It's certainly something that I've been thinking of when I've been scheming my own magnet designs.jrevillug said:
Yes, but it has been calculated that the smallest groove variation on an LP is the same size as the wavelength of light- are you THAT sure that there is no movement? Can you even see the vibrations of your stylus? Just a thought.
It might be cartridge dependent too, and (obviously) a non-magnetic cartridge body would have to be preferred here. Of course there will always be a magnet inside the cart. and we might find that MM's show a higher sensitivity to being affected than Moving Coil's (?)
(Presuming there is any affectation at all)
having seen and heard Ynoans deck i'm pretty sure theres no cogging, the deck is rock solid stability wise with no audible flutter or pitch distortion.
i think the trick is the fact that the float is so small, only a few mm.
and it's definately quieter from a bearing point noise point of view than any ball, thrust plate or inverted bearing i've sat next to...
the side load from the drive belt pails into insiginificance compared with the weight of the platter on the bearing shaft.
i think the trick is the fact that the float is so small, only a few mm.
and it's definately quieter from a bearing point noise point of view than any ball, thrust plate or inverted bearing i've sat next to...
the side load from the drive belt pails into insiginificance compared with the weight of the platter on the bearing shaft.
BE ADVISED, ALL INFORMATION IN THIS POST IS AUDIO TECHNICAL WANKERY, AND IN NO WAY BASED ON SOLID OR EVEN TESTED FACT. 
INTENDED SOLEY FOR 'Hmmmm' CONSIDERATION
Yep, that's what I was looking at as well. I like the gigantic spherical magnet, man they look fun! I'm sure I could invent a use for that, necessitating it's purchase. The rings magnets I saw were a bit small though. Ideally you'd want one that was the same size, or bigger, than the platter's diameter to offer the highest stability. Because obviously if it's smaller, it's more and more like balancing a plate on your finger.
Good point re: modifying the force, you'd want to be careful selecting if you just went with two ring magnets. It should be possible to calculate the required magnet strength with some calculator time.
In a perfect system, everything but the stylus would be rigid. Anything less than that means increasing the degree of signal aborbed in the compliance of the moving parts. The only reason for the tonearm not being rigid is so that it can move. Obviously, we need to be realistic here....
Cogging is likely low even with separate magnets because you have quite a lot of them and the platter's mass will require quite a bit of force to accelerate up and down. With the platter spinning, the up and down cogs of the individual magnets will be averaged by the platter mass, in the same way a capacitor averages the bumps on rectified AC.
You're right about an electromagnet system. If you have one on the underside of the platter, you also need some way of power it. Depending on the guage of wire used to wrap it and the way it's wrapped, neither is it a perfectly homogenous field. E.g. imagine wrapped a one layer deep coil with copper pipe. Now one with tens or hundreds of depth layers from ultra fine coiling wire. The second will produce a much more even field. I started doing the work to check the currents and wire sizes needed for a CD sized magnet to lift a platter but went to bed instead, I've just had a weeks worth of maths and physics work every night.
You're also correct on air bearings, they're not fundamentally better, just a comparable I'd guess.
Good point about the positioning of the magnetics under the label, that should help. And also entirely contradicts my stability point about wider magnetics. So that's one contradiction in design terms.
Yep. I'd do away with trying to watch for changes in height and just see if you can measure them using the cartridge. One way to do that would be to excite some kind of vertical motion on the platter. So you'd play a groove with vertical modulation without the magnetics and look at the skew rate and amplitude on a scope. Then put the magnetics in place and look again. If any of the signal is being absorbed in the magnetic field, the skew rate and amplitude of the signal on the scope will decrease.
That's a great idea! I got super excited when I read this. Although, from what I understand, one of the few magnetic materials that can survive long term with it's poles in close opposition is neodymium. I think a lot of speakers use ceramics or something like that right? So I'd want to check about life time of the bearing if I used old speaker magnets.
Some of the newest, high power speakers might have neodymium in them I guess.
