for a large mass magnetically suspended t/t platter!
Written by Christos P. Skaloumbakas / 09 Dec 03This paper is mainly devoted to an extreme upgrade of a high quality turntable and I would urge people to give much thought to such kind of modifications, since it is possible that they would end-up worse than it was before...
Having said that, I am giving below most of the technicalities of the flywheel design and...
I am taking the chance to review the whole analogue set-up, covering the following subjects/topics:
Since, I will be talking below, about:
- moment of inertia (or rotational inertia) and
- kinetic energy (rot)...
I will like to remind from physics that:
- Moment of inertia is defined as the resistance to any change in rotation of a body.
- Just as mass resists in any change in linear motion, so does moment of inertia resist in any speeding up or slowing down of rotational motion.
- Kinetic energy is the product of the moment of inertia, multiplied by rotational speed.
Since, what we are aiming at, is not figures in absolute units but maximizing these figures, absolute figures may not be correct but relations among listed analogue designs as far as inertia and energy are true. (Some interesting stuff about theory & lab measurements can be found here)
Turntable's current state and description
Parts description
Platter - Chassis - Magnetic field
Platter & chassis are described in paper02.htm and no 'deviations' from the original have been applied (obviously...). Only the platter's spindle has been modified by putting a couple of small neodymium ring magnets on its bottom to add a little bit more repulsion needed to contemplate for the extra weight of the disc-o-ring used on the platter. These 2 magnets act against a plate magnet put on the top of the sub-chassis axis. In addition, another small ring magnet is inserted on the top of the spindle, so oriented to attract the redesigned record clamp. More on that, in the record clamp section...
Another change was the replacement of the WD-40 used for lubrication (recommended by the manufacturer) by an extra light mil-spec oil used for the flywheel shaft, as described later, thing that reduced even more the friction between the shaft and the bearing of the platter. This excellent oil is as thin as the WD40 but with lubrication properties and it stays rather than drains out after a while, as the WD-40 does. Think of it: From 33.33 RPM rotation, when not connected with the motor, the platter takes 9 minutes to stop instead of 3.5 needed before!
Some words about the magnetic field:
There are only 3 t/t manufacturers in the world using magnetic repulsion for platter suspension: aas gabriel from Switzerland, Verdier from France and TL Acoustics from Hellas - Greece (+ possibly another one or two DIY'ers). The reason they are so few I believe is because some people still believe that the magnetic field used to levitate the platter is not safe for cartridges!!! Only from the fact that I am using a 10,000 Euro cartridge on such a 'dangerous' environment should prove enough but just for enjoying myself, I conducted a small experiment:
I placed on the bare platter a very strong Neodymium rod, a magnetic steel ball and my Insider 'reference' cartridge. Only with the first item, moving it along the surface of the platter you slightly feel the field. The ball can be placed and hold still wherever I put it without feeling anything at all and the cartridge has been playing music for years without any problem... For your info, moving coil cartridges have their small magnets (possibly 1000 times weaker than the rod in the picture) oriented perpendicular to the arm and this is the orientation the red magnetic rod preferred to rest on the platter during this small experiment!
OTOH, even with the almost complete absence of bearing friction, the platter (after the belts taken off) remains still at any position, floating perfectly without any rotary movement (preference to a special position) what so ever! And when it rotates, no even the slightest vertical movement or vibration along the axis is exhibited, reproducing the cleanest and the mostly rumble and wow and flatter free music one could dream of... So, no need for more discussion on the whole subject...
Power supply
The power supply hasn't changed. The new aas gabriel t/t model uses a 65VA/12V car battery but I reckon the small one is adequate enough for powering-up the motor, (even after the addition of the huge flywheel), for at least 10 hours of continuous battery operation of the rotating system, before I have to recharge it. For this 10 hours period, it gives me plenty of good voltage range of 12.8V to 12V, (from the on-line-rotating 13.25V, reducing to 12.8 in 3 min., from 12.8 to 12.7 in 80 minutes and then decreasing linearly to 12 Volts), to work with - listen to the music, for a whole evening listening session.
If time comes and motor has to be replaced for any reason, I will either upgrade it to the aas gabriel current p.su. model paired with a new motor, or find another solution. (i.e. Maxon strong motor paired with a Maxon battery operated p.su. - controller - digital encoder etc.)
Motor - Arm base
The motor, as mentioned, remains the original. It proved very strong in comparison with the currently available Maxon motors, which they were thought to be a replacement for it, if excessive friction was encountered. I do not know what is going to happen in the future though...
Motor spindle has passed through many changes (it is the smallest part and the most easy to modify anyway...). You can see 3 different shapes in the 3 pictures on the right. The first one is the original and the second one has been used for almost a year just to have the illusion that I am using some flywheel effect until I build the big one. It was not a bad at all and in connection with the cassette tape used as a belt instead of a thread, it gave me more punch to the music.
The third motor spindle configuration is the one currently used, with one drive at the beginning of the flywheel project and 2 drives when the flywheel took its final heavier shape, absolutely essential, otherwise slipping occurred!
The arm base, also described in paper#02 has been modified, regarding the placement of it on Roller Blocks in the past, (see paper#12). This paper is no valid now as far as this. With the inclusion of the flywheel, I thought the arm base should've been heavier and put on spikes for extra stabilization and rigidity. Three stainless steel cylinders have been screwed on the bottom of the base and 3 blind threads have been opened (one/cylinder) to accept spikes similar to the ones used for the huge turntable sub-chassis by the manufacturer. So now, the arm base consists of 3 materials: 3 stainless steel cylinders on the bottom, one Lead cylinder in the middle and 'Permalin' on the top. I am sure; this caused a large improvement to the reproduced sound...
Motor Housing
The motor housing, also described in the paper#02, is placed on a lead base and hasn't been modified. The only thing that is modified is its 3 threaded spikes, which replaced the silicon pads, originally used and the Roller Blocks placed under later (paper#12). But what has deeply changed and which redefined what complete silence in music means, is the place the motor was sited.
From the beginning the motor was sited on a separate shelve, independent of the shelve the sub-chassis/platter was placed on and away from it. After the flywheel first designs, it has been realised that the rotation of the flywheel was best balanced and stabilised when put in the middle, between the platter and the motor and on a straight line with them. So, since this small separate shelve has been devoted for the flywheel, I had to add another rack for the motor and rack being impossible to be placed between the left speaker and the main rack, the best solution was the idea of a tripod there.
I was lucky enough to find a very good one, which I obviously modified by increasing its weight, mainly at the bottom, and filling it-in with lead shots for resonance absorption. Now the motor was acoustically completely 'invisible' by the platter... and using the rotating swivel of the tripod, I could also optimise very accurately the threads' tension between the motor and the flywheel! Impressive isn't it?
Flywheel project
Introduction
Flywheels are not extensively used among t/t manufacturers. Out of so many t/t models I have personally entered pictures of in the web site's T/T Gallery, only 3 manufacturers use a flywheel (in their top models that is...). The pictures shown are VPI's and Forsell's top models and the Australian Aura. VPI were the first to employ such an apparatus in their TNT models and in fact the first idea about inserting a flywheel in my aas gabriel (something that sounded to me at the beginning almost ridiculous for such a heavy mass platter) was from a friend working for VPI.
Later on, as this idea was drivelling my mind, I begun to think seriously about it and started to count pros and cons of such a design.
Just to put things into prospective:
The aas gabriel alone (+ disc-o-ring but no flywheel) is producing a good 8.1, almost as high as the f-w equipped top VPI TNT. And to be honest, there are about 10 heavy platters in the t/t market, which are producing the same, more or less, momentum of inertia and none of them is employing a flywheel for extra energy storage. They do not need it or... you think they don't!!!
As I had designed my custom rack quite large to accommodate the platter and motor in separate shelves, I had the space to play with - meaning I could make a large OD flywheel and also being able to place it away from the platter, I could make the drive OD small enough, (and still have grasp), for high rotation. Large OD (or more explicitly more mass concentrated outside) and high rotation are the secrets of high kinetic energy (rotational)... and here come the cons of such designs - extreme designs that is:
Preliminary designsMy idea was to use such a rotary speed, which would not give, if something went wrong with the machining, resonance frequency that would affect the acoustical spectrum. So, I was playing with a frequency of 10Hz or 600rpm for the 45 speed. (10Hz is not heard but I know my speakers produce it). So I took this as a limit. To make things safer, I had initially thought of a nice round figure of 400rpm / 33.33RPM, which finally has been reduced to 375.
Machining & Magnets:
These were the things, which had held the flywheel project for 1½ year! In the mean time, I was playing with my fancy Excel sheet (see picture) and exchanging tons of e-mails with a a gentleman of High End called Dave Whittaker - N. Zealand, the manufacturer of Aura t/t, (see picture above), who had always a practical solution in-hand to my extreme engineering ideas. But all that on papers of course. It seemed custom Neodymium magnets that I was looking for, were very expensive to experiment with and I couldn't find the proper people to do the job. All this time, to ease my pain, I had just modified the motor spindle to a flywheel-shaped one with smaller drive, so I was working my motor with almost 2000rpm (D.W's idea) and using a cassette tape instead of a thread to drive the platter. Not bad at all - many manufacturers recommend high motor rotation anyway - and with this 'substitute', I had almost given up the intermediate flywheel idea all together!
