DIY Air Bearing Linear Arm

Some of you may know I built a ball bearing linear arm. However, this kind of linear arm excels with low compliance cartridge and requires high VTF to do the job right. My ball bearing arm does sound wonderful with Denon DL-103R(VTF=2.2, low compliance) after fine turning, especially for jazz. Bass is excellent. Denon DL-103R usually exhibits some shrillness in high frequency, but not on my linear arm.

In my opinion, ball bearing linear arm works better with high VTF and low compliance cartridge because two reasons.

1. Heavy moving mass. In most cases, there are no too many options to reduce moving mass.
2. Ball bearings have simply too high friction for high compliance cartridges.

So, the answer is to get ride of friction in order to use high compliance cartridges. Frictionless air bearing becomes the only option.

For air bearing, the most difficult part is air compressor, specially for New Way air brushing kind of arm because it requires high air pressure. The higher air pressure, the stiffer the arm is. I used a small fridge compressor to build quiet air compressor. The end result is excellent. It runs very quiet. In my listening room, the noise level is about 25 db in night. I measured the air compressor noise level. It doesn't even register on the sound pressure meter. The noise level is just same as a small refrigerator. I also add two cooling fans. These fans bring noise level up to about 30 db. It is no problem for me because I put the air compressor in the closet. I can hear nothing on my listening seat.

Please see attached my diy air compressor diagram. My compressor has two zones, high and low pressure zones.

1. Air Filter. It prevents small particles into the air line.
2. Compressor. It is small fridge compressor. In my case, it stops for 30 mins and runs about 20 mins. The compressor can be a little more powerful so the run time will be shorter. But in my case, it is ok.
3. 1st Stage In-line Filter. I put it first because it is cheap and can easily be replaced.
4. High Pressure Regulator. This is 2nd stage filter with high pressure regulator.
5. Check Valve. Check valve permits air only going one direction because the compressor can't hold high air pressure for long time. It also prevents high pressure air goes back into the compressor.
6. Pressure Switch. Once the tank reaches 80 psi, it will shut the compressor off and turn the compressor on once it comes down to 40 psi.
7. Safety Valve. It is set at 125 psi.
8. Air Tank. I bought 10 gallon, but 5 gallon should be fine. Air tank will hold high pressure air. Whenever I want to listen, it will be ready. I bought the tank from Homedepot for $34.00. They sell 5 gallon for $30.00.
9. Cooling Fans. These fans are from my old PC power supplies.
10. Temperature Control Switch. I set the temperature at 70 F.
11. 12 V DC power supply for the fans.
12. Main Power Switch.
13. On/Off Air Valve. Juts in case I need to shut off the air to feed the arm.
14. Flow Control Valve. It slows down air flow so I can have low pressure air to feed the arm.
15. Low Pressure Regulator. It regulates air pressure in low pressure zone. It controls the pressure to feed the arm. It has filter too. The filter element is 10 micron.
16. 4th Stage Filter. It has even finer filter. .3 micron.
17. 5th Stage Filter. It is auto fuel filter. Its function is to prevent anything small particles into the air line.

I like my diy air compressor so much. It is quiet and runs cool. I don't think you can buy anything commercially available air compressor better than my diy air compressor even you may spend a lot of money. The total cost of my diy air compressor is about $350.
 

Attachments

  • air-compressor-dia.jpg
    air-compressor-dia.jpg
    184.7 KB · Views: 3,225
  • air-bearing6.jpg
    air-bearing6.jpg
    397 KB · Views: 3,200
  • air-bearing7.jpg
    air-bearing7.jpg
    358.5 KB · Views: 3,073
  • air-bearing8.jpg
    air-bearing8.jpg
    336.7 KB · Views: 2,924
Last edited:
Here is the diy arm.

In Photo 1

1. Two dial indicators to measure VTA.
2. The arm slides on CNC machine rails.
3. For arm adjustments.
4. 1/16" latex tubing. It is very flexible and has very little resistance.
5. Micrometer head to do fine back and forward adjustments.

In Photo 2

1. Magnets. It works perfectly. Slight push. The arm base will be off the micrometer head. But once it close to micrometer head, it will pull the arm base in about 1 mm distance. Once the magnet touches the micrometer head, it will hold the arm in place.
2. Micrometer head.

