DIY Turntable

I am just in the planning stages of building one too. This is my suggestion. For the rotating platter, consider using a flywheel from a manual transmission car. Several reasons.

Cheap
Readily available
Heavy e.g. 11kg for 6 cyl one
Already balanced
Has gear teeth on perimeter for speed sensing or strobe (137 on mine)
Usually has flat area next to gear for drive belt to go
Holes already drilled around centre with great accuracy

For a drive motor I am thinking of using either a dc servo motor out of an old HP Deskjet. This has virtually zero cogging effect (lumpy torque), and has an inbuilt 400 pulse per rev rotary encoder disk, or perhaps an old IDE hard disk with a flywheel glued to the disk and a spindle out the centre. The motor in these is a 3-phase synchronous type so no speed feedback necessary.
Have a look also at a vcr capstan drive motor. They sometimes have an integral flywheel and speed sensor.

Let us know how you go.

GP.
 
I would also recommend checking out Redpoint Audio Design. They are the later iteration of the Teres project...

I've spoken to Thom a couple of times and he is absolutely great! I am basing my table on their bearing and platter... Thom can give you a run-down of what has proven to work and what hasn't... I also like the idea of a vinyl platter for a better impedance coupling to the vinyl record...

Also take a look at their bearing... About the nicest that I've seen...

They can be found here:

http://www.redpoint-audio-design.com/index.html

Good luck,

Steve
 
GP,

Cheap Readily available Heavy e.g. 11kg for 6 cyl one Already balanced Has gear teeth on perimeter for speed sensing or strobe (137 on mine) Usually has flat area next to gear for drive belt to go Holes already drilled around centre with great accuracy

great idea! nothing beats rotatioanl inertia. Except more of it. :) Any further idea how to combine that flywheel on top with a thick layer of PVC/Vinyl?

For a drive motor I am thinking of using either a dc servo motor out of an old HP Deskjet. This has virtually zero cogging effect (lumpy torque), and has an inbuilt 400 pulse per rev rotary encoder disk, or perhaps an old IDE hard disk with a flywheel glued to the disk and a spindle out the centre. The motor in these is a 3-phase synchronous type so no speed feedback necessary. Have a look also at a vcr capstan drive motor. They sometimes have an integral flywheel and speed sensor.

if the servo motor you mention is a mechanically commutated DC motor (similar to the maxon), go with it! Low-to-zero torque ripple make live life easy here! Moreover, this sort of motor is an asynchronous one. Means: no "rotational spring constant" like with synchonous motors. TTs with synchronous motor belt drive are 2nd order system in best cases (if the belt is non-elastic, stiff against extending) whereas DC-motor belt drive is a 1st order system hence incapable of rotational oscillations.

If you should need such a motor, i have some left, you could buy one or two from me.

Hint: keep the rotational inertia of motor low, the one of the platter high. The motor is not meant to store rotational energy but the platter is.

Another hint: steer clear of ball bearings for the platter spindle, no matter how high the temptation is (i.e. you get the spindle of an old monster harddisk for free). Stick with plain bearings. With friction bearings. By any means!!

And yet another hint: stick with non-elastic belts. Use magnetic tape, does a fine job here. And you have to compensate for the belt's lack of elasticity, the motor has to sit on a sled or a lever pulling moderately outwards.

Have fun!
 
Not sure what to put on top of the platter yet. Yep, the motor is a commutator type made by Buhler or Buehler, can't remember how you spell it. I nevere thought about oscillations in a synchronous motor in this application. Thanks! Actually I have read that in sizes of >1000 hp this oscillation thing can become quite an issue.

Bearings, yeah. At first I was trying to use a big stepper motor and go the direct drive route (did I hear someone fainting there ? ;) ) but even after lots of fiddling with the drive waveform harmonics I still couldn't get a satisfactory result so I have ditched that idea.

Belts stretchiness is also someting I had never imagined would cause a problem but if the drive motor had lots of iertia as well, then I can see a disaster just waiting to happen.