Which is one of the plus points for an electromagnet, that it doesn't wear down. Another consideration could be a composite. Pricey ring magnet on the platter underside, electromagnet coil on the plinth.
Re: weird vibrations... I've been thinking that my design might have the motor and plinth arrangement hanging on springs and critically damped. So I'd have the motor damped at the frequency it produces most noise for example. I'm just wondering if lead will be strong enough to make the platter from if it's an inch or two thick.... I don't want it to start sagging or warping.
INTENDED SOLEY FOR 'Hmmmm' CONSIDERATIONVinyl-Addict said:Bismuth has been mentioned as a good shielding material. Here is a source in the US for ring magnets.
http://www.unitednuclear.com/magnets.htm
Yep, that's what I was looking at as well. I like the gigantic spherical magnet, man they look fun! I'm sure I could invent a use for that, necessitating it's purchase. The rings magnets I saw were a bit small though. Ideally you'd want one that was the same size, or bigger, than the platter's diameter to offer the highest stability. Because obviously if it's smaller, it's more and more like balancing a plate on your finger.
YNWOAN said:
Good point re: modifying the force, you'd want to be careful selecting if you just went with two ring magnets. It should be possible to calculate the required magnet strength with some calculator time.
In a perfect system, everything but the stylus would be rigid. Anything less than that means increasing the degree of signal aborbed in the compliance of the moving parts. The only reason for the tonearm not being rigid is so that it can move. Obviously, we need to be realistic here....
Cogging is likely low even with separate magnets because you have quite a lot of them and the platter's mass will require quite a bit of force to accelerate up and down. With the platter spinning, the up and down cogs of the individual magnets will be averaged by the platter mass, in the same way a capacitor averages the bumps on rectified AC.
You're right about an electromagnet system. If you have one on the underside of the platter, you also need some way of power it. Depending on the guage of wire used to wrap it and the way it's wrapped, neither is it a perfectly homogenous field. E.g. imagine wrapped a one layer deep coil with copper pipe. Now one with tens or hundreds of depth layers from ultra fine coiling wire. The second will produce a much more even field. I started doing the work to check the currents and wire sizes needed for a CD sized magnet to lift a platter but went to bed instead, I've just had a weeks worth of maths and physics work every night.
You're also correct on air bearings, they're not fundamentally better, just a comparable I'd guess.
Good point about the positioning of the magnetics under the label, that should help. And also entirely contradicts my stability point about wider magnetics. So that's one contradiction in design terms.
jrevillug said:
Yes, but it has been calculated that the smallest groove variation on an LP is the same size as the wavelength of light- are you THAT sure that there is no movement? Can you even see the vibrations of your stylus? Just a thought.
I was thinking about this today, and I agree with eeka chu that a rigid bearing providing the location for the platter, with (permanent) magnets taking 99% of the weight is the way to go. Although with the electromagnet the thrust plate would take a hell of a beating when the magnet is off, especially if there was a power cut while listening to your 78s.
Yep. I'd do away with trying to watch for changes in height and just see if you can measure them using the cartridge. One way to do that would be to excite some kind of vertical motion on the platter. So you'd play a groove with vertical modulation without the magnetics and look at the skew rate and amplitude on a scope. Then put the magnetics in place and look again. If any of the signal is being absorbed in the magnetic field, the skew rate and amplitude of the signal on the scope will decrease.
AuroraB said:
That's a great idea! I got super excited when I read this. Although, from what I understand, one of the few magnetic materials that can survive long term with it's poles in close opposition is neodymium. I think a lot of speakers use ceramics or something like that right? So I'd want to check about life time of the bearing if I used old speaker magnets.
Some of the newest, high power speakers might have neodymium in them I guess.
Which is one of the plus points for an electromagnet, that it doesn't wear down. Another consideration could be a composite. Pricey ring magnet on the platter underside, electromagnet coil on the plinth.
Re: weird vibrations... I've been thinking that my design might have the motor and plinth arrangement hanging on springs and critically damped. So I'd have the motor damped at the frequency it produces most noise for example. I'm just wondering if lead will be strong enough to make the platter from if it's an inch or two thick.... I don't want it to start sagging or warping.