Surfing on the INTERNET one day, (May 2003), I caught a fellow Hellenic - Greek DIYer, for the time, who was posting pictures from his t/t designs in a DIY Forum.
That was it! Him and a friend of his were just forming a Company named 'TL Acoustics' in Patras - south of Athens to start making a small production of high quality turntables (in kits and in complete set-ups). They were 2 dedicated people with great enthusiasm for analogue. Christos Livadas, an expert in handling the lathe; Konstantinos Tzivleris, a specialist in electronics, p.su.'s etc. That was what I was looking for! And the flywheel project came to life again...
We got in touch and they proposed to me to make the flywheel. They accepted many elements of my design and they put some of their own on the table. They took the magnetic suspension idea (in fact they are employing magnetic suspension in their t/t design ever since...) and they proposed shaft OD and lubrication method. They had also revealed to me one source of Neodymium magnets, a US manufacturer who proved extremely knowledgeable and helpful! Since this was in fact an experiment for me - and for them - we agreed on 'materials only' cost for the project.
Finally we agreed on the dimensions and the design details (drives, magnets, bearings etc), they sent me some AutoCad 3-D pictures to have an idea, I placed the first magnets Order and we kicked off...
Flywheel construction
As I said, this project was almost an experimental one. No apparatus of such a design has ever been tried before, we had no idea as to what extra friction we would encounter or any other problems might crop up. I couldn't take risks and this is why we thought that I should move the whole aas gabriel set-up (motor and p.su. included) from home to Patras, 250 Km south of Athens.
So, 130Kg have been dismantled, put on my car and off they went to TL's workshop. (Not a very pleasant thing to do during summer vacations). We had to have them there to watch how the whole system would perform after the flywheel addition and take the necessary correcting steps. These might have been, different motor spindle drive OD or different material used for belts or even a stronger motor, if needed.
First thing changed from the original design were the drive belts. The first flywheel was designed to accept a cassette tape from the motor and an open reel tape from the flywheel to the platter. There is much talk lately about the advantages of the ¼" reel tape for driving platters and one of the pioneers of this idea is my friend D. Whittaker who uses it from his flywheel to the platter in his Aura t/t design. I was also using a cassette tape when I had modified the motor spindle to a flywheel-look one as I have mentioned before and I had been impressed by the sound improvement!
Unfortunately, with the extra rotating weight of the flywheel and such small drive OD's, both tapes were slipping. (You can't have it all...). Fishing line has also been tried but this was also slipping a bit. Perfect material at that time was found to be the 0.8mm. wax less dental floss. So drives of the flywheel had to be machined out and one drive in and 3 drives OUT were cut to accept one string of dental floss from the motor spindle and 3 strings to the platter.
Bearings - Magnets - Oil
Second thing that proved inadequate was the shaft OD. We thought at the beginning that a small diameter of the rotating axis would benefit as far as friction and noise are concerned. But the high rotation of such large flywheel was causing a small trembling - bending of the axis which was almost eliminated but not disappeared if we put motor flywheel and platter on a straight line (something that we applied from there on any way), and there comes the need of a tripod to place the motor on, since the flywheel would take its place on the rack. So, next modification was the increase of the axis OD. Shaft and bearings had to be modified too...
Magnets and bearings configuration has also been changed due to the increase of the axis OD. The rotating f-w bell is repelled using Neodymium magnets, cleverly placed to the f-w and to the f-w body... I want to emphasize that many combinations and placement patterns have been tried in order to repel such an enormous mass, rotate at this high speed without clogging effects - achieve smooth rotation (one magnet would not drive, or rote the other one with magnetic force, if they are standing still...), etc.
The rotating axis is working in an oil bath with milspec light oil, for minimizing friction. We had agreed that the same oil should've been used for both the platter and the f-w and for a while we wanted to try WD-40 (as the t/t manufacturer recommends for its platter) for the flywheel oil bath as well. After the testing and the friction comparisons between WD-40 and this milspec light oil I talked earlier, we decided we should shift to the latter for both, platter lubrication and flywheel lubrication.
Another essential change was the drives' number and configuration, aiming to minimize slippage and eventually kinetic energy (rotational) loss. The picture above is the last configuration that we have ended-up. More of that in the following section:
F-W mass - Drives
As I realised that skill was not something missing from my friends in 'TL Acoustics', when we decided to increase the rotating axis OD to make it stiffer, I thought we should try to increase the rotating mass a bit and so the 2nd and last version of the rotating bell - cut out of a fresh 'surgical grade solid stainless steel 316' - was weighting 11.1Kg, had an OD of 180mm and it was exercising a kinetic energy (rotational) figure of 18.5 units (believe it or not). Try now to add this to the kinetic energy (rotational) figure of the platter alone and you will end-up to an incredible figure of 26.6 Kg.m2.rpm for the 33.33RPM speed. (3.5 times more than the inertia figures exhibited by the 3 t/t's mentioned in the introduction!!)
Pictures of the 4 different flywheel configurations during the project are shown below. (2 rotating f-w bells paired with 4 drives configurations!)
At the same time, we cut 5 drives OUT, to bond the flywheel better with the platter, (flywheel and platter must work as one body, if possible), which, after a while, were reduced to the final 4 (and with deeper configuration). As for the drives IN, initially there was only one but in order for the whole system to respond quicker from the motor (eliminating some slipping there), I thought we should try a double drive connecting the motor to the f-w and so the motor spindle was also made exactly with the same pattern.
So, final: 2 threads f/ the motor & 4 threads to the platter. (Currently playing and testing various thread materials for the best operational and acoustical combination...)
I have mentioned above, about the difference in the drives' OD between IN & OUT and this is shown in the pictures. We call this gear ratio and is applied when we want to increase the torque of a rotating object. Gear ratio is used in all t/t's since we make a very small motor spindle (5 - 30mm OD) and we have a very large platter (300 - 310mm OD) but I haven't encountered applied in t/t flywheels (they all work with the same drive OD IN & OUT). So, if you enter a 'no-gear' f-w in a system, theoretically the rpm of the motor should not change for turning the platter with the same speed (added friction not taken into account). Since I had the 'permission' from my motor to turn really quick, I thought why shouldn't I try?
Finished analogue system set-up - Accessories
Placement - Disc-o-ring - Record clamp
In a paper I had entered a 3 years ago, I was writting why I preferred Symposium RollerBlocks to place the motor and the arm base on. Since the flywheel entered the rotating system and the tripod for the motor was used, RollerBlocks were abandoned. With the mass increase of the arm base, RollerBlocks used there have been removed too.
Now, both motor and arm base are sited on spikes and stainless steel pads. These were also manufactured by 'TL Acoustics' during the running of the flywheel project. For the platter's spikes, extra heavy pads were manufactured filled-up with absorbent/resonant-free material.
No plinth/base is used for the whole analogue set-up. Platter & f-w are sited on different shelves of the same DIY heavy rack (see details of the rack design here) and the motor is placed on a tripod. Distance between motor and flywheel is 450mm and between flywheel and platter 600mm. All 3 are placed on a straight line. Area needed for the whole set-up (phono stage incld) is 1500 x 500mm and total weight of the turntable (tripod incld but not motor p.su. & phono stage) is a 'delicate' 330lb/150Kg...
My DIY disc-o-ring remains the same - always placed on the record when I play - and details can be found here. Talking about momentum of inertia here, I wish to emphasize the contribution the 1.6Kg-315mm OD ring makes to it. It is calculated to 1.16 Kg.m2.rpm (33.33RPM), figure similar to what a 'normal' t/t flywheel contributes, (between the VPI's and the Forsell's f-w value, believe it or not!). You know why? Because all its mass is concentrated at the perimeter...
My DIY record clamp, used for some years, (original model here), has been redesigned and manufactured by 'TL Acoustics' during the running of the f-w project. External dimensions have not been altered but the material made of has changed. Now it's from stainless steel, the small shaft is made from an anti-resonant material (also used to make a contact ring which touches the record) and a stronger Neodymium ring magnet is used on the bottom. This magnet is paired with the magnet put on the top of the platter's spindle as mentioned earlier (see picture).
Energy absorption/drainage
All configurations below have been tested using a stethoscope and verified with music listenings:
Record mats
Record mads or pads play an important role in music reproduction and they can add color or be more neutral or sound more pleasent or whatever, especially if we are talking of small platters. Theory, mainly supported regarding record mats, says that the material of the mat should be close to the properties of vinyl. This is why many platters are made of acrylic material and no mat is used since acrylic is similar to vinyl. Metal platters though need some kind of mat to put the record on. For large and heavy platters though, IMHO, the record mat's role becomes less aparent.