In Photo 3

1. + and - wires wrap together to reduce interference.
2. Carbon fiber to enforce the head shell and to slow down the transmission of resonance.
3. Arm rest.
4. Stainless steel bearing shaft. I polished it with 2000 grit sand paper to mirror image.
5. Teflon insulated silver wires.

In Photo 4

1. Pressure gauge by the arm to indicate air flow.
2. Shut off valve.

In Photo 5

1. The air bearing arm is suspend with table plinth.
2. Micrometer heads to do fine adjustments.

The arm sounds wonderful without any fine tuning so far. It can work with any cartridges, high compliance and low compliance with VTF 1.0 g. The arm is very quiet without any artificial click, pops due to bearing frictionless and correct tracking. The sound is very 3d like, natural. I feed the arm with 35-40 psi air pressure. The compressor runs about 20 mins then stops for about 30 mins.

Any improvements? Yes. I realize that the head shell can be shortened a little more than half inch, see photo 6.
 

Attachments

  • air-bearing9.jpg
    air-bearing9.jpg
    341.5 KB · Views: 1,601
  • air-bearing5.jpg
    air-bearing5.jpg
    388.4 KB · Views: 1,312
  • air-bearing4.jpg
    air-bearing4.jpg
    386.7 KB · Views: 1,322
  • air-bearing3.jpg
    air-bearing3.jpg
    338.3 KB · Views: 1,360
  • air-bearing2.jpg
    air-bearing2.jpg
    328.5 KB · Views: 1,565
  • air-bearing1.jpg
    air-bearing1.jpg
    351 KB · Views: 2,871
This all seems like a huge degree of complication for a simple job imho.
That said kudos for your engineering efforts.
I have a couple of junked wide format dot matrix printers.
To me the rails/stepper motors assy are just begging to be donor parts for a optical deviation sensor controlled linear TT running a long tone arm, shorter arm if that is a better solution.
There have been mass production examples of this method using lower mass/heft mechanics.
Is air bearing the correct solution....any mass in the system allows/encourages tracking angle error/overshoot according to disk eccentricity.
In my thoughts are using inner groove (or outer groove played backwards) error signal to drive a disk centering auto/manual disk positioning system.

Is there real advantage in running a very short tone arm ?.

Dan.
 
super10018,

Your arm looks VERY nice !!! You have gained lots of experience from your ball bearing arm, and have put it to good use. Years ago I had an Eminent technology 2 air bearing and it worked very well with various cartridges from Shure v15 3 to Dynavector 17d however the low bass seemed a little light in weight when compared to an older Dynavector 505, although both arms sounded excellent. I wonder if you have done a comparison of both the arms you made with the same cartridge. Very nice work !!!
 
Dan,

Actually, it is very simple. The arm does a very basic action, linear motion. All the efforts are to achieve one thing, i.e. short arm. Why do I go all the way to make short arm? Because short arm has good bass, high dynamics, esp. marco dynamics and natural sounding. I am inspired by Versa Dynamics arm. My own experience confirms it, too.

My arm's moving mass is not low at all. It is 67 grams. But I think it is lighter than Kuzma airline moving mass. It is said that Kuzma's is 80 grams. I use 1/2" air bushing, which is 33 grams by itself, just because I was not sure about high moving mass. Kuzma uses 20 mm air bushing. It weights 54 grams. However, I understand now. 3/4" or 20" should have no problem at all. Its high mass may even enhance bass and dynamics. I plan to make another heavier but shorter magnesium head shell. In the meantimes, I may change the counter weight mechanism to add some mass, too.

I have read the thread about disk eccentricity with great interest. But eccentricity is just part of LP playback. What can we do about? I don't think we can do anything. Cut a big hole in every disks? No. You can't do this. It is not my concern at all. If you try to find a solution, we are really complicating a simple problem. In my opinion, trying to force a pivot arm does linear action is making a simple problem complicated too. To force a pivot arm does linear motion is illogic mechanically although some of designs are very clever and inspirational.