Before I built the prototype I have posted photographs of I did in fact build a more simple version to check that the principle would work. The earlier experiment used the same magnets but half as many were used and they were spaced apart by over double the distance (a gap of 7mm existed between each magnet). This first bearing was made so that both halves would rotate on a shaft. No cogging effect was noted even when the bearing was spinning at very low speeds. If a cogging problem exists it will manifest itself as the bearing comes to a halt and when no flywheel is used; the first prototpe weighed only a few ounces.
YNWOAN said:
How about using one ring magnet on the stationary surface and multiple smaller magnets (as you chose) on the moving part of the bearing? Wouldn't this approach improve or eliminate cogging?
FWIW, My current platter weighs 36 pounds so I'd like to find a magnet structure to elevate/repell this weight.
Here is a calculator for disc (not ring) magnets.
http://www.magnetsales.com/Design/Calc_filles/Pullandpushbetween2Discmagnets.asp
http://www.engconcepts.net/List_Of_Ring_Magnets.asp
another source of ring magnets. the R1410 ring near the bottom of the page seems ideal, 4.5" diameter.
another source of ring magnets. the R1410 ring near the bottom of the page seems ideal, 4.5" diameter.
YNWOAN said:
At first, I'd been planning to find some wide OD pipe from somewhere, weld some plate over a 1" length of it and then pour lead into that, then machine it down to size. The pipe and plate would then form something easier to machine to tolerance and less likely to get scratched and so forth. However, finding someone who can give me an off cut of pipe in those dimensions would be difficult. If I wanted stainless steel, it'd only get harder (and probably less likely to get it for free from the off cuts bin). Maybe I could use a section from a suitably sized gas cylinder... lots of work though.
In terms of playback, lead should theoretically be one of the best choices I would have thought. The harder and lighter something is, the more prone it is to exictation. The softer and heavier, the less.
I was thinking of lead mainly because I literally have a bin full of it and a crucible spare.
Vinyl-Addict said:
That approach would eliminate it and allow for changing the field strength. Myself, I'm not planning to fully float the platter. I'll just put one ring magnet in the platter, one in the base and then adjust the base one closer and closer until most of the platters weight is off the taper bearing. That way, you can use AAA precision bearings, which should be absolutely smoking.... very hard to find high precision bearings on eBay, so that'd have to be a list price purchase. The lower the load on the bearing, the cheaper it is.
Not fully floating the platter also has the advantage that any errors in aligning the magnets parallel to each other won't be directly represented as vertical errors in the platter position as it spins (bobbing up and down), it will only change the coupling of the platter to the bearing and possibly noise production as the loading on the bearing changes throughout the rotation.
Cool link... I've been trying to calculate how strong a magnet is required. Unfortunately, a lot of the hobbyist magnet sites only grade magnets as 'strong, stronger, ultra strong'. Email a lot of these sites asking for ampere / m^2 ratings and you'll probably get nothing back.
The repulsion force at contact should be reasonably close to the force needed to separate the magnets, which is easier for most people to measure. The field decays 8 times for every one unit of distance (from memory).
One problem you would still have with this arrangement is that you still have individual magnetic elements in the platter, spinning under the pickup coil in the cartridge (depending on where you put them on its radius of coarse). In effect, that's a ripple generator. Having both as solid rings would remove that. Whether or not it's an issue is another thing. It should be fairly easy to measure if it's having an effect by simply watching the cartridges output on a scope for an AC signal at the frequency the magnets in the platter are spinning past the cartridge.
If you use a taper bearing on the platter, you could also use a disc magnet dropped into the centre of the bearing.
Tuukka said:
Yep, I saw this site too, they do have quite a lot on there. That big one would win on stability, but as YNWOAN mentioned earlier, having a small OD also means that the magnetic field it produces is kept further away from the cartridge.
