The telescope uses an alt-az mount. There are single azimuth, altitude, and rotation motors. There are three actuators attached to the secondary mirror to allow motorized focus and tilt (but not translation) adjustment. In September 2008, a new tertiary mount was installed that has a motor for tertiary rotation.
The entire telescope rests on the main azimuth bearing at the base of the cone. This bearing cannot be visually inspected; no serious investigation of it was made until May 2004. At that time, it was determined that the bearing is likely NOT the bearing called out in the drawings, it appears to be a size smaller. Jon Davis believes that it is an SKF bearing.
During May 2004, the telescope developed difficultly moving in azimuth, prompting the investigation just mentioned. At this is time, it was determined there was a felt seal at the top of the bearing that was partially out of place, allowing some access to the bearing (for dirt, etc.) Insertion of some cable ties suggested there was little or no lubrication on the bearings. A visual probe was attempted and it was determined that the inner surface of the bearing does appear to be covered, with just a small access from the inner diameter that appears to also have a felt seal around it. At this point, it was decided that grease lubrication could only help the situation, so about 1/3 tube of grease was applied (Mobil fairly lightweight, type ???; this was done through a small grease port that did not have a grease fitting; we used such a fitting but the thread of the hole apparently didn't match the fitting well, so we could not get the fitting to stay in. After lubrication, a new felt seal was installed using masking tape.
After the lubrication, the telescope appeared to move much smoother, but we are still waiting to see if this is a long lasting improvement. The smoothness was quantified by attaching a force meter to the telescope structure near the rotator; it took about 40 lbs of force to get the telescope moving, and then about 25 lbs to keep it moving smoothly. Prior to the lubrication, it took 50 lbs to get it going and 35-40 lbs, irregularly, to keep it moving; the improvment was fairly dramatic.
Some additional lubrication was injected in November 2006, as the telescope seemed to be getting a bit more difficult to move.
Bill K. suggested that Mobil SHC 1500 grease was used, but Ben H. thinks that is unlikely and suggests Mobil SHC 460 was used. Nick M. says that all of the Mobil SHC greases are compatible and recommends SHC 460.
The behavior of the azimuth axis warrants monitoring with time, and presumably routine lubrication, although one cannot continue to add grease indefinitely without cleaning out the bearings at some point. Servicing the bearing, however, will require lifting the entire telescope.
In 2010, the pointing appeared to get noticeably worse. In November 11, Nick MacDonald helped to inspect the system. He recommended getting replacement bearings for the azimuth wheel bearings, and to replace the cam followers. We also increased the pre-load on the soft springs a bit.
Nick recommended Applied Industrial Technology in Albuquerque as a source for bearings. He determined that the az drive bearings are Fafnir S10PP2C3 (there are 16 of these), and that the az drive idlers are Migill CF1S (there are 16 of these).
The idlers/cam followers for both az and alt were rebuilt by Ben Harris summer 2013.
Az drive re-greased 2014 May 30 (see log).
From the drawings, it appears that the altitude and rotator bearings are sealed type bearings, i.e. without maintenance needs.
In Nov 11, Nick M. looked at the altitude encoder, and saw that lots of metal flakes were coming off of the altitutde drive. We removed the encoder and cleaned it up. When remounted, we left off on bracket that appeared to be overconstraining the system (put it in the toolchest). We also noticed that the fit of the encoder housing is very tight over the wheel, and wondered whether it should be remachined.
Nick recommended Applied Industrial Technology in Albuquerque as a source for bearings. He determined that the alt drive bearings are Nic 5368 (there are 4 of these), and that the alt drive idlers are Migill CF1/2S (there are 4 of these).
In Decebmer 2012, we pulled off the altitude encoder and replaced the bearings and cam followers. The bearings are press-fit into the housing. However, one of them was quite loose (the one where the encoder is), so we put it in with some LocTite to try to keep it in place. The internal shaft is loose in the bearings, but we think that with preload, it will engage the bearing (and this appears to be the case).
The telescope az, alt, and rotator drive motors were manufactured by Parker-Compumotor. Their technical support number is (800) 358-9070. The motors are Dynaserv series DR motors. There are several copies of the manual for the motor drivers around. There are copies in the Yellow Telescope Manual and the Big Blue Notebook. The motors and their drivers are matched sets. The model number of each motor is noted on the motor drivers. (Hard to see for sure, but best guess: Az drive DR1070E-115 (resolver, 70N, E means 8 inch, 115V) SN 93020400314, altitude DR1015B-115 (15N, B 6-inch) SN 93030900107, rot DR1060B-116 (60N, 6-inche) SN93031000075). These are Generation 1 drives (no longer made), and must be matched set.
( 2/12 Information on potential replacement drives from Parker/Computmotor: now any 1070 drive can match with any driver. New ones are digital. Motors have not changed in output. Footprint is different. Cabling to motor will be slightly different. Interface to the drivers will be quite different. New drives have a 36 pin connector instead of old 50 pin connector. Consult with Peter Mack, who has done stuff like this. Need to know what the I/O voltages used for controlare. ROM quote: DRG-1060B-115V motor+driver $6729. )
The motor drivers are located in the left-hand side of the computer rack behind the weather station. If a motor ceases to function, sometimes, you can gain clues to the problem by looking at the digital readout on the face of the driver. The digital readout will read "0" for normal operation. Other codes that might be displayed are in the Compumotor manual.