During all these years I am 'playing' with the aas gabriel t/t, I have testeed many record mats, starting with the original, made out of synthetic leather. For 3 months I was using on the platter, a very good mat from carbon fibers hand made by Pluto, but ... the 'cheap' idea came to my mind. I moved to a design, described in this paper (scroll down), by sticking something like 60 self adhessive silicon 'tears', as I call them, on the platter. This design sounded to me better than the original mat and not inferior to Pluto's expensive design... Later on, I tested the ExtremePhono inexpensive non-felt mat (see review here) and this mat sounded better than my 'tears' solution and stayed on my platter for some time.
Pictures of 4 different record mats been tested so far are shown below.
When I was playing with some bitumen sheets for combating resonance, having made a lot of pads filled-up with this material, I thought I should try to make a bitumen record mat (this is close to vinyl properties I think...) So, I picked-up a 1mm thick bitumen sheet, I cut out a circle (using me disc-o-ring ID as a calliper!), I made a hole in the middle and that was it. These sheets are sold very cheap, used in the construction business, isolating roofs etc. and comes in rolls. As it is covered with a thin plastic membrane on both sides, I removed the one (face down to the platter) and kept the other one (face up); I wouldn't like my record to touch the bitumen substance! Results were far better than everything I had used before. Slam, neutral sound, analysis what have you...
Later on, I have removed the bitumen record mat and since the first flywheel arrived, I am playing my records directly on the platter with no mat at all! Music seems more absolute and controlled this way; problem is, it is difficult for me to put my fingers under the record when I want to remove it from the rotating platter. (I change records with the platter spinning.)
Recently my friend D. Whittaker sent me for testing an anti-slip record mat he is using on his own design the Aura t/t. My impressions are that it is very much worth the try because it sounded to me better than everything else I had tried before, except (again...) the 'no-mat' solution. It seems the 'no-mat' solution works in my case, due to the massive and special material platter, which absorbs the most of the resonance produced by the cartridge, not permitting any feedback. In my case, anything I put between the record and the platter acts as a reflection boundary, keeping resonance within the record mass, which feeds back the cartridge, muddling the sound. Certainly, direct placement of the record on the platter is also favored, (becoming more effective), from the disc-o-ring I am putting on, which flattens out the record, bonds the record with the platter and drains out any remaining vibration from it.
Driving threadsFor 7 years or so I have owned this turntable, I gained some experience regarding the material belts should be made from, depending on the distance you place the motor from the platter, the bond you want to make, grip, thread angle etc.
The picture shows the various threads I have used until today. From left to right (and chronologically): 3 types of silk threads, 2 types of fishing lines (0.2mm and 0.3mm) 2 types of tapes (6mm reel to reel tape & 4mm metal type cassette tape), wax-less dental floss and 3 types of rubber O' Rings (1mm, 1.5mm & 2mm). Nearly all of them performed adequately, if no flywheel was used, with the slim fishing line slightly ahead and later on with the 4mm cassette tape giving more dynamics.
When the flywheel got in the way, things became more demanding and complicated and some of those materials were proven insufficient. As I have mentioned before, theory is nice as far as maximizing momentum of inertia of the flywheel, but what we are really after is maximizing the kinetic energy (rotational) of the whole system, and this cannot be accomplished unless we will have a proper bond between the flywheel and the platter. Elasticity & slipping are the 'enemies'...
Two contenders were 'competing' at the beginning: dental floss & 0.2mm fishing line. It seemed the latter was gaining some points with simple knots done very easy in both of them and after that, equally stressed in-situ. Simple knots can be applied to silk threads as well but they cannot be stressed afterwards to be exactly equal and there is also the problem of slipping. So silk threads had been abandoned from the beginning. Tapes had also been excluded, due to slipping and difficulty to make a proper groove profile for them.
Later on, some friends gave me the idea of custom rubber belts, so I looked and discovered that I could use thick rubber O' Rings, offered in chords, (for large mean diameters I am interesting in), and with a very special glue they give with them, I could cut the desired length (90 degrees cut please!) and connect the 2 ends to make perfect and exactly equal rubber belts.
I have tested 3 thicknesses: 1mm, 1.5mm and 2mm profile OD and found that the ideal thickness was the thick one offering less elasticity, absolute grip and small vibration during rotation. I cannot completely eliminate the violin string vibration of the belt (due to the large distance between platter and flywheel but I prefer this way...), but the thick one behaves better in this respect and flaps around less. 8-10% stretching seems adequate not to produce rubber deformation and permanent elongation of the 2mm O' Ring.
In both places, these rubbers performed excellently (I think D. Whittaker also abandoned the idea of using 6mm tape between f-w and platter in his Aura design and turns back to 2 x 1.6mm OD rubber belts he was using before...) and for contemplating the elasticity of them, (less than that of a flat rubber belt but still more than any other material used before), I am thinking of increasing the number of rubber belts, driving the platter, from 4 to 5, making the bond of the platter to the flywheel almost anelastic... Some final machining must also be prformed to these 5 drives section of the flywheel to make their profiles semi-circular. I will eventually have 1 belt to drive (dental floss or rubber) & 5 belts to bond (rubber)...
Equivalent system mass - Comparisons
You have heard the statement "Mass is never bad for turntables" and I agree with this statement. Thing is, how much this mass could be for proper driving (motors, belts etc...), easy maintenance (lifting if required...), quiet operation (suspension, bearings, shafts etc...), stability and other nightmares that somebody would encounter applying this simple statement. It is true that in recent years, more and more tables are getting thicker and heavier. You can see pictures of them in our T/T Gallery, like Bogdan Audio, Pluto, Stratosphere, Symphonic Line, Ulysses, VYGER, Verdier, a couple of top Teres t/t models, some fancy DIY designs etc.
Imagine now a platter, heavy enough as the above turntables, paired with an extra heavy f-w and you've got another thing coming... And to put things above the same denominator for easy comparisons, using again my excel sheet, my 20Kg-33.33RPM platter paired with the 11.1Kg-375rpm f-w, gives a total kinetic energy (rotational) of 26.6 Kg.m2.rpm, which is equivalent to a massive 77.3Kg platter of 12" OD, if such a platter existed and be operational... But, as you can realize, a platter of this mass wouldn't be feasible to exist in one piece, since there is no way to suspend it, lift it, or rotate it without excessive wear in the bearings, noise etc.
Now, to be honest with you, the kinetic energy (rotational) figures of the platter and the flywheel are not exactly addable. Two rotating objects, connected via any kind of threads, do not perform like one body. There is always an energy loss, due to the elasticity and slipping of these threads. This is why I configured 4 drives, (and possibly 5 in the future...), to connect the f-w with the platter, to combine rigidity and flexibility of the connecting medium. But still, I do not believe that I get more than 80% of the flywheel's kinetic energy (rotational) to the platter, (with the thick rubber O' Rings that is).Talking about friction & energy losses, during the break-in period my p.su. custom volt-meter shows 5.16 Volts driving the motor for the 33.33 RPM speed. Taking the flywheel out of the rotating system, the same motor spindle needs 2.35 Volts of power to rotate the platter directly with the same speed... Theoretically, if they weren't any friction losses, the voltage value would've been 2.35 x gear ratio = 3.9 Volts. That means that we have a 32% additional friction loss, by adding the flywheel rotation, friction that causes the motor to work harder (higher voltage for the same work)...Never the less, a nice figure indicative of the rotating system's inertia is the time the system needs to stop (from the 33.33RPM rotation status to standstill). When the platter is connected via a fishing line with the motor and no f-w is used, it takes 45 seconds to stop. If the f-w is inserted in the system and rotate at 33.33RPM again, this time becomes 3 minutes!! (2 minutes with the rubber belts)
Obviously you would ask:
'How long this monster takes to gain 33.33RPM rotation from standstill?'
Well, it needs 2 min, with some manual spinning to help it a bit, when I switch on, but it really needs ½ to 1 hour to gain the exact RPM it was turning at the end of the previous day's listening session! (Oil bath to warm-up and become lighter, presumably - healthy reaction I think...) Well, I do not care a bit; many people wait even longer to warm up their precious valve amplifiers anyway... From 33 to 45RPM and vice versa, it needs only a couple of minutes and this is more essential to me: Lift off the record, put it in its sleeve, take out the next one and put it on the platter, doesn't need less time than that...
And there is another advantage (the best advantage of all IMHO to the sound!). This VERY SLOW system's reaction to speed changes means very slow reaction to speed flactuations or speed irregularities, if there are any (and not more than of the 0.05RPM range anyway - less than 0.1% - due to the excellent battery operated p.su.), leading to corrections of a frequency way out of the acoustical spectrum and so not affecting and confusing the sound reproduction what so ever! Now, I can explain the very clean notes I receive, down all the way to the frequency spectrum, after the flywheel entered the game...
Listening comments (info entered on 21 Jun 2005)
As you may have realised, it took me quite a while to finalize my listening impressions, because of the fact that the design went through many stages, with each one taking its time to break-in and "relax"... checking for the best threads combination to be used, getting the p.su. stronger etc... took much time before I calm down and put my final words regarding listening differences of the whole project.