I was trying to guestimate the required magnet strength I might need to lift most of the platter weight off the bearing and remembered seeing ring magnets an inch or so in diameter on eBay that could lift 145kg for about £25 I think. So two together should stick with ~290kg of force, which means they should repell with roughly that much on contact with each other. Since a platter weighs 10 - 20kg, I'd guess magnets like those would repell the platter by quite a considerable distance. If you're only trying to lift some of the weight off the bearings, you'd either need to separate them by quite a lot (good because they're even less likely to depolarise over time) or you could just go with weaker magnets.
Why do you feel that a taper bearing is the way to go? I would have tought that it is way too noisy for this aplication. The purpose of the magnetic bearing (for me) is to remove the grinding effect of the ball on the thrust pad in a more conventional turntable bearing. The sleeve and shaft arrangement used by almost all turntables is already hugely quieter (and much lower friction) than a taper bearing and I still feel that it is too noisy. I can't help but feel that you need to do some real prototypes.
YNWOAN said:
I would be worried about noise and you're right that hands on doing is required!
There are just so many things to make mistakes on and wish you could go back on that I want to consider as much as possible before I start buying anything... I'm a student as well, which means £££s is a premium. The lead idea is purely so I don't have to buy plates of steel or aluminium. I also have a more than unhealthy urge to cast some of it.
Bearing wise, I'm planning to use something like one of Gamet's precision taper bearings. These beauties are so well made you have to warm them up to cup of tea temperatures to slide them into place without damaging them. So I'm hoping they should also proove to be virtually silent at ~1 RPS. Although, I need to check the price of them because it could be frighteningly expensive.
http://www.gmnbt.com/gamet_bearings.htm
The radial bearing in your own table will be generating noise, but barely any because there is next to no radial load on it. The only load on your radial bearing will be from platter balancing imperfections and the pull from the drive mechanism. By magnetically lifting 99% of the weight off a taper bearing, a similar effect would be produced.
Why a taper bearing? The design I saw used a thrust plate and radial bearing. A taper bearing does both.
The thrust / radial bearing would be better in terms of stability because the thrust plate can be made to have a massive diameter, so the platter absolutely cannot tilt. But the tolerances on bearings goes down as size goes up (the manufacturers even publish specifications for it and it makes sense, it's not audiophile estimation), so bearing noise goes up as size does. You also need to be sure you have the thrust bearing precisely centred under the platter to avoid weird dragging effects due to speed differences about it's circumfrence (think about how the inside tyre of a car goes slower than the outer one when you go round a corner, because the two aren't exactly the same distance from the centre of the turning circle) and now have two sources for noise, the thrust and radial bearing.
I just chose a taper bearing because it seemed slightly cheaper and less noisey by comparison with the thrust and radial bearing design I'd seen, with the only drawback being that it's more like the plate on a finger example.
The bearing in my design would be next to useless in terms of load carrying. I'd set the platter to be almost floating, like yours, only just about resting on the bearing with a few N of force. That means that any forces developed by the modulation on the groove only has the play of the bearing to push down through (the play in its tolerance).
I'm trying to get access to our university's machine shop at the moment. I'm an electronic engineering student, but even still... they're not really used to having people in there I think, it's for the techicians to make stuff for the department. I keep walking past and drooling at all the stuff in there. It's about the size of a two semi detatched houses and packed with lathes, mills, saws etc... If they'll let me in there and it's open during the holidays, I might stay here over some of my holidays and work on it in the quiet.
taper bearing as in cartridge bearing? 
no 3 tier bearing will be silent, ever,no matter how well made.
thats why nearly all the turntable manufacturers use a sleeve bearing. they are massy,with very high surface area and as such quiet within their specified load.
a 3 tier bearing will never be as quiet if machine dto the same tolerances.
you might be better of making a wide sleeve bearing with ph'bronze centre and vesconite sleeve. make the bearing as wide as it is tall to maximise resistance to axial loads.
no 3 tier bearing will be silent, ever,no matter how well made.
thats why nearly all the turntable manufacturers use a sleeve bearing. they are massy,with very high surface area and as such quiet within their specified load.
a 3 tier bearing will never be as quiet if machine dto the same tolerances.
you might be better of making a wide sleeve bearing with ph'bronze centre and vesconite sleeve. make the bearing as wide as it is tall to maximise resistance to axial loads.