Here is the text I posted to a newsgroup recently of my adventure a long time ago with a turntable belt : "Twenty-one years ago, having a good hifi was *the* most important thing as far as I was concerned. So I went out and bought myself my first ever belt-drive turntable. (they were for records, son). Anyway, it ran just fine, and here I was one day just feasting my ears and eyes listening to, and watching it play. It had written on the front "FG servo" which I assumed meant that it had some kind of oscillator driven motor and an incremental encoder of sorts on the platter for feedback. There was also a neon stroboscope next to the rim of the platter for messing around with speed adjustment if you wanted. I discovered this day that if you slowed down the rotation of the platter with your finger, the speed would dip a little then pull back up to what it was. Conversely, when you lifted your finger the speed would rise momentarily and then drift back to the set point. I would dab my finger on and off while watching in childlike wonder the transient response of this feedback loop while contemplating the complexities of the proportional, integral and derivative factors of this beautiful servomechanism. Oh joy! Then one fateful day I decided I just had to see what was inside this awe inspiring machine, so I opened it up to see and what a letdown! Gasp! There was no incremental encoder driven from the platter. It turned out that what I imagined was the undershoot and overshoot of some fancy feedback loop was merely the stretching and contracting of the drive belt under a changing load. D'oh! If only I had left it unopened...

GP.
 
GP,
(did I hear someone fainting there ? )
no, just vomitting. :)
(serious now) if a direct drive i designed in a proper way, it can be terrific. All the designer has to care for is attaching proper wall thicknesses to the bearing bushing and shaft and see that no space lcaimed by the bearing is usd by the motor. Quite a task, but not impossible.

But usually DD is used for budget designs.

Belts stretchiness is also someting I had never imagined would cause a problem but if the drive motor had lots of iertia as well, then I can see a disaster just waiting to happen.
even with comparatively low inertias,
a stretchy belt eats up µdynamix, detail resolution, pace rhythm and timing. Take my word on it!
 
Maybe a small fluid coupling between the motor and it's pulley would damp any bad effects from belt stretchiness the same way a manual car can go jerk-a-jerk-a-jerk sometimes but an auto tranny one just won't because the torque converter slips and damps these oscillations.

You can get big ones for industrial motors that have to start high-inertia loads, but I have never heard of a flea-power one. Would be interesting I think.

GP
 
AuroraB,

IMO a belt should not be sticky, quite the opposite. Look at it, any fast motor speed variation, oscillation, whatever, drowns in the belt being slippery.

Moreover, stickyness is not as uniform as you may wish it. Causing something equivalent to torque variations by this.

Have the belt slippery and under well controlled tension.
 
It probably would. I have never checked the external field of a cartridge but I would have though for a well designed one the flux would have been almost completely self-contained. If it is not then there may indeed be problems. I'll get a very thin steel feeler gauge and wave it underneath my cartridge and see if it is attracted. But first I better see if it is not stainless and therefore magnetic!

GP.
 
Wuffwaff,

thanx alot for reminding us,
indeed a magnetic platter would influence an MC cartridge.

GP,
no, the flux is not self-contained at all, on the contrary! Best MC cartridges have non-magnetic housings or are are naked, they have no shielding for sure, and very strong magnets, SaCo or Neodymium or similar.
The magnetic field is shaped like an apple, highly non-linear and with heaps of stray field; in the region of the coils the field still is nonlinear but, by careful adjustment, symmetric.

If the platter surface is magnetic, the MC cart simply sits down on its *** and the cantilever goes CRACK!! (i once tried to find out how much spacing to record surface a used MC still had and used a 1.0mm drill bit for it --- SNAP!! the drill bit slipped out of my fingers and snapped onto the cartridge body, just oh just missing the cantilever. Phew, had it hit the cantilever, there wouldn't have been any anymore. And the drill bit was hard to remove.

So what can you do? You can put a very thick layer (say, 4", say100mm !?) of vinyl on top of your flywheel, maintaining a proper distance between cartridge magnet and magnetic part of the platter. And get sonic benefits from the mechanical impedance coupling (between record and platter) the vinyl provides.

You'll get gorgeous results! and very low surface noise.