I felt already from the beginning extremely happy from the fact that mechanically the rotation of the flywheel was completely silent and produced not even the slightest lateral movement when rotating!!! You have to bear in mind that the energy produced by this high rotation flywheel, is equivalent (f-w alone...) to a 55Kg platter! So, first thing I had noticed is more silence, something I wouldn't imagine it would show so clearly. Possibly this was also due to the fact that the 'noisy' motor got out of the way. And this noise absence is what I first commented on...
Now, as far as music itself, the first thing you notice is the increase in the dynamics. Secondly, someone notices that contrasts between music notes have become much-much more evident. Calm passages have become calmer and softer, strong passages have become so LIVE, they scare you to death. You listen to Mahler's Symphonies for example and you get the idea what live music is and what Composer of contrasts means...
Piano notes, drum hits, vibraphones etc. have re-gainned their real slam and energy, that you forget that this is a system that produces them. Violins on the other hand give you the impression that got back their correct tune and speed and depth and flow and there are instances, when you have the impression that particularly violins were a bit off-tuned before! Saxophones and trumpets give you the impression that they are louder and more precise in their dimension, and if the player moves you see them moving because music volume and energy differences are much better distinguishable...
What else can I say? Just the fact that when you hear music, you tend to critisize the recording and you have stoped talking (for ever) about driving abilities or frequency response or controlled bass or room tunes or anything of the sort... All these seem crap and nonsense!!! Because when you put the record on the platter and switch-off the lights, the system isn't there...
Music program highlights:
System:
Latest Developments (p.su. updates and driving belts final selection - info entered between 21 Jun & 29 Dec 2005)
Since I first entered this paper and almost completed it, there were some details pending, as far as the most suitable driving belts needed. I had also experienced a luck of strength of the original battery power supply, which had also to be considered. So, here we are with the latest (and final...) updates of the design:
For practical reasons and easy handling of the p.su., the battery was removed, because there was no need for it now. I was not using it because it couldn't last for long before it needed to be recharged again... What the use of it. I couldn't see any sound difference without using it anyway, in fact I was having the impression that with the power switched on (no pure battery drive that is), the sound was better. In addition I housed the external voltage monitor digital display inside the p.su box and I have replaced the trimmers, with high quality rotary buttons on the faceplate for easy handling the micro-speed calibration. That was phase I of the p.su. update.
phase II was more essential. The whole voltage regulation circuit has been removed and another one took its place, giving an output voltage rate of 1.5V to 15V and higher amperage than the original. So, the p.su., has become stronger almost 3 times, to put it in simple.
Let's move now to the VERY importand factor of rotation, which is driving belts, but before reading on, I would insist to review what is written about threads so far... OK?
So, starting from the last paragraph, as far as the FW-platter section, I had to dig another groove to the top section of the flywheel (to make the drives 5) and make them also semi-circular with 2mm groove ID to accept the round belts. So I did.
Rubber belts seemed OK, but still during rotation, they couldn't last for long, because they were glued for both ends to meet, even with the right configuration the grooves now had. On the other hand, surfing and searching the Internet for drive belts, I discovered that Polyurethane belts of several grades seemed more appropriate for the job. They are much less elastic than rubbers and as sticky as them (even more...) so they do not slip around the stainless steel of the flywheel's drives. So after some samples they sent me to check the apropriate length I was needing (not to be either loose or very stressed...), I finally ended up, with 5 x 2mm Polyurethane round belts of a hard grade to be used for the flywheel - platter section.
Now, here comes the difficult section: the motor-FW section. Anti-slipage and anti-elastic properties of the belt is a must here! The belt also MUST be DURABLE, because we are talking about much higher rotation here Dental floss is an easy to find solution, it is not slipping (not much...) it is anelastic, but... it doesn't last for long (see picture on the left). Fishing string is also anelastic, it lasts forever, but... it slips! In order to have some kind of hold, I had tried 5 rounds of them (that means one connection) but speed stability was not satisfactory and it was making noise as it was turning... So I abanoned the idea.
Again Polyurethane material came to my rescue...
Due to small drive diameters here, flat belts are appropriate. Again, samples were sent to me to find the correct hardness and length of the belt (a single belt is used here) and...
I ended up with 1mm width 0.7mm thickness (the slimmest) polyurethane flat belts (7.5% stressed recommended for the harder type offered), with the possibility in the future to open-up the motor capstan's & bottom FW's drive groove for a 2mm width flat belt, for best performance. No slippage long duration no noise!
These polyurethane belts that I am talking about are fused, with temperature to connect the 2 ends. So some may question the problem of exactly equal lenghts needed for the 5 round belts, conecting the flywheel with the platter. The tolerance they assured me is 1-2mm, which for the length needed is 0.1%, entirelly acceptable for the elasticity of the material. After 24 to 48 hours of stretching them in place (here, 3% stretch is used for each belt), even the smallest length difference will disappear!
Finally, and this is the last part of the "Play..." the bottom drive of the flywheel was opened to 2mm and a flat 2mm width polyurethane belt is implemented to rotate all those weights with no slippage. In fact, I have used the original aas Gabriel motor capstan but with the 2mm belt going around its body rather than inside a groove. It is definitely better and with much less rotational noise... I think it is a clever idea to use also 2 long threads (used to secure the capstan around the motor axis) instead of using 2 short ones, to work as a kind of flywheel effect...
Having said that, I am giving below most of the technicalities of the flywheel design and...
I am taking the chance to review the whole analogue set-up, covering the following subjects/topics:
- The current state of the analogue set-up
- The description of the t/t elements
- Platter - Chassis - Magnetic field
Power supply
aas gabriel t/t, w/ 20Kg platter & custom 11.1Kg-375rpm f-w, both magnetically suspended 
Analogue system as shown from the back wall and from the top. Motor, flywheel, platter and arm base placed on a straight line. Phono stage at the back of the arm - Motor - Arm base
- Motor housing
- The flywheel project
- The finished analogue system set-up - Accessories
- Listening comments
- Latest Developments (p.su. updates and driving belts' final selection)
Since, I will be talking below, about:
- moment of inertia (or rotational inertia) and
- kinetic energy (rot)...
I will like to remind from physics that:
- Moment of inertia is defined as the resistance to any change in rotation of a body.
- Just as mass resists in any change in linear motion, so does moment of inertia resist in any speeding up or slowing down of rotational motion.
- Kinetic energy is the product of the moment of inertia, multiplied by rotational speed.
Since, what we are aiming at, is not figures in absolute units but maximizing these figures, absolute figures may not be correct but relations among listed analogue designs as far as inertia and energy are true. (Some interesting stuff about theory & lab measurements can be found here)
Turntable's current state and description- aas gabriel heavy mass Swiss made & DIY improved.
- Platter: 44lb/20Kg high quality alloy, with a ring magnet inserted on the bottom and a custom aluminium spindle bearing a small ring magnet.
- Chassis: 84lb/38Kg high quality alloy, on a lead base, bearing a ring magnet on the top.
- Arm base: Outboard, standalone, from high quality alloy, on a lead base and add-on 'surgical grade solid stainless steel 316' cylinders, weighting 57lb/26Kg in total.
- Motor: High quality German made DC, with a polished metal motor spindle (no groove).
- Motor housing: Outboard, standalone, on a 33lb/15Kg lead base, placed on a dedicated 48lb/22Kg tripod.
- Power: Outboard, standalone p.su., with extra strength and 12V battery removed. High quality micro speed adjustment knobs for the 33 and the 45 rpm's.
- Flywheel: Outboard, standalone, from 'surgical grade solid stainless steel 316'; 25lb/11.1Kg - 180mm OD bell, rotating at 375rpm via 6 drives, floating on a 35lb/16Kg chassis, placed away from the platter.
- Drives:
- 1 x 45mm. OD drive, using 1 x 2mm Polyurethane flat belt f/ motor to f-w
- 5 x 27mm. OD drives, using 5 x 2mm OD Polyurethane round belts f/ f-w to platter. - Suspension: Platter and f-w floating by magneticrepulsion.
- Bearings - Magnets - Oil: Details below...
- Disc-'O'-Ring: 3.5lb/1.6Kg nickel coated brass, custom made.
- Record clamp: 1.3lb/0.6Kg 'surgical grade solid stainless steel 316', custom made, with a small ring magnet on bottom for extra push.
- Placement: No plinth/base provided. Motor - 450 (± 10mm) - flywheel - 600 (± 30mm) - platter on a straight line, with platter & f-w on the rack (different shelves) and motor on a tripod. Area needed for the whole set-up (phono stage incld): 1500 x 500mm.
- Total weight (tripod incld but not p.su. & phono stage): 330lb/150Kg.
