A variety of drive systems have been devised for precision telescope pointing and tracking. Such drive systems include worm drives, chain drives, friction drives and hybrid band-worm drives. Each type of drive system possesses several advantages and disadvantages. In some cases, disadvantages can be minimized through proper design and implementation of the drive system. This article's purpose is simply to educate the reader about the potential advantages and disadvantages for each type of drive system. To this end, this article does not examine any specific implementation of a particular type of drive system either produced or used by any telescope manufacturer.

Please note: Both potential and inherent advantages, and potential and inherent disadvantages are presented for each type of drive system. This does not imply that either the potential or inherent advantages or disadvantages actually exist or are present to a substantial degree within a particular manufacturer's implementation of a particular type of drive system. Again, this article is intended to educate the reader of all possible advantages and disadvantages with each type of drive system.

The structural design of the telescope's mount, optical tube assembly (OTA) and optical supports are subject to flexure when the telescope is pointed in different orientations. These inherent flexures impose limits on the telescope's pointing accuracy, and impose limits on the telescope's tracking accuracy during extremely long duration exposures. Additional factors which affect pointing and tracking accuracy include RA-DEC axis non-perpendicularity, OTA-DEC axis non-perpendicularity (optic tilt), and polar alignment errors in altitude and azimuth relative to the true pole. Fortunately these problems can be compensated for by using pointing correction software such as T-point. Pointing correction software performs a remarkable job of correcting for these pointing errors as long as the telescope structure does not exhibit any significant amounts of hysteresis (random and unpredictable errors). Thus the telescope's drive system becomes the only remaining component which can also affect the pointing and tracking performance of the telescope.

There are several types of drive systems which have been or are currently used for observatory class telescopes. Each type of drive system has unique features and possible or inherent benefits, as well as possible or inherent limitations. Obviously there are several factors which must be taken into account when deciding which type of drive system is ideally suited both for the telescope and the user's intended applications. Some of these factors are the drive system's performance, the local environmental conditions encountered during routine operation of the telescope, the drive system's cost, the drive system's expected lifetime, and the drive system's immunity to damage from contamination or from a lack of maintenance.

With this in mind, let's take a look at each type of drive system which is available for or is currently used on observatory class (large aperture) telescopes:

Worm drives employ a worm which is mated to a matching gear, and may include a built-in slip clutch mechanism. The slip clutch facilitates manual positioning of the telescope and prevents damage to the drive system should a power outage occur or a physical obstruction be encountered that prevents movement of the telescope. Worm drives usually are attached directly to the telescope's axes of rotation.

Worm drives can have certain advantages:

• Moderately high stiffness.
• Excellent tracking if precisely manufactured and if periodic error correction hardware or software is used.
• Very low cost for small to medium gear sizes.
• Relatively easy to install.

Moderately High Stiffness: The stiffness of a worm drive is limited by the flexure of the thread on the worm and a single tooth on the gear, and by any flexures, stress or strain within the worm block assembly and its associated pivot point(s) which engage the worm to the gear.

Worm drives can have several potential disadvantages:

• High loading on the gear's teeth, since only one or two teeth are in contact with the worm at any given moment.
• Periodic error with each revolution of the worm.
• Vibrations during very high speed slews due to the inherent periodic error.
• Tooth-to-tooth errors, due to the limiting accuracy obtainable during manufacture.
• Drive backlash, caused by too much clearance between the worm and gear or by excessive play in the worm's pivot point.
• Gear decentering errors, usually caused by a poorly designed clutch mechanism.
• Throat height error, caused by adjustment of a poorly designed clutch mechanism.
• Tangent angle error, caused by adjustment of a poorly designed clutch mechanism.
• Susceptibility to damage by foreign contaminants or due to the lack of a clutch mechanism.
• Frequent maintenance since the grease lubricating the worm and gear must be periodically cleaned off and replaced.

High Loading: Conventional worm drives inherently allow only one or two gear teeth to be in full contact with the worm at any given moment during each revolution of the worm. Contributing factors which produce high loading upon the gear's teeth are friction within the telescope's bearings, an improperly balanced telescope, and the telescope's inertia during the acceleration or deceleration phases of a slew. This high loading may be greatly reduced by using lightweight materials in the construction of the telescope in order to reduce the telescope's inertia, by making sure the telescope is properly balanced, by making sure that the drive train and bearings are properly greased to keep friction to a minimum, or by using extremely large worm gears. If the telescope is not properly balanced for all orientations, then the gear may wear unevenly, resulting in a somewhat elliptical or other shape being introduced into the gear over time. This will cause increasing pointing and tracking errors over time.