Thank you for your input!
Jim
 
httattoo,

I have compared two cartridges on different arms so far. Both cartridges are high compliance and low VTF. One of cartridges is zyx. On the ball bearing arm, zyx may produce some kind of distortion, and buzzing for some music passages. I know its cantilever is fighting with the friction of ball bearing when that happens. On the air bearing arm, it works wonderfully. It is my favorite cartridge on the air bearing arm so far. I also tried another Lyra MC cartridge, which was recently re-tipped by Soundsmith. On the ball bearing arm, its bass is completely killed at 1.7 g. But I can't use VTF lower than 1.7 because it may mis-track. On the air bearing arm, the Lyra runs smoothly even at VTF 1.5 g. The Lyra produces good bass with lower VTF. I even tried the Lyra at 1.2 g. It tracks very good. Generally speaking, air bearing arm is certainly better than ball bearing arm. Ball bearing arm has its limitations. It works well with low complain cartridges. Ball bearing arm can be very very good if it has the right cartridge to go with. Air bearing may well be the best mechanism you possibly can get for a tangential tone arm. Of course, air bearing arm requires air compressor so it is not as easy as using ball bearing arm.
 
Last edited:
I have read Stereophile Michael Fremer's review of Kuzma Air Line tonearm recently. In his review, he raised a lot of interesting questions about air bearing linear-tracking arm. Here is the link.

Kuzma Air Line tonearm | Stereophile.com

He wrote:" The one glaring omission in the Air Line's design is a damping trough. A linear design—even one using a stationary rail—must laterally move a large mass. The grooves of most LPs are not concentric because of pressing inconsistencies, which means the groove is constantly shifting the tonearm's mass back and forth. Because the stylus is at the end of a spring mechanism (the cantilever's suspension), if you use a cartridge with too high a compliance—too floppy a spring—you can have the tail wagging the dog, in that the motion of the groove spiral will move the cantilever from its central position. The risk is then that with a frictionless bearing the stylus might be slammed from one groove wall to the other."

The following is manufacture's comment. "A system will resonate only when disturbing forces appear at the resonance frequency. If there are no disturbing forces, then there are no problems. However, if a system is overdamped, then instead of one resonance, two smaller resonances occur, one below and one above the previous resonance, which can create further problems. There is, in fact, a level of effective damping on the Air Line tonearm. The cantilever suspension itself and the air supply tube add damping. Our choice was for either too little or adequate damping; we chose the latter."

Basically, Kuzma is saying that external damping is provided by the resistance of air supply tubing. On my air bearing arm, I can turn the cooper tube therefore to increase the resistance of air supply tubing. Please see photo. However, I don't know how effect it is and how much damping can be generated. I have seen someone uses oil kind of liquid to damp the moving force of arm mass. But it is not easy to implement. Here is my proposal to add a damping device. Please see the sketch. It is simple and easy to implement. Weights can be adjusted accordingly base upon how much damping you need. At the center of string, there are two small springs.

All suggestions and critics are welcome.
 

Attachments

  • air-bearing10.jpg
    air-bearing10.jpg
    173 KB · Views: 1,365
  • air-bearing11.jpg
    air-bearing11.jpg
    165.2 KB · Views: 1,014
Last edited:
He wrote:" The one glaring omission in the Air Line's design is a damping trough. A linear design—even one using a stationary rail—must laterally move a large mass. The grooves of most LPs are not concentric because of pressing inconsistencies, which means the groove is constantly shifting the tonearm's mass back and forth. Because the stylus is at the end of a spring mechanism (the cantilever's suspension), if you use a cartridge with too high a compliance—too floppy a spring—you can have the tail wagging the dog, in that the motion of the groove spiral will move the cantilever from its central position. The risk is then that with a frictionless bearing the stylus might be slammed from one groove wall to the other.
Exactly as above are my concerns also with this frictionless air 'lubricated' linear bearing concept with relatively heavy cartridge/arm/sled system.

In the condition of off centre record, this high mass is encouraging sled lateral overshoot/resonance and stylus cantilever deviation from it's central position within the cartridge magnetic circuit, all due to cantilever lateral compliance.
Distortion and mistracking are the results of this behaviour.
Higher tracking force is a remedy, but not the proper solution.

There are numerous production examples of motorised arm mounting sled driven by tracking error signal generated by angular offset sensors (usually optical) of more conventional twin pivot arm.
Lateral run out of the disc will cause a tracking error signal AC component, which is heavily filtered and the resultant averaged/smoothed DC component drives a sled motor servo system....typical CD players also use this method.

My solution would to to use the tracking error signal to drive a centre reading MC meter (or other suitable display)
With the stylus placed in the final groove of the record, and the record rotated manually, it would be a simple matter to precisely centre the record whilst viewing the display.
The metering system would require DC offset and gain controls.
The platter would require an undersized centre spindle.
Once the record grooves are exactly centred, horizontal stylus deviation from neutral position is of much less concern, if not eliminated.