The motors can be tuned as described in the Compumotor manual, and this has been done several times, and should be periodically checked. To tune the motors, one uses an oscilloscope attached to the POSN and AGRND leads on each motor driver. You put the motors/controllers in a test mode via a recessed switch on the motor driver. This is supposed to generate a square wave which you can see on the oscilloscope. There are three adjustments that can be made via small pots: the FC (characteristic frequency) switch, the lim integral switch, and the DC gain. The manual describes these in a bit more detail. Generally when we have done this, we have never been able to get a great square wave out of the controllers, we see a steep rise, a gradual decline at the top of the wave, and then a steep decline. Adjusting the switches can change things, more from the FC switch than from lim integral; often one can see a "peak" at the beginning of the top of the wave, or a nonlinear top of the wave; generally we have tried to adjust these features so they are removed. Some of the adjustments can cause the motors to make loud, unhealthy noises, so one should make the ajustments gradually and be prepared to back them off reasonably quickly.
The altitude motor controller was adjusted July 2004: the FC switch was changed from 9 to B, limit integral was left at F, and DC Gain was left at 12:00. The azimuth motor was found at FC=9, Lim integral=5, and DC Gain=2:30.
The motors were adjusted 2/4/2006. At this time, the rotator controller exhibited some peculiarities, reported an error code occasionally referring to encoder connector problems, but no specific problem was identified, and eventually it started to function OK (I believe that this was (much) later identified as being due to bad cabling, which was replaced). The altitude motor was adjusted to better contact the altitude drive, causing notable improvement in slewing behaviour. Motor adjustments were as follows: rotator FC from D to 9, limit integral F to F, gain to 45 degrees (highest before noise), altitude B to F, limit integral F to F, gain to highest before noise, azimuth FC 9 to 6, limit integral 5 to F, gain up.
In October 2006, the altitude and rotator motors were adjusted to attempt to better mate them to the drive surfaces. The rotator was chattering significantly. I was unable to get a very good mating of the rotator, but it was much quieter after adjustment.
The azimuth motor started to develop some serious vibrations in fall 2006. We had a several hour loud vibration in early November. After this, the mating was inspected with the boroscope and didn't seem too bad. After consideration of the motor mounting, we decided to back way off on the spring at the bottom of the motor, as it seemed to be overconstraining the motor in the wrong direction. The soft spring was adjusted out until moving the azimuth wouldn't move the motor, then back in until the motor moved both directions when the telescope was rotated, then tightened 1/4 turn further in.
We looked again at motor tuning with Ed in late Jan 2013. Made some adjustments based both on test waveforms and also by tuning while looking at encoder feedback at various positions on sky. On azimuth, set to FC=7, LIM=E, ACGAIN=MIN DCGAIN=tick 4 (all no change). On altitude, FC=C->E, LIM=E->1, AC=MIN, DCGAIN=2.25->3.0. On rotator, FC=C, LIM=F, AC=MIN, DC=3.75 (all no change). Saw some poorer behavior (according to encoders) in az at high az rate, tried to increase gain, but that led to higher frequency vibration.
Encoders are currently implemented on azimuth and altitude axes. At one point, an encoder on the rotator was considered, but it is not currently implemented.
The telescope encoders are all manufactured by Computer Optical Products of Chatsworth California. Their phone number is (800) 340-0404. Their website is http://www.opticalencoder.com. The encoder part numbers are written on the sides. All are model CP-850-HCE-128K (alt CP-850-HCE-128-F?) encoders. The differences in the three axes all have to do with their mounting configuration. The azimuth encoder has a square flange mount and cable coming from the back of the encoder. The elevation has the cable coming from the back of the encoder, but is manufactured without the flange. The rotator encoder has the flange but a side-cable. These are important distinctions if replacements are ordered.
Limit switches are implemented on the azimuth and altitude axes. No limit switches are implemented on the rotator which means that extra care needs to be taken to avoid rotator cable wrap issues.
The limit switches throughout the 1-meter system are commonly available microswitches. Part numbers are printed on the switch housings and generally the switches can be ordered through such vendors as DigiKey and Mouser Electronics. The smaller roller-rocker switches used in the mirror covers are available at Radio Shack.
Most of the limits in the system are easy to understand with inspection. The limits that are more complicated are the azimuth limits. The azimuth limits make use of two microswitches and a toggle switch to give a total range of 540 degrees. The azimuth limits work as follows: When the toggle switch is down, the south microswitch is active; when the toggle switch is up, the north microswitch is active. When the telescope is at the south limit switch, it is pointed almost east, i.e. at azimuth approximately 450. As the telescope is rotated counterclockwise (from above) from this limit, it will trip the toggle switch at azimuth approximately 265, which is just after the telescope has passed the north limit switch. Then the telescope gets another 360 degrees of rotation before it hits the north limit switch at azimuth approximately -90.
The toggle switch is activated by a Delryn trigger mounted to an L-bracket from the azimuth disk. If the toggle switch fails, the azimuth limits fail. The toggle switch is a simple DPDT switch commonly available at Radio Shack. Also, in an emergency