Theory first. An arm should remain at right angle to the groove no matter what direction it is traveling. Only in this way can the channels be read properly by the stylus. This must be done with maximum rigidity and minimum friction. Energy put into the arm tube must be damped or evacuated.
The knock on unipivots is that they can allow the arm tube to roll so that the cartride is no longer tangent to the groove. In most unipivots this is addressed by silicone or oil damping. If one were to take a stroke while swimming, then one would see that the part of the stroke taken in air has less resistance than the part in water. If one moves the arm slowly through the water there is little resistance, but the faster the arm moves in the water the more resistance. In theory, a damped unipivot will be excited to roll by a quick, powerful event, the onset of a pinch warp for instance, the oil will resist this event and keep the arm, mostly tangent. Sounds pretty iffy, no? Even some captured bearing arms use damping to address resonant and dynamic behaviour though. Since the unipivot uses a pivot point that is essentially a nail in a cup, the pivot point acts as a tip toe, with a single path for the evacuation of stored energy. There is also the full weight of the arm resting on this point and that give the bearing a lot of integrity. Additionally, since nothing in the bearing hinders horizontal and vertical movement, there is very little friction and very little is added as the arm ages. Sounds pretty bad for the poor unipivot, no?
Captured bearing arms always maintain the same tangency to the grove as their rolling motion is controlled by the vertical and horizontal pivots, usually cone and ball races. Provided that there is no play in the pivots, they will keep whatever tangency they are set up with. The arm is damped by the pivots resisting any motion that is not vertical or horizontal and by the verical bearings on the sides of the arm being equal paths for the evacuation of energy in addition to any damping inherent in the arm tube. This type of arm has some issues also. If you think about it, the ideal is to have infinitely minimal friction accompanied by infinitely tight bearings to prevent play in the bearing assemblies from letting the arm tube wander around and the cone banging around in the race. The two requirements are mutually exclusive in this type of bearing. The more that you tighten it to remove play, the more friction is introduced. These types of bearings are rated on the ABEC scale which basically is a measurement of how imperfect the bearing is. That said, modern machining techniques can make a very, very, good bearing. Two other things to look for in captured bearings are Brinelling, which is caused by external forces rolling the arm against the bearings thus flat spotting them. This can happen during mounting of the arm; excessive twisting causing the ball bearings to deform. Once deformed it is all over as the cone will no longer have a smooth surface to ride against. Captured bearing arms may also have wear issues in the bearings themselves but this is mostly related to materials technology and the initial integrity of the bearing. That said, as the market demonstrates, there are really excellent examples of both types of arms and execution is more important, in my mind, than type. There are also excellent designs using fluid bearings, air bearings, knife-edge bearings, etc. Don't get too hung up on this stuff. As always, listen and compare.
The knock on unipivots is that they can allow the arm tube to roll so that the cartride is no longer tangent to the groove. In most unipivots this is addressed by silicone or oil damping. If one were to take a stroke while swimming, then one would see that the part of the stroke taken in air has less resistance than the part in water. If one moves the arm slowly through the water there is little resistance, but the faster the arm moves in the water the more resistance. In theory, a damped unipivot will be excited to roll by a quick, powerful event, the onset of a pinch warp for instance, the oil will resist this event and keep the arm, mostly tangent. Sounds pretty iffy, no? Even some captured bearing arms use damping to address resonant and dynamic behaviour though. Since the unipivot uses a pivot point that is essentially a nail in a cup, the pivot point acts as a tip toe, with a single path for the evacuation of stored energy. There is also the full weight of the arm resting on this point and that give the bearing a lot of integrity. Additionally, since nothing in the bearing hinders horizontal and vertical movement, there is very little friction and very little is added as the arm ages. Sounds pretty bad for the poor unipivot, no?
Captured bearing arms always maintain the same tangency to the grove as their rolling motion is controlled by the vertical and horizontal pivots, usually cone and ball races. Provided that there is no play in the pivots, they will keep whatever tangency they are set up with. The arm is damped by the pivots resisting any motion that is not vertical or horizontal and by the verical bearings on the sides of the arm being equal paths for the evacuation of energy in addition to any damping inherent in the arm tube. This type of arm has some issues also. If you think about it, the ideal is to have infinitely minimal friction accompanied by infinitely tight bearings to prevent play in the bearing assemblies from letting the arm tube wander around and the cone banging around in the race. The two requirements are mutually exclusive in this type of bearing. The more that you tighten it to remove play, the more friction is introduced. These types of bearings are rated on the ABEC scale which basically is a measurement of how imperfect the bearing is. That said, modern machining techniques can make a very, very, good bearing. Two other things to look for in captured bearings are Brinelling, which is caused by external forces rolling the arm against the bearings thus flat spotting them. This can happen during mounting of the arm; excessive twisting causing the ball bearings to deform. Once deformed it is all over as the cone will no longer have a smooth surface to ride against. Captured bearing arms may also have wear issues in the bearings themselves but this is mostly related to materials technology and the initial integrity of the bearing. That said, as the market demonstrates, there are really excellent examples of both types of arms and execution is more important, in my mind, than type. There are also excellent designs using fluid bearings, air bearings, knife-edge bearings, etc. Don't get too hung up on this stuff. As always, listen and compare.