Parts descriptionPlatter - Chassis - Magnetic fieldPlatter & chassis are described in paper02.htm and no 'deviations' from the original have been applied (obviously...). Only the platter's spindle has been modified by putting a couple of small neodymium ring magnets on its bottom to add a little bit more repulsion needed to contemplate for the extra weight of the disc-o-ring used on the platter. These 2 magnets act against a plate magnet put on the top of the sub-chassis axis. In addition, another small ring magnet is inserted on the top of the spindle, so oriented to attract the redesigned record clamp. More on that, in the record clamp section...
 |
| The 38Kg 'aas gabriel' chassis, with the huge bottom ring magnet (not as strong as neodymium, possibly ceramic...) |
 |
| A Neodymium magnetic rod, a steel ball and a moving coil cartridge resting on the platter |
There are only 3 t/t manufacturers in the world using magnetic repulsion for platter suspension: aas gabriel from Switzerland, Verdier from France and TL Acoustics from Hellas - Greece (+ possibly another one or two DIY'ers). The reason they are so few I believe is because some people still believe that the magnetic field used to levitate the platter is not safe for cartridges!!! Only from the fact that I am using a 10,000 Euro cartridge on such a 'dangerous' environment should prove enough but just for enjoying myself, I conducted a small experiment:
I placed on the bare platter a very strong Neodymium rod, a magnetic steel ball and my Insider 'reference' cartridge. Only with the first item, moving it along the surface of the platter you slightly feel the field. The ball can be placed and hold still wherever I put it without feeling anything at all and the cartridge has been playing music for years without any problem... For your info, moving coil cartridges have their small magnets (possibly 1000 times weaker than the rod in the picture) oriented perpendicular to the arm and this is the orientation the red magnetic rod preferred to rest on the platter during this small experiment!
OTOH, even with the almost complete absence of bearing friction, the platter (after the belts taken off) remains still at any position, floating perfectly without any rotary movement (preference to a special position) what so ever! And when it rotates, no even the slightest vertical movement or vibration along the axis is exhibited, reproducing the cleanest and the mostly rumble and wow and flatter free music one could dream of... So, no need for more discussion on the whole subject...
Power supply |
| The 'small' 7VA/12Volts battery psu for the aas gabriel DC motor |
 |
| the original motor spindle for silk or fishing line thread; Drive OD=8.5mm, rpm=1190 (no f-w) |
 |
| Modified flywheel-shaped motor spindle, for cassette tape 'belt'; Drive OD=5.3mm, rpm=1912 (no f-w) |
 |
| Latest version of the motor spindle (covered for protection with a nice transparent cap), with 2 drives (for extra grip) for dental floss or fishing line threads; Drive OD=10mm, rpm=1690 (w/ f-w) |
If time comes and motor has to be replaced for any reason, I will either upgrade it to the aas gabriel current p.su. model paired with a new motor, or find another solution. (i.e. Maxon strong motor paired with a Maxon battery operated p.su. - controller - digital encoder etc.)
Motor - Arm baseThe motor, as mentioned, remains the original. It proved very strong in comparison with the currently available Maxon motors, which they were thought to be a replacement for it, if excessive friction was encountered. I do not know what is going to happen in the future though...
 |
| Stand alone arm base, with 3 stainless steel cylinders screwed on bottom for increased weight and better stabilisation |
The third motor spindle configuration is the one currently used, with one drive at the beginning of the flywheel project and 2 drives when the flywheel took its final heavier shape, absolutely essential, otherwise slipping occurred!
The arm base, also described in paper#02 has been modified, regarding the placement of it on Roller Blocks in the past, (see paper#12). This paper is no valid now as far as this. With the inclusion of the flywheel, I thought the arm base should've been heavier and put on spikes for extra stabilization and rigidity. Three stainless steel cylinders have been screwed on the bottom of the base and 3 blind threads have been opened (one/cylinder) to accept spikes similar to the ones used for the huge turntable sub-chassis by the manufacturer. So now, the arm base consists of 3 materials: 3 stainless steel cylinders on the bottom, one Lead cylinder in the middle and 'Permalin' on the top. I am sure; this caused a large improvement to the reproduced sound...
Motor Housing  |
| The aas gabriel German made DC motor |
 |
| The motor on a stand alone heavily modified tripod (between the rack and the left speaker) has acoustically disappeared! |
From the beginning the motor was sited on a separate shelve, independent of the shelve the sub-chassis/platter was placed on and away from it. After the flywheel first designs, it has been realised that the rotation of the flywheel was best balanced and stabilised when put in the middle, between the platter and the motor and on a straight line with them. So, since this small separate shelve has been devoted for the flywheel, I had to add another rack for the motor and rack being impossible to be placed between the left speaker and the main rack, the best solution was the idea of a tripod there.
I was lucky enough to find a very good one, which I obviously modified by increasing its weight, mainly at the bottom, and filling it-in with lead shots for resonance absorption. Now the motor was acoustically completely 'invisible' by the platter... and using the rotating swivel of the tripod, I could also optimise very accurately the threads' tension between the motor and the flywheel! Impressive isn't it?
Flywheel projectIntroductionFlywheels are not extensively used among t/t manufacturers. Out of so many t/t models I have personally entered pictures of in the web site's T/T Gallery, only 3 manufacturers use a flywheel (in their top models that is...). The pictures shown are VPI's and Forsell's top models and the Australian Aura. VPI were the first to employ such an apparatus in their TNT models and in fact the first idea about inserting a flywheel in my aas gabriel (something that sounded to me at the beginning almost ridiculous for such a heavy mass platter) was from a friend working for VPI.
Later on, as this idea was drivelling my mind, I begun to think seriously about it and started to count pros and cons of such a design.
First I checked the motor if it could handle such a mass and friction addition, so I played with the trimmers of the p.su. and found that the working voltage range was wide enough to accept extra friction (2.22-7.13V), plenty to play with... (Motor originally working with 2.73V / 33.33RPM)
VPI TNT HR-X, with 2 motors & a 3.6Kg-230rpm f-w (owner: V. Gountanis, Athens - Hellas - Greece) 
Aura t/t with a 6.85Kg-106rpm f-w (designer: D. Whittaker, N. Zealand) 
Forsell Air Reference with a 1.2Kg-500rpm f-w (owner: S. Bonamis, Brussels-Belgium) 
'Aerex' stand alone flywheel (6.5Kg-150rpm rotating mass) - Then, I made a model in an Excel sheet for calculating kinetic energy (rotational) of platters spindles etc. to see what would be the effect of inserting a flywheel in the rotating chain.
Just to put things into prospective:
- The VPI shown has a kinetic energy (rotational) figure (in Kg.m2.rpm) of 7.29 without the f-w and 8.36 with the f-w (14.7% increase).
- The Aura has 5.8 without and 6.6 with its f-w (12.1% increase) and...
- The Forsell has 4.15 without and 6.34 with the f-w (52.9% increase).
- The Aerex stand alone flywheel contributes to any t/t is inserted in, with 0.93 Kg.m2.rpm kinetic energy (rotational).
- (These figures are calculated for the 33.33RPM speed and they are just simple estimations from the dimensions shown in the pictures but, IMHO, should not deviate more than 10% from reality...)
- Just for comparison:
- a Denon DP57L direct driven platter has 0.93 Kg.m2.rpm and...
- a belt driven 5Kg platter has 1.75 Kg.m2.rpm kinetic energy (rotational).
The aas gabriel alone (+ disc-o-ring but no flywheel) is producing a good 8.1, almost as high as the f-w equipped top VPI TNT. And to be honest, there are about 10 heavy platters in the t/t market, which are producing the same, more or less, momentum of inertia and none of them is employing a flywheel for extra energy storage. They do not need it or... you think they don't!!!
As I had designed my custom rack quite large to accommodate the platter and motor in separate shelves, I had the space to play with - meaning I could make a large OD flywheel and also being able to place it away from the platter, I could make the drive OD small enough, (and still have grasp), for high rotation. Large OD (or more explicitly more mass concentrated outside) and high rotation are the secrets of high kinetic energy (rotational)... and here come the cons of such designs - extreme designs that is:
Preliminary designsMy idea was to use such a rotary speed, which would not give, if something went wrong with the machining, resonance frequency that would affect the acoustical spectrum. So, I was playing with a frequency of 10Hz or 600rpm for the 45 speed. (10Hz is not heard but I know my speakers produce it). So I took this as a limit. To make things safer, I had initially thought of a nice round figure of 400rpm / 33.33RPM, which finally has been reduced to 375.- PROBLEM ONE: Friction from such rotation figure. Would it add noise to my almost friction free platter? Solution:
- Why not apply Hanspeter Gabriel's philosophy of magnetic suspension and use strong magnetic field for f-w repulsion? (Ball bearing was out of the question)
- Employ friction(noise)-free shaft bearings.
- PROBLEM TWO: Precise machining. Solution:
- The people who would do the job should've been acquainted with such designs and do not look at you like a strange animal! Where to find them in Greece? There was no Hellenic - Greek t/t manufacturer at the time I was searching - no one until...
Machining & Magnets: |
| Various inertia calculations put in an 'excel' sheet to find the best combination between weight, OD, ID and rotation of the flywheel (pdf file just for demonstration) |
Surfing on the INTERNET one day, (May 2003), I caught a fellow Hellenic - Greek DIYer, for the time, who was posting pictures from his t/t designs in a DIY Forum.