Periodic Error: Worm gears always possess an inherent periodic error. While the telescope is tracking, the periodic error may be easily compensated for by using periodic error correction (PEC) software. The periodic error of a conventional worm drive may change over time due to wear. Thus the periodic error correction circuitry may have to be reprogrammed on a fairly regular basis. Note that if the worm is correctly aligned to the gear, if a method is provided for maintaining precise and full worm to gear engagement, and if the engagement mechanism's pivot point is located on a line tangent to the worm/gear point of contact, then the periodic error could actually decrease over time as the worm becomes even more precisely lapped to the gear.

There are several available design means to reduce periodic error. The simplest method is to use a very large number of teeth on the gear in order to reduce the pitch angle of the teeth. The second method is either to reduce the thickness of the gear or to increase the diameter of the worm. Reducing the thickness of the gear results in teeth which have a smaller contact surface area, thus increasing the loading on the teeth. Increasing the diameter of the worm reduces the pitch angle of the teeth, but increases the possibility of the worm chattering against the gear even when the highest quality grease is used for lubrication. The final method is to make the gear very large in diameter such that the gear's inherent periodic error causes less angular deflection of the telescope. Any of these methods may be combined to reduce periodic error to an absolute minimum.

Vibrations Due to Periodic Error: While periodic error may be easily corrected using PEC software, large amounts of inherent periodic error can cause vibrations when the telescope is rapidly slewed. During slewing, the worm drive's periodic error is essentially causing the telescope to accelerate and decelerate slightly with every revolution of the worm. This causes vibration and results in in increased loading between the worm and each tooth of the gear as the telescope is rapidly slewed. This increased loading during high speed slews can rapidly increase the rate of wear for the worm drive.

Tooth-to-Tooth Errors: Quality worm gears are manufactured on expensive gear hobbing machines. Nevertheless, all gear hobbing machines have a limiting tolerance to which they can cut the teeth on a gear. This limiting tolerance results in slight spacing errors between each tooth on the gear. These spacing errors, or tooth-to-tooth errors, may be substantially reduced or nearly eliminated by performing precision lapping operations of the worm to the gear. Generally the larger the worm gear then the smaller the tooth-to-tooth errors will be.

Backlash: Drive backlash will develop over time if the worm's support mechanism does not feature a method for maintaining full engagement between the worm and the gear, however this particular form of backlash is rarely seen since most manufactures use mechanisms to maintain full engagement of the worm and the gear. Another source of backlash is any play within the pivot point(s) of the worm/gear engagement mechanism. Again, design means usually are used to prevent any play within the engagement mechanism.

Decentering: Decentering of the gear may be caused by either the mounting method used to attach the gear to the telescope axis, by the gear's inherent clutch mechanism, or by errors made when manufacturing the gear blank. Slip fit mounting mechanisms are commonly employed to attach a gear to a telescope's axis of revolution. Such slip fit mounting mechanisms will obviously not precisely preserve the coaxial alignment of the gear's axis relative to the telescope's drive axis. Nevertheless, this axial misalignment and resulting pointing errors are easily compensated for by pointing correction software. A gear's built-in clutch mechanism can also cause decentering errors. The gear must be free to revolve within the clutch mechanism while being constrained only by the friction of the clutch mechanism. This normally requires a small clearance between the gear's inner bore and the clutch mechanism's hub. This clearance can allow the gear's axis to shift laterally relative to the telescope's axis of revolution. Generally the larger the gear then the less pronounced will be any decentering errors due to a built-in clutch. Careful design and manufacture of the clutch mechanism can minimize or prevent any substantial decentering errors, assuming the gear's bore is concentric to the gear's outside edge. Decentering errors caused by usage of the clutch are not predictable and thus are not compensated for by pointing correction software.