If I may be a little more direct, I would say that a taper bearing is completely inappropriate for a platter bearing. The stylus/cartridge rests directly on the platter (obviously) and the platter is directly linked to the main bearing. My own experiments indicate that the main bearing is the source of a significant degree of noise and this is assuming a sleeve and point contact form of bearing. I have used taper bearings a number of times in different applications and can assure you that they are much noisier than the sleeve and point bearing - they also exhibit much greater levels of friction. Please note - all bearings that consist of one surface rubbing against another will produce friction and noise - this cannot be avoided - but it can be minimised.
If the platter is properly balanced absolute stability is not an issue if the bearing is made to sufficiently high tolerance.
"Forces developed by the modulation of the groove" are inconsequential with respect to the main bearing of the platter.
A radial bearing consists of many smaller bearings (rollers) arranged in a radial pattern forming one larger bearing. Each of these smaller bearings will generate noise and friction. Attempting to lift most of the load from such a bearing would be an attempt to limit the inherent limitations of this type of bearing.
If the platter is properly balanced absolute stability is not an issue if the bearing is made to sufficiently high tolerance.
"Forces developed by the modulation of the groove" are inconsequential with respect to the main bearing of the platter.
A radial bearing consists of many smaller bearings (rollers) arranged in a radial pattern forming one larger bearing. Each of these smaller bearings will generate noise and friction. Attempting to lift most of the load from such a bearing would be an attempt to limit the inherent limitations of this type of bearing.
While I'm certainly willing to listen, noise is caused by vibration, and vibration can only occur when something is a loose fit and there's a force acting on it. I'm talking about using bearings specified to microns, you don't get anywhere near those tolerances with a normal lathe turned part like a sleeve. Or even a regular bearing.
Higher bearing load rating is always expensive or troublesome. If you can remove most of it with a few cheap magnets, lifting most of the load from the bearing simply makes sense. It also removes most of the force that can cause vibrations through the tolerance of the part.
I have seen tables for $10K+ with roller / ball bearings in them. One of which includes a bearing heater to expand the bearing to the correct tolerances.
Higher bearing load rating is always expensive or troublesome. If you can remove most of it with a few cheap magnets, lifting most of the load from the bearing simply makes sense. It also removes most of the force that can cause vibrations through the tolerance of the part.
I have seen tables for $10K+ with roller / ball bearings in them. One of which includes a bearing heater to expand the bearing to the correct tolerances.
taper bearings specified to microns, is that single microns or 10's of microns?
consider for a moment the size of the rollers in a taper bearing. lets say 5mm in length and 2-3mm in diameter, now take its tolerance and factor that in to create an angular error across it's length.
now do the same for a typical sleeve/ball turntable bearing thats 20x as long built to the same tolerance,or even 10 times worse tolerance.
which has the least angular error? least rocking and hence least noise? and that's before you factor in that all 20-50 roller parts need to be 'identical' not to introduce further error.
and then of course the reason detre' of taper bearings is that they are less noisy under rated load than when used under say 1/10 load. they are just totally unsuited to use as turntable bearings where the load is reduced through levitation.
i don't doubt that somewhere someone will be using the very highest spec taper bearings in a turntable application, but they'll be having to design around the problems it introduces.
i mean heating, come on why bother,just use a vesconite sleeve and oil impreg' phosphor bronze.
consider for a moment the size of the rollers in a taper bearing. lets say 5mm in length and 2-3mm in diameter, now take its tolerance and factor that in to create an angular error across it's length.
now do the same for a typical sleeve/ball turntable bearing thats 20x as long built to the same tolerance,or even 10 times worse tolerance.
which has the least angular error? least rocking and hence least noise? and that's before you factor in that all 20-50 roller parts need to be 'identical' not to introduce further error.
and then of course the reason detre' of taper bearings is that they are less noisy under rated load than when used under say 1/10 load. they are just totally unsuited to use as turntable bearings where the load is reduced through levitation.
i don't doubt that somewhere someone will be using the very highest spec taper bearings in a turntable application, but they'll be having to design around the problems it introduces.
i mean heating, come on why bother,just use a vesconite sleeve and oil impreg' phosphor bronze.