That was it! Him and a friend of his were just forming a Company named 'TL Acoustics' in Patras - south of Athens to start making a small production of high quality turntables (in kits and in complete set-ups). They were 2 dedicated people with great enthusiasm for analogue. Christos Livadas, an expert in handling the lathe; Konstantinos Tzivleris, a specialist in electronics, p.su.'s etc. That was what I was looking for! And the flywheel project came to life again...
We got in touch and they proposed to me to make the flywheel. They accepted many elements of my design and they put some of their own on the table. They took the magnetic suspension idea (in fact they are employing magnetic suspension in their t/t design ever since...) and they proposed shaft OD and lubrication method. They had also revealed to me one source of Neodymium magnets, a US manufacturer who proved extremely knowledgeable and helpful! Since this was in fact an experiment for me - and for them - we agreed on 'materials only' cost for the project.
 |  |  |  |
| Preliminary AutoCad 3-D pictures made by 'TL Acoustics' from my engineering drawings | |||
Finally we agreed on the dimensions and the design details (drives, magnets, bearings etc), they sent me some AutoCad 3-D pictures to have an idea, I placed the first magnets Order and we kicked off...
Flywheel constructionAs I said, this project was almost an experimental one. No apparatus of such a design has ever been tried before, we had no idea as to what extra friction we would encounter or any other problems might crop up. I couldn't take risks and this is why we thought that I should move the whole aas gabriel set-up (motor and p.su. included) from home to Patras, 250 Km south of Athens.
 |
| So, I took the big decision: Take everything apart and ... |
 |
| Flywheel base - shaft/drives - rotating 'bell' |
 |
| ... move it to the machine shop in Patras - Peloponnese! |
 |
| The 2 guys from 'TL Acoustics', posing in front of the first f-w model... (mid-July 2003) |
First thing changed from the original design were the drive belts. The first flywheel was designed to accept a cassette tape from the motor and an open reel tape from the flywheel to the platter. There is much talk lately about the advantages of the ¼" reel tape for driving platters and one of the pioneers of this idea is my friend D. Whittaker who uses it from his flywheel to the platter in his Aura t/t design. I was also using a cassette tape when I had modified the motor spindle to a flywheel-look one as I have mentioned before and I had been impressed by the sound improvement!
Unfortunately, with the extra rotating weight of the flywheel and such small drive OD's, both tapes were slipping. (You can't have it all...). Fishing line has also been tried but this was also slipping a bit. Perfect material at that time was found to be the 0.8mm. wax less dental floss. So drives of the flywheel had to be machined out and one drive in and 3 drives OUT were cut to accept one string of dental floss from the motor spindle and 3 strings to the platter.
Bearings - Magnets - OilSecond thing that proved inadequate was the shaft OD. We thought at the beginning that a small diameter of the rotating axis would benefit as far as friction and noise are concerned. But the high rotation of such large flywheel was causing a small trembling - bending of the axis which was almost eliminated but not disappeared if we put motor flywheel and platter on a straight line (something that we applied from there on any way), and there comes the need of a tripod to place the motor on, since the flywheel would take its place on the rack. So, next modification was the increase of the axis OD. Shaft and bearings had to be modified too...
 |
| The shaft with the bearings and the special mil-spec light oil inside (impossible to find...) |
 |
| Drives from the motor spindle and drives for the platter, with different OD's for extra torque! (testing here: 4 rubber 1mm OD O'Rings used for driving the platter and 2 fishing lines 0.2mm OD used for driving the f-w) |
Another essential change was the drives' number and configuration, aiming to minimize slippage and eventually kinetic energy (rotational) loss. The picture above is the last configuration that we have ended-up. More of that in the following section:
F-W mass - DrivesAs I realised that skill was not something missing from my friends in 'TL Acoustics', when we decided to increase the rotating axis OD to make it stiffer, I thought we should try to increase the rotating mass a bit and so the 2nd and last version of the rotating bell - cut out of a fresh 'surgical grade solid stainless steel 316' - was weighting 11.1Kg, had an OD of 180mm and it was exercising a kinetic energy (rotational) figure of 18.5 units (believe it or not). Try now to add this to the kinetic energy (rotational) figure of the platter alone and you will end-up to an incredible figure of 26.6 Kg.m2.rpm for the 33.33RPM speed. (3.5 times more than the inertia figures exhibited by the 3 t/t's mentioned in the introduction!!)
Pictures of the 4 different flywheel configurations during the project are shown below. (2 rotating f-w bells paired with 4 drives configurations!)
 |  |  |  |
| pic#1: Flywheel No#1 with 8Kg weight, rotating at 400 rpm and having 2 drives (1+1, for open reel tape and cassette tape threads) -//- pic#2: Flywheel No#2 with 8Kg weight, rotating at 400 rpm and having 4 drives (3+1, for dental floss threads) -//- pic#3: Flywheel No#3 with 11.1Kg weight, rotating at 400 rpm and having 6 drives (5+1, for dental floss threads) -//- pic#4: Flywheel No#4 with 11.1Kg weight, rotating at 375 rpm and having 6 drives (4+2, for dental floss or fishing line threads). Note: No modification to the first f-w base (sub-chassis) was made - only the rotating f-w body (bell) and the drives had been modified during the project. | |||
At the same time, we cut 5 drives OUT, to bond the flywheel better with the platter, (flywheel and platter must work as one body, if possible), which, after a while, were reduced to the final 4 (and with deeper configuration). As for the drives IN, initially there was only one but in order for the whole system to respond quicker from the motor (eliminating some slipping there), I thought we should try a double drive connecting the motor to the f-w and so the motor spindle was also made exactly with the same pattern.
 |
| Motor, Flywheel and Platter in-line, with the 6 strings transmitting the rotation... |
I have mentioned above, about the difference in the drives' OD between IN & OUT and this is shown in the pictures. We call this gear ratio and is applied when we want to increase the torque of a rotating object. Gear ratio is used in all t/t's since we make a very small motor spindle (5 - 30mm OD) and we have a very large platter (300 - 310mm OD) but I haven't encountered applied in t/t flywheels (they all work with the same drive OD IN & OUT). So, if you enter a 'no-gear' f-w in a system, theoretically the rpm of the motor should not change for turning the platter with the same speed (added friction not taken into account). Since I had the 'permission' from my motor to turn really quick, I thought why shouldn't I try?
- Starting from the fact that we have a standard platter OD (aas gabriel's is 12"), turning at a standard RPM and given the fact that I decided for a 375 f-w rotation, I ended-up with a 27mm. f-w drive OD to platter.
- Starting from the fact that I have a specific motor, turning within a certain range of rpm's (output voltage range) and given the fact that a good motor spindle OD (permitting various belt types to twist around it and have a good grip), is about 10mm., I ended-up with a 45mm. f-w drive OD to motor. (A good safety margin for encountered excess friction was taken into account)
- At a later stage, (after the break-in period), a more accurate motor spindle drive OD could be machined, for the motor to rotate within the recommended operating range for maximum torque. The break-in time is needed for the stabilisation of the total rotary friction of the system, so that the spindle's final drive OD would accurately contemplate it...
Finished analogue system set-up - AccessoriesPlacement - Disc-o-ring - Record clamp |
| Analogue system as shown from the left speaker |
 |
| Analogue system as shown from the right speaker (disc-o-ring on platter) |
Now, both motor and arm base are sited on spikes and stainless steel pads. These were also manufactured by 'TL Acoustics' during the running of the flywheel project. For the platter's spikes, extra heavy pads were manufactured filled-up with absorbent/resonant-free material.
No plinth/base is used for the whole analogue set-up. Platter & f-w are sited on different shelves of the same DIY heavy rack (see details of the rack design here) and the motor is placed on a tripod. Distance between motor and flywheel is 450mm and between flywheel and platter 600mm. All 3 are placed on a straight line. Area needed for the whole set-up (phono stage incld) is 1500 x 500mm and total weight of the turntable (tripod incld but not motor p.su. & phono stage) is a 'delicate' 330lb/150Kg...
 |
| Record clamp as shown from the bottom with the white 'acetal' material along the shaft and the perimeter and the small Neodymium ring magnet. Same magnet is shown on the top of the spindle |
My DIY record clamp, used for some years, (original model here), has been redesigned and manufactured by 'TL Acoustics' during the running of the f-w project. External dimensions have not been altered but the material made of has changed. Now it's from stainless steel, the small shaft is made from an anti-resonant material (also used to make a contact ring which touches the record) and a stronger Neodymium ring magnet is used on the bottom. This magnet is paired with the magnet put on the top of the platter's spindle as mentioned earlier (see picture).
Energy absorption/drainageAll configurations below have been tested using a stethoscope and verified with music listenings:
- The motor housing has been placed on steel cones and pads, directly on the tripod, for sound drainage downhill where it will propagate and eventually dissipated by the tripod's lead shots infill. Tensioning of the threads (motor to f-w) is done by the tripod's swivel.
- The flywheel base has been placed on steel cones and pads, directly on the shelve, for the same reason. The idea is, vibrations from the rotation to be transmitted to the shelve, through the lateral bearings and the cones and from there, to be dissipated by the heavy rack. A thin cloth has been placed under the pads for easy f-w alignment and proper tensioning of the threads (f-w to platter).