Throat Height and Tangent Angle Error: If the design of the clutch mechanism is not identical for each side of the gear, then the pressure exerted by the clutch can slightly warp the face of the gear. This warpage will alter the throat height alignment between the worm and the gear. If the warpage is severe, then the warpage will also alter the tangent angle alignment between the worm and the gear. Adjustment of the gear's clutch mechanism may cause the gear's clutch pad to compress or decompress slightly, additionally altering the throat height alignment between the worm and the gear. Since the gear's teeth inherently have a pitch angle, any change in throat height alignment will introduce pointing errors in the form of pointing offsets. Any alterations by the clutch to the throat height and/or tangent angle alignments will increase the gear's periodic error and result in uneven and improper wear patterns on both the gear and the worm.

Susceptibility to Damage: Since a worm drive features a worm which rotates against a gear, the surfaces of the worm and teeth can become scored if any grit or other contaminants get between the worm and the gear. Visually imagine running the flat of a knife through a tub of butter. The knife will dig out some butter while leaving a depression which is bordered on either side by a slight raised edge. The same process occurs when either the worm or a tooth becomes scored by grit. While the slight raised edges will rapidly be worn away, the trough still remains because material has been gouged out and is lost. Thus grit and other contaminants increase the rate of wear within a worm drive.

Many worm driven telescopes do not use clutch mechanisms for some or all of the reasons mentioned above. However, a clutch mechanism is desirable since it serves the important function of preventing damage to the worm and gear should a power outage occur during a high speed slew or a physical obstruction be encountered while slewing. Worm drives without a clutch mechanism could be damaged under such conditions, particularly if the gear is driving a very heavy telescope.

Frequent Maintenance: The grease lubricating the worm and gear must be periodically cleaned off and replaced. If the grease isn't periodically replaced, then minute particles which have accumulated from the normal wear of the drive will accelerate the wear within the drive.

Chain drives are mentioned here merely for historical purposes since chain drive systems are rarely used anymore for large telescopes. Readers may recall the instruments manufactured by Autoscope which employed chain drives. Chain drives consist of one or more chains mounted under tension around two disks of unequal size. The chains are not unlike the chains commonly found on bicycles, but are of higher precision. Multiple chain drive reduction stages may be employed to achieve even greater pointing an tracking accuracy, and to increase the reduction between the telescope and the motor.

Chain drives have two inherent advantages:

• Very easy to implement since precise alignment between the disks is not very critical as long as forced tracking for the chain(s) is used.
• Extremely low cost compared to other types of drive systems.

Chain drives have several inherent disadvantages:

• Very low stiffness.
• Velocity errors arising from the thickness of each link, the length of each link, and the sizes of the drive disks.
• Breakage since the chain must be held under fairly high tension, which causes the pivot points in the chain's links to prematurely wear and eventually break.
• Frequent maintenance since the chain must be periodically oiled and checked for excessive rust and wear.

Very Low Stiffness: The stiffness of a chain drive is the poorest of any type of drive system examined within this article. The stiffness is limited by the average cross sectional area of each link within the chain, is further limited the compressibility of the cylindrical pins which couple the links together, and is further limited by the actual surface area of contact between each pin and mating link.

Velocity Errors: A chain, consisting of numerous links of given thickness and length, cannot precisely follow the ideal circular arc around either drive disk. This results in velocity errors as each link engages/disengages each drive disk. These velocity errors may be reduced, but never completely removed, by using relatively large disks compared to the thickness and length of each link within the chain, by using thin chains with closely spaced links, by using multiple chains with each chain's links being offset relative to the each other, or by using multiple chain drive reduction stages. Combinations of these methods may be used to greatly reduce velocity errors.

Breakage: A chain must be held under fairly high tension in order to achieve high pointing accuracy and to prevent telescope wind loading from causing tracking errors. This high tension causes rapid wear within each link's pivot and, eventually, breakage of the weakest link.

Constant Maintenance: If the chain is not periodically oiled, then the chain may eventually break due to the high friction and wear within the links of the chain. Conventional steel chains are also subject to rust.

Friction drives were developed to overcome the inherent deficiencies found in both worm drives and chain drives. A friction drive consists of a primary disk which is directly driven by an auxiliary disk of much smaller diameter. The auxiliary disk is in direct frictional contact with the primary disk. Additional friction drives may also be incorporated to further increase the reduction between the telescope and the motor.

