make (hugely so if extremely close tolerances between mating components is sought). One of the most common ways of achieving tight tolerances is to match mating components, so slightly larger diameter shafts are matched to slightly larger diameter sleeves to achieve an overall tolerance between the two that is very tight, for example. However, despite some claims, Linn do not (and never have done) supplied their inner and outer bearings as matched pairs. As an aside, about a year ago a friend reported to me that he had just bought a new Cirkus bearing and it was a very tight fit - clearly the tolerance between the two bearing parts were close to their extremes (or likely one was beyond that specified*). Another aspect to consider with very tight tolerance bearings is the expansion and contraction of dissimilar materials according to temperature. This is a very real issue with very tight tolerance components and can alter friction levels considerably.

* Tolerance drift is inevitable as machine tools wear. Replacing cutting tools more regularly increases cost.
Mark
User avatar
YNWaN
Site Admin
Posts: 6797
Joined: Wed Apr 27, 2011 2:02 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by YNWaN »

Nobeone wrote: Mon May 11, 2020 11:57 am Ah I had not spotted that Mark, "Linn tight" on a Keel is an expensive mistake perhaps?
Yep, very much so. Particularly given how ham-fisted and mechanically unaware many people/dealers are. I reckon that with sufficient enthusiasm you can significantly distort, or even crack, the mounting points of a Keel.
Mark
User avatar
YNWaN
Site Admin
Posts: 6797
Joined: Wed Apr 27, 2011 2:02 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by YNWaN »

A couple of points about the Tranquility. So the Tranquility modifies the bearing in a way that no amount of DLC can match. In fact, the only way to achieve the same (unintended pun alert) impact is to use a much lighter platter and that will obviously have other implications.

The Tranquility releases load on the point of the bearing by partially supporting the platter on a compressed magnetic field. By doing so it moves the majority of the supported load directly to the chassis and bypasses the thrust pad. The overall load is still the same but the thrust pad now carries very little of it. If you reduce the load on a bearing you increase its life and reduce noise - this is fundamental engineering fact. In addition the Tranquility provides lateral stability by dint of its larger diameter and its positioning close to the platters COG.

The Karousel may (but I don’t guarantee it) be quieter than the Cirkus bearing but it won’t be as quiet as a Tranquility equipped Cirkus because the loads are unchanged and so is the fundamental design.
Mark
User avatar
YNWaN
Site Admin
Posts: 6797
Joined: Wed Apr 27, 2011 2:02 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by YNWaN »

So let us look at ball race bearings. These are very good at high speed and high load. Basically, two tracks (races) rotate relative to each other and the load is transferred by a number of balls coupling the two race surfaces. Ideally, the balls rotate, rather than slide, and this reduces friction compared to a sleeve bearing where the surfaces slide against each other. However, for a ball race bearing to work there must be slack in the system otherwise the the balls will not rotate and the bearing will be locked. 6 the same token, too much free p,aBy encourages wear and increases noise. For this reason ball race bearing often have a pre-load applied to them (very common in tone arms). This is where the inner and outer races of the bearing are pushed in opposite directions and so reducing the free play in the assembly. This does increase friction though and, in such a bearing, there is an inherent trade off between friction and free play - this is one* of the reasons I consider them inherently unsuitable for arm bearings.

* The other reason has to do with the range of movement, or rather the lack of movement, that a tone arm goes through. When the inner race rotates relative to the outer race the balls rotate only a relative fraction. Given an arm only ever moves twenty odd degrees and never does a full rotation, it is very unlikely a ball in the any of the races even does one single rotation is its entire life. This issue is further compounded b6 the need to lubricate such bearings, pre loading friction, dust seal friction and the drying up of lubricants with time.

——-

So the Mober uses twin ball races with the inner race removed and the bearing shaft acting directly upon the balls. As there is now no inner shell it is not possible to apply a pre-load to the bearings. This means there is both metal to metal running surfaces and free play in the system which means increased noise and decreased rigidity. The reason these problems are not glaringly obvious in use is because of the extremely slow surface velocity of the bearing (how slowly it turns) and the very low side load the bearing experiences.
Mark
al2002
Posts: 201
Joined: Sun Jun 26, 2016 4:09 pm

Re: LP12 main bearings: Cirkus + Tranquility, Mober, Karousel

Post by al2002 »

“1 and 4 are inherently linked and pretty much always overlooked by those discussing the correct oil to use in a turntable bearing. In essence, the smaller the gap the ticker the oil or, the smaller the gap the thinner the oil”

In your post above, there are typos. I believe you wanted to write “ In essence, the larger the gap the thicker the oil or, the smaller the gap the thinner the oil”
Post Reply