- The stand alone arm base has been placed on steel spikes and pads, directly on the shelve. The idea is also the same. Resonance and vibration from the arm rail and the cartridge (when tracking the groove) to find a path to drain out and not go or bounce back to the cartridge and distort the sound signal. The added weights to the arm base certainly help a lot on that, by absorbing some of this energy.
The main table has been placed on steel spikes and steel pads filled with compliant material. The idea now is a bit different. The special massive platter - thick single layer of Permalin material - is absorbing much of the vibration by the cartridge tracking the groove. It has been established that direct placement of the record onto the platter, (no use of record mat), really helps in this the most, especially with the use of my custom DIY disc-o-ring on the record! The remaining of the vibration energy drains out through the lateral bearings to the 38Kg lead sub chassis, with the absence of a ball bearing, which usually produces rumble (platter floating 2mm. above sub chassis). So, no further grounding - escape of resonance is needed to the shelve and by this isolation, vibrations escaping from the arm and going downhill to the shelve are not allowed to enter the sub chassis and to the platter.
This is the 12-leg Monster, with steel cones and pads for energy drainage (motor, f-w and arm base). Main turntable on compliant material filled in pads for energy isolation...
Record matsRecord mads or pads play an important role in music reproduction and they can add color or be more neutral or sound more pleasent or whatever, especially if we are talking of small platters. Theory, mainly supported regarding record mats, says that the material of the mat should be close to the properties of vinyl. This is why many platters are made of acrylic material and no mat is used since acrylic is similar to vinyl. Metal platters though need some kind of mat to put the record on. For large and heavy platters though, IMHO, the record mat's role becomes less aparent.
During all these years I am 'playing' with the aas gabriel t/t, I have testeed many record mats, starting with the original, made out of synthetic leather. For 3 months I was using on the platter, a very good mat from carbon fibers hand made by Pluto, but ... the 'cheap' idea came to my mind. I moved to a design, described in this paper (scroll down), by sticking something like 60 self adhessive silicon 'tears', as I call them, on the platter. This design sounded to me better than the original mat and not inferior to Pluto's expensive design... Later on, I tested the ExtremePhono inexpensive non-felt mat (see review here) and this mat sounded better than my 'tears' solution and stayed on my platter for some time.
Pictures of 4 different record mats been tested so far are shown below.
 |  |  |  |
| pic#1: The aas gabriel original mat from leather. No record clamp was used at that time -//- pic#2: Self adhesive silicon 'tears', put on the platter (out-performed mat#1) - the platter spindle and an early design of the record clamp up-side down are clearly seen -//- pic#3: 1 mm thin bitumen sheet (out-performed mat#2) -//- pic#4: Anti-slip thin and soft material, lately used for the Aura t/t, kindly offered by D. Whittaker, currently tested vs. the 'no mat' solution... Note: The dental floss knots, appearing in the pictures, do not pose any problem what so ever in the operation... | |||
When I was playing with some bitumen sheets for combating resonance, having made a lot of pads filled-up with this material, I thought I should try to make a bitumen record mat (this is close to vinyl properties I think...) So, I picked-up a 1mm thick bitumen sheet, I cut out a circle (using me disc-o-ring ID as a calliper!), I made a hole in the middle and that was it. These sheets are sold very cheap, used in the construction business, isolating roofs etc. and comes in rolls. As it is covered with a thin plastic membrane on both sides, I removed the one (face down to the platter) and kept the other one (face up); I wouldn't like my record to touch the bitumen substance! Results were far better than everything I had used before. Slam, neutral sound, analysis what have you...
Later on, I have removed the bitumen record mat and since the first flywheel arrived, I am playing my records directly on the platter with no mat at all! Music seems more absolute and controlled this way; problem is, it is difficult for me to put my fingers under the record when I want to remove it from the rotating platter. (I change records with the platter spinning.)
Recently my friend D. Whittaker sent me for testing an anti-slip record mat he is using on his own design the Aura t/t. My impressions are that it is very much worth the try because it sounded to me better than everything else I had tried before, except (again...) the 'no-mat' solution. It seems the 'no-mat' solution works in my case, due to the massive and special material platter, which absorbs the most of the resonance produced by the cartridge, not permitting any feedback. In my case, anything I put between the record and the platter acts as a reflection boundary, keeping resonance within the record mass, which feeds back the cartridge, muddling the sound. Certainly, direct placement of the record on the platter is also favored, (becoming more effective), from the disc-o-ring I am putting on, which flattens out the record, bonds the record with the platter and drains out any remaining vibration from it.
Driving threadsFor 7 years or so I have owned this turntable, I gained some experience regarding the material belts should be made from, depending on the distance you place the motor from the platter, the bond you want to make, grip, thread angle etc.  |
| Various material used for driving the platter (and lately the flywheel) |
When the flywheel got in the way, things became more demanding and complicated and some of those materials were proven insufficient. As I have mentioned before, theory is nice as far as maximizing momentum of inertia of the flywheel, but what we are really after is maximizing the kinetic energy (rotational) of the whole system, and this cannot be accomplished unless we will have a proper bond between the flywheel and the platter. Elasticity & slipping are the 'enemies'...
Two contenders were 'competing' at the beginning: dental floss & 0.2mm fishing line. It seemed the latter was gaining some points with simple knots done very easy in both of them and after that, equally stressed in-situ. Simple knots can be applied to silk threads as well but they cannot be stressed afterwards to be exactly equal and there is also the problem of slipping. So silk threads had been abandoned from the beginning. Tapes had also been excluded, due to slipping and difficulty to make a proper groove profile for them.
 |
| One 2mm thick rubber O' Ring driving the flywheel and 4 (soon 5!) for connecting the f-w with the platter... |
I have tested 3 thicknesses: 1mm, 1.5mm and 2mm profile OD and found that the ideal thickness was the thick one offering less elasticity, absolute grip and small vibration during rotation. I cannot completely eliminate the violin string vibration of the belt (due to the large distance between platter and flywheel but I prefer this way...), but the thick one behaves better in this respect and flaps around less. 8-10% stretching seems adequate not to produce rubber deformation and permanent elongation of the 2mm O' Ring.
In both places, these rubbers performed excellently (I think D. Whittaker also abandoned the idea of using 6mm tape between f-w and platter in his Aura design and turns back to 2 x 1.6mm OD rubber belts he was using before...) and for contemplating the elasticity of them, (less than that of a flat rubber belt but still more than any other material used before), I am thinking of increasing the number of rubber belts, driving the platter, from 4 to 5, making the bond of the platter to the flywheel almost anelastic... Some final machining must also be prformed to these 5 drives section of the flywheel to make their profiles semi-circular. I will eventually have 1 belt to drive (dental floss or rubber) & 5 belts to bond (rubber)...
Equivalent system mass - ComparisonsYou have heard the statement "Mass is never bad for turntables" and I agree with this statement. Thing is, how much this mass could be for proper driving (motors, belts etc...), easy maintenance (lifting if required...), quiet operation (suspension, bearings, shafts etc...), stability and other nightmares that somebody would encounter applying this simple statement. It is true that in recent years, more and more tables are getting thicker and heavier. You can see pictures of them in our T/T Gallery, like Bogdan Audio, Pluto, Stratosphere, Symphonic Line, Ulysses, VYGER, Verdier, a couple of top Teres t/t models, some fancy DIY designs etc.
Imagine now a platter, heavy enough as the above turntables, paired with an extra heavy f-w and you've got another thing coming... And to put things above the same denominator for easy comparisons, using again my excel sheet, my 20Kg-33.33RPM platter paired with the 11.1Kg-375rpm f-w, gives a total kinetic energy (rotational) of 26.6 Kg.m2.rpm, which is equivalent to a massive 77.3Kg platter of 12" OD, if such a platter existed and be operational... But, as you can realize, a platter of this mass wouldn't be feasible to exist in one piece, since there is no way to suspend it, lift it, or rotate it without excessive wear in the bearings, noise etc.
Now, to be honest with you, the kinetic energy (rotational) figures of the platter and the flywheel are not exactly addable. Two rotating objects, connected via any kind of threads, do not perform like one body. There is always an energy loss, due to the elasticity and slipping of these threads. This is why I configured 4 drives, (and possibly 5 in the future...), to connect the f-w with the platter, to combine rigidity and flexibility of the connecting medium. But still, I do not believe that I get more than 80% of the flywheel's kinetic energy (rotational) to the platter, (with the thick rubber O' Rings that is).Talking about friction & energy losses, during the break-in period my p.su. custom volt-meter shows 5.16 Volts driving the motor for the 33.33 RPM speed. Taking the flywheel out of the rotating system, the same motor spindle needs 2.35 Volts of power to rotate the platter directly with the same speed... Theoretically, if they weren't any friction losses, the voltage value would've been 2.35 x gear ratio = 3.9 Volts. That means that we have a 32% additional friction loss, by adding the flywheel rotation, friction that causes the motor to work harder (higher voltage for the same work)...Never the less, a nice figure indicative of the rotating system's inertia is the time the system needs to stop (from the 33.33RPM rotation status to standstill). When the platter is connected via a fishing line with the motor and no f-w is used, it takes 45 seconds to stop. If the f-w is inserted in the system and rotate at 33.33RPM again, this time becomes 3 minutes!! (2 minutes with the rubber belts)
Obviously you would ask:
'How long this monster takes to gain 33.33RPM rotation from standstill?'