Friction drives can have several potential advantages:

• Very high stiffness.
• No periodic error, assuming that the secondary disk is held precisely centered upon its axis.
• Potentially free of backlash, if auxiliary speed reduction components are carefully designed or if additional friction drive reduction stages are employed.
• High precision pointing, limited only by any disk decentering errors.

Friction drives do possess several inherent weaknesses:

• High precision tolerances for the primary and the secondary disks.
• Extremely small area of contact between the two drive disks.
• Extremely high engagement pressure, necessary to prevent slippage, at the point of contact between the primary and secondary disks.
• Precision bearings, necessary to precisely center and support the small secondary disk, and which are capable of withstanding the extremely high engagement pressures.
• Precise alignment, required to prevent rapid wear of the primary and secondary disks.
• Surface deformation, caused by the extremely high engagement pressures and the extremely small area of contact between the disks.
• Susceptibility to slippage.
• Susceptibility to contaminants which can permanently damage the drive system.

High Precision: The surfaces of the primary and secondary disks must be ground to extremely high tolerances since any surface errors will be directly transmitted to the telescope's drive axis.

Extremely Small Area of Contact: The primary and secondary drive disks are only in contact along the line where the two disks actually touch each other. The surface area of contact along this line is theoretically zero. The actual surface area along this line of contact is in reality not zero, but rather is a very small surface area. The actual total surface area of contact is related to each disk's microscopic surface roughness and the engagement pressure between the disks. The microscopic surface roughness widens the surface area of contact to a finite value. The engagement pressure slightly deforms the surfaces of each disk along the line of contact, thereby further widening the total surface area of contact. Nevertheless, the total surface area of contact between the two disks remains extremely small.

Extremely High Engagement Pressure: Due to the extremely small area of contact between the friction disks and the inertia of the telescope, it follows that the smaller disk must be engaged to the larger disk with extremely high pressure in order to prevent slippage between the disks. This strong engagement pressure can easily approach the surface deformation limits of the structural materials used for the disks, resulting in surface hardening over time.

Since the actual surface area of contact between the two disks is extremely small, it follows that the surface of the large disk can become deformed over time due to the surface hardening and wear occurring from the repeated back and forth slewing of the telescope. Slewing the telescope to all observable areas of the night sky uses only a portion of the large disk, with the portion corresponding to the zenith area generally being used the most. As a result, the larger disk can develop a slightly elliptical shape (or other shape) over time, dependent upon the range of slewing operations which are normally conducted with the telescope. Surface wear and deformation can be alleviated by the use of very large primary drive disks which permit the use of reduced engagement pressures, extremely hard disk materials such as stainless steel or titanium, or by using methods to harden the surfaces. Combinations of these methods may be employed to significantly reduce surface wear and deformation.

Precision Bearings: The bearings which support the smaller disk MUST be of very high precision and be capable of withstanding the high pressure necessary to couple the two disks in strong frictional contact. Any decentering errors or errors due to wear within the bearings supporting the small disk can result in observable tracking errors. Large friction driven telescopes generally compensate for these errors by using Reneshaw tape encoders, mounted around the larger drive disk and operating in closed loop fashion, to give continuous feedback of the exact position of the telescope while slewing or tracking.

Precise Alignment: The two disks must be carefully aligned such the axes of revolution of both disks are precisely parallel. Any substantial misalignment reduces the surface area of contact between the disks, possibly allowing the engagement pressure to exceed the surface deformation limits of the disks, and will cause premature wear to the disks. An analogy would be the rapid and premature wear to a car's tire tread if one wheel is out of alignment compared to the car's other tires.

Susceptibility to Slippage: Only so much engagement pressure may be applied before the engagement pressure approaches the surface deformation limits of the disk materials. As a result, a friction drive is susceptible to slippage if the telescope is operated in a significantly out of balance condition. Slippage will cause very slight wear and thus permanent damage to the surfaces of the drive disks.

Susceptibility to Contaminants: Any contaminants which get between the two disks can become embedded into the surfaces of the disks due to the necessarily high engagement pressure required to maintain frictional contact without slippage. Embedded contaminants can result is permanent damage to the disk(s) and sudden tracking errors. The duration and magnitude of these sudden tracking errors is dependent upon the diameter of the primary drive disk and the diameter and surface height of the embedded contaminant.