Sleeve bearings arent just lathe turned, there then precision ground and polished and are also specified in microns.
Noise can be created by friction, not just from a loose fit.
A roller bearing (or ball race bearing) has many more parts to create problems, and a taper bearing relies on the needles being in contact with the inner and outer races.
A well made sleeve bearing doesnt actually touch metal to metal at all as there is a thin film of oil between the two (not really possible with a taper needle race bearing).
In a good sleeve bearing the only metal to metal contact is at the ball (i use saphire) that is MASSIVELY smaller contact area than a taper needle bearing.
Im afraid i agree, but by all means try if you must. The world would be nowhere without people trying
In my opinion taper bearings should stay in car wheel hubs.
Didnt see the above reply, i think were both getting at the same point
Noise can be created by friction, not just from a loose fit.
A roller bearing (or ball race bearing) has many more parts to create problems, and a taper bearing relies on the needles being in contact with the inner and outer races.
A well made sleeve bearing doesnt actually touch metal to metal at all as there is a thin film of oil between the two (not really possible with a taper needle race bearing).
In a good sleeve bearing the only metal to metal contact is at the ball (i use saphire) that is MASSIVELY smaller contact area than a taper needle bearing.
Im afraid i agree, but by all means try if you must. The world would be nowhere without people trying
In my opinion taper bearings should stay in car wheel hubs.
Didnt see the above reply, i think were both getting at the same point
The tolerances are usually given as TIR (total indicated runout), which means that all the rollers, races and cups have already been taken into account. If I'm remembering properly.
hey, at least you get somewhere to heatsink regulators to...
Where abouts did you get the sapphire ball from?
I may answer back a lot, but only in an attempt to sift the truth out of the endless bins of audiophile myth I'm now accustom to. I'm not stupid enough to flat out ignore advice from people who've done it before.
And will probably end up with something very close to YNWOAN's table myself. I was thinking on the way back from my maths lecture this morning, it should actually be possible to quantify the vertical deflection of the platter in respect to a load placed on it.
Imagine two platters. One is made of MDF and needs next to no magnets under it to levitate it 5mm. The other weights 30kg say, and has a gigantic magnet under it to achieve 5mm of separation.
The first will have a field under it that's much more compliant than the latter, so more of the force intended for the stylus would be bypassed by the compliance of the field.
Theoretically at least, it should be possible to use some maths to work out values for the effect, since you can measure the separation and platter weight to determine what the strength of the field supporting it must be. I've not looked into that in much detail but I expect it might get involved. I may see if my maths lecturer can help me with that since he loves differential equations and integration.
Funnily enough, I don't think there would be a difference at all; the period of oscillation, and therefore its interaction with the arm/cartridge compliance would be the SAME.
This is because the period is constant for a given deflection of a mass on a spring - it doesn't actually matter what the stiffness of the spring is, because the ratio of all possible mass:spring stiffnesses is the same for a given deflection. This falls out of simple harmonic oscillator theory, not a lot of calculus required
BTW The stiffness of magnetic repulsion varies as inverse-cube of deflection IIRC. Shouldn't alter the result above though.
I got the saphire ball from here
http://www.dejaydistribution.co.uk/balls8.html.
They do all sorts, minimum order was £25 if i remember right.
I bought a ceramic too, and a few others for other people to make the numbers up.
Cant remember price, but around £5 each.
http://www.dejaydistribution.co.uk/balls8.html.
They do all sorts, minimum order was £25 if i remember right.
I bought a ceramic too, and a few others for other people to make the numbers up.
Cant remember price, but around £5 each.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.