Well, it needs 2 min, with some manual spinning to help it a bit, when I switch on, but it really needs ½ to 1 hour to gain the exact RPM it was turning at the end of the previous day's listening session! (Oil bath to warm-up and become lighter, presumably - healthy reaction I think...) Well, I do not care a bit; many people wait even longer to warm up their precious valve amplifiers anyway... From 33 to 45RPM and vice versa, it needs only a couple of minutes and this is more essential to me: Lift off the record, put it in its sleeve, take out the next one and put it on the platter, doesn't need less time than that...
And there is another advantage (the best advantage of all IMHO to the sound!). This VERY SLOW system's reaction to speed changes means very slow reaction to speed flactuations or speed irregularities, if there are any (and not more than of the 0.05RPM range anyway - less than 0.1% - due to the excellent battery operated p.su.), leading to corrections of a frequency way out of the acoustical spectrum and so not affecting and confusing the sound reproduction what so ever! Now, I can explain the very clean notes I receive, down all the way to the frequency spectrum, after the flywheel entered the game...
Listening comments (info entered on 21 Jun 2005)As you may have realised, it took me quite a while to finalize my listening impressions, because of the fact that the design went through many stages, with each one taking its time to break-in and "relax"... checking for the best threads combination to be used, getting the p.su. stronger etc... took much time before I calm down and put my final words regarding listening differences of the whole project.
I felt already from the beginning extremely happy from the fact that mechanically the rotation of the flywheel was completely silent and produced not even the slightest lateral movement when rotating!!! You have to bear in mind that the energy produced by this high rotation flywheel, is equivalent (f-w alone...) to a 55Kg platter! So, first thing I had noticed is more silence, something I wouldn't imagine it would show so clearly. Possibly this was also due to the fact that the 'noisy' motor got out of the way. And this noise absence is what I first commented on...
Now, as far as music itself, the first thing you notice is the increase in the dynamics. Secondly, someone notices that contrasts between music notes have become much-much more evident. Calm passages have become calmer and softer, strong passages have become so LIVE, they scare you to death. You listen to Mahler's Symphonies for example and you get the idea what live music is and what Composer of contrasts means...
Piano notes, drum hits, vibraphones etc. have re-gainned their real slam and energy, that you forget that this is a system that produces them. Violins on the other hand give you the impression that got back their correct tune and speed and depth and flow and there are instances, when you have the impression that particularly violins were a bit off-tuned before! Saxophones and trumpets give you the impression that they are louder and more precise in their dimension, and if the player moves you see them moving because music volume and energy differences are much better distinguishable...
What else can I say? Just the fact that when you hear music, you tend to critisize the recording and you have stoped talking (for ever) about driving abilities or frequency response or controlled bass or room tunes or anything of the sort... All these seem crap and nonsense!!! Because when you put the record on the platter and switch-off the lights, the system isn't there...
Music program highlights:
TEST RECORD 'Timbre' (Opus 3 8000)
C.L and K.T, from 'TL Acoustics'. This is how they looked when they first heard the results in my listening room... (mid-November 2003) - LAUDATE II 'Baroque Music f/the Duben Collection' (Proprius PROP 7860)
- CANTATE DOMINO (Proprius PROP 7762)
- JANACEK 'Fanfares' (CzStPh/Serebrier, Ref. Rec., RR-65)
- Bill Evans Trio 'Waltz for Debby' (Analogue Productions APJ 009)
- Rob Wasserman with ... 'DUETS' (Alto Analogue AA011)
- Michael Ruff 'Speaking in Melodies' (Sheffield Lab TLP-35)
- Clair Marlo 'Let it Go' (Sheffield Lab TLP-29)
- SOUNDS UNHEARD OF! (Analogue Productions 3009)
- Gregorio Paniagua 'LA FOLIA' (ATR 013)
- Bruce Katz Band 'Crescent Crawl' (AudioQuest AQ 1012)
- John Williams & the Boston Pops 'Swing, Swing, Swing' (Philips 412 626-1)
BIG BAND (Ref. Rec. RR-14)
Vinyl cutter Neumann VMS 70 
Vinyl cutter Technics SP 02 SP 10mk3' - Jacintha 'Danny Boy' (Groove Note Records GRV1001-1, the 45rpm attached record was used)
- TAFELMUSIK Popular Baroque Masterwoks (Canada's Orig. Instr. Baroque O., Ref. Rec. RR-13)
- THE TEMPEST 'Highlights f/ the Ballet' (San Francisco Performing Arts O./LeRoux, Ref. Rec. RR-10)
- WILD CHILD BUTLER (APO-004)
- HARRY JAMES 'Comin' from a good place' (Sheffield Lab 6)
- HARRY JAMES 'Still Harry after all these years' (Sheffield Lab 11)
- LES BROWN goes 'DIRECT TO DISC' (The Great American Gramophone Company)
- STRAVINSKY/DEBUSSY 'Firebird'/'Afternoon of a Faun' (Los Ang. P.O/Leinsdorf, Sheffield Lab LAB-24)
System:
- Genesis 300/301 - Rowland 8T amplifier - 'Consummate' line preamp. (with 450W ultra fast p.su.) - DACT CT100 phono module (custom silver wired with 550W dual ultra fast p.su.) - aas gabriel turntable (upgraded - see t.o.p.) - insider reference wood on Souther linear arm modified.
- Wiring: All interconnects & sp. cables are 'DARA custom silver'.
- Power cables: Xaitas 'black signature'.
- Electrical install.: 2X35A dedicated power lines + V4/V3 power isolators.
Latest Developments (p.su. updates and driving belts final selection - info entered between 21 Jun & 29 Dec 2005)Since I first entered this paper and almost completed it, there were some details pending, as far as the most suitable driving belts needed. I had also experienced a luck of strength of the original battery power supply, which had also to be considered. So, here we are with the latest (and final...) updates of the design:
 |
| p.su. with no battery & trimmers replaced |
 |
| p.su. 3 times stronger (and simpler...) than the original! |
phase II was more essential. The whole voltage regulation circuit has been removed and another one took its place, giving an output voltage rate of 1.5V to 15V and higher amperage than the original. So, the p.su., has become stronger almost 3 times, to put it in simple.
Let's move now to the VERY importand factor of rotation, which is driving belts, but before reading on, I would insist to review what is written about threads so far... OK?
 |
| How dental floss, used as driving string, f/ motor to fw gets deteriorated after a week's use |
 |
| 5 x 0.2mm OD round fishing strings f/motor to fw. Durable but slipping. Knots make noise too! |
Now, here comes the difficult section: the motor-FW section. Anti-slipage and anti-elastic properties of the belt is a must here! The belt also MUST be DURABLE, because we are talking about much higher rotation here Dental floss is an easy to find solution, it is not slipping (not much...) it is anelastic, but... it doesn't last for long (see picture on the left). Fishing string is also anelastic, it lasts forever, but... it slips! In order to have some kind of hold, I had tried 5 rounds of them (that means one connection) but speed stability was not satisfactory and it was making noise as it was turning... So I abanoned the idea.
 |
| 1mm flat polyurethane driving belt f/motor to fw |
 |
| 2mm flat polyurethane driving belt f/motor to fw. The original motor capstan is finally implemented but with the groove obviously not used |
 |
| 1mm flat polyurethane belt f/motor 5 x 2mmOD round belts to platter |
 |
| here is the last picture of the 'final product'... |
Due to small drive diameters here, flat belts are appropriate. Again, samples were sent to me to find the correct hardness and length of the belt (a single belt is used here) and...
I ended up with 1mm width 0.7mm thickness (the slimmest) polyurethane flat belts (7.5% stressed recommended for the harder type offered), with the possibility in the future to open-up the motor capstan's & bottom FW's drive groove for a 2mm width flat belt, for best performance. No slippage long duration no noise!
These polyurethane belts that I am talking about are fused, with temperature to connect the 2 ends. So some may question the problem of exactly equal lenghts needed for the 5 round belts, conecting the flywheel with the platter. The tolerance they assured me is 1-2mm, which for the length needed is 0.1%, entirelly acceptable for the elasticity of the material. After 24 to 48 hours of stretching them in place (here, 3% stretch is used for each belt), even the smallest length difference will disappear!
Finally, and this is the last part of the "Play..." the bottom drive of the flywheel was opened to 2mm and a flat 2mm width polyurethane belt is implemented to rotate all those weights with no slippage. In fact, I have used the original aas Gabriel motor capstan but with the 2mm belt going around its body rather than inside a groove. It is definitely better and with much less rotational noise... I think it is a clever idea to use also 2 long threads (used to secure the capstan around the motor axis) instead of using 2 short ones, to work as a kind of flywheel effect...