The hybrid band-worm drive system consists of a small disk coupled to a much larger disk via a stainless steel band. The large disk is directly coupled to the telescope's axis of revolution. A small high precision worm drive assembly is coupled to the small disk. The small disk and its coupled worm drive assembly are supported by a tensioning mechanism for the band drive component which also prevents ambient temperature fluctuations from affecting the pointing and tracking performance of the drive system.

Hybrid band-worm drive systems possess several advantages:

• Extremely low pressure between the stainless steel band and disks, due to the very large surface area of contact. The average pressure can be well under 100 psi. This very low pressure, unlike the high engagement pressure inherent in friction drives and the high tooth loading inherent in worm drives, prevents any kind of wear from occurring within the band drive system.
• Extremely low load between worm and gear within the worm drive component, due to the speed/load reduction provided by the band drive component.
• Virtually zero vibration during high speed slews since any vibration caused by periodic error within the worm drive component is substantially reduced by the reduction ratio of the band drive component.
• Immunity to damage from grit or contaminants due to the very low pressures between the band and the disks, and the extremely low load between the worm and the gear. The pressures are magnitudes below the surface deformation limits of the surfaces involved, particularly when compared to either conventional worm drive or friction drive systems.
• Extremely small periodic error since the periodic error of the worm drive component is reduced by the band drive component to about 1 arc second without employing any form of PEC software.
• Zero wear within the band drive component over the lifetime of the drive system.
• Extremely slight wear within the worm drive component over the lifetime of the drive system due to the very low load presented to the worm and gear by the band drive reduction. This also means that any programmed sub-arcsecond periodic error correction for the worm drive component will remain valid for years.
• High precision pointing, limited only by any disk decentering errors.
• High precision tracking, limited only by any periodic error within the worm gear component which is not corrected by PEC software.
• Very low cost compared to large friction and worm drive systems.

Hybrid band-worm drive system weaknesses:

• Moderate stiffness.

Moderate Stiffness: The only inherent weakness within a hybrid band-worm drive is that the stiffness of the band drive component is less than a worm drive yet is significantly greater than a chain drive. A properly designed hybrid band-worm drive which uses a single stage band drive component is capable of accurate tracking in winds up to 20 MPH, and is capable of accurate tracking in even stronger winds if the telescope is housed in a dome. The stiffness of a hybrid band-worm drive may easily be increased by a factor of four simply by implementing an additional band drive reduction stage. This permits the use of a much larger secondary pulley and a much thicker band on the band drive component's primary stage.

Worm drive systems have been around for decades and are still commonly employed as a reasonably cost effective solution for small to medium sized telescopes and mounts. Large worm drive systems are expensive but, if carefully implemented, can yield very good performance for large aperture telescopes. Worm drive systems do have several inherent design limitations, most of which can be addressed by the manufacturer. Only worm drive systems with very small amounts of periodic error are suitable for high speed slewing requirements.

Chain drive systems enjoyed a very brief period of popularity, particularly due to the influence of Autoscope during that company's brief period of existence and due to the limited degree of robotic automation and pointing precision which was available or necessary at the time. Chain drive systems have an extremely low stiffness value that makes them unsuitable for precision tracking applications if the telescope is exposed to significant amounts of wind. Chain drives may be slewed at high speeds without damage, yet noticeable telescope vibration is likely to occur.

Friction drive systems are expensive yet provide observatory class performance and possess very high stiffness. Friction drive systems are suitable for use in extremely windy conditions as long as the wind does not cause the friction drive to slip. However, friction drives can be easily damaged by contaminants or by slippage between the disks. Friction drive systems may be slewed at high speeds if the telescope is well balanced.

Hybrid band-worm drives combine the inherent advantages of the other types of drive systems discussed while discarding virtually all of their disadvantages. Hybrid band-worm drives yield observatory class pointing and tracking performance. Hybrid band-worm drives are immune to damage from contaminants and are immune to damage from slippage. Hybrid band-worm drives may be slewed at high speeds even if the telescope is somewhat out of balance. Additional band drive reduction stages may be employed to substantially increase the stiffness such that the telescope is capable of accurately tracking in extremely windy conditions. Hybrid band-worm drives will not slip even in extremely windy conditions or when the telescope is significantly out of balance.

Researched and written by,

Michael Marcus
Senior Engineer, ObservatoryScope