My other concern of these very short arm systems is VTA modulation caused by record warp.
To cure this problem I would look to a vacuum clamping system....not too difficult to implement in my opinion.

One day when I get a round tuit, it I will give it a go.

Dan.
 
Super,


That is a very nice arm. I must say it is more heavily engineered than even the lathes that a laquer is cut on :). From what I have been reading thus far is that some lateral friction/damping is needed to be added to the air bearing?. This was my observation even with using bearing that there is a happy medium between over and under damped lateral motion. I also learned the same applied to a lesser degree to vertical motion. With the 2 bearing 2 tube I did this with displacement to center of gravity. As Fremer states correctly, in a nutshell , the larger the mass the harder to stop the inertia. In a bearing you can also carefully tailor the two to respond ideally.


Colin
 
Since we want arm to be moved easily on air bearing but also want to damp it Wouldn't it be little difficult to achieve ? A little offbeat newbie approach. Can the wire coming out of tonearm be used to dampen the movement with varying stiffness of the wire ?
Regards
 
I have done some experiments in regarding damping and disk eccentricity. I played a eccentric LP with the air bearing arm. The air bearing kept moving side way from left to right and from right to left. The distance of moving is between 2-3 mm. I used a Lyra high compliance mc cartridge at recommended VTF 1.7 g. The sound was normal and had no noise. But I couldn’t tell if there were distortion or not. Then, I changed VTF from 1.7 to 1.2 g, which is much lower than manufacture recommended VTF. I could hear some kind noise at a particulate frequency at same area when the needle was being thrown on right side of groove wall. This experiment indicates that as long as you keep the correct VTF, the noise caused by lack of damping isn’t audible usually. In most of cases, it will be fine even without damping. I couldn’t tell if there are mis-tracking and distortion by just listening. In theory, I believe mis-tracking and distortion exist. In other words, it is ok not to damp as long as VTF is correct in most of cases.

So, in my opinion, Michael Fremer’s comment is over simplified. He didn’t consider VTF. Kuzma is right. If VTF is correct, it adds damping. Even for pivot arm, if VTF is not correct, needle will be thrown on the right side of groove wall. But in the meantime, Kuzma’s explanation about damping is not adequate for such high priced arm. Air supply tubing may add sampling but it is not good enough.

In my case, my arm is shorter than Kuzma’s and mass is lower than Kuzma’s. The disturbing force to throw needle off center of groove may be amplified through long arm wand. Once the disturbing force happens, it is more difficult to get the needle back to original position due to heavier mass. This is why I was reluctant to use 3/4” air bushing at beginning and went all the way to make a short arm wand.

In reality, it is probably ok without special damping device. There are a lot of factors contributing to LP playback distortion besides damping. But I am a perfectionist. I want to do it right.
 
Last edited:
I see you don,t have a vacuum platter? And with the relatively short pivot distance, you might have some sideways movement due to warps?
I,m in the process of doing a ET2 with a short 1.4" pivot arc distance with silicone damping.
For what it's worth, I will use a big 15" sub that I can see cone movement to adjust the viscosity, so on the worst off center discs I have , a somewhat stable woofer movement will be the goal. This is all low frequency stuff.
Don,t know if your spring system will work
You feel good that the airline to the arm isn,t causing any issue at the start and end of a record?

Regards
David
 
I see you don,t have a vacuum platter? And with the relatively short pivot distance, you might have some sideways movement due to warps?
I,m in the process of doing a ET2 with a short 1.4" pivot arc distance with silicone damping.
For what it's worth, I will use a big 15" sub that I can see cone movement to adjust the viscosity, so on the worst off center discs I have , a somewhat stable woofer movement will be the goal. This is all low frequency stuff.
Don,t know if your spring system will work
You feel good that the airline to the arm isn,t causing any issue at the start and end of a record?

Regards
David

David,

I have an outer clamp, which is about 6 lbs. Please see photo 6. Therefore, warp is not a problem for me. Otherwise, I won't try to make short arm. I don't see any problem at the start and the end of a record. In fact, the quality of sound for the air bearing arm is very consistent. I can't hear any difference. I have been using Graham 2.2 for almost 9 years. Over 90% of the times, I could hear the difference when the cartridge was approaching the ends of the records. The only thing I need to be careful is to put the needle down at the start of the records. The air bearing is frictionless. A very light push will move the arm way into the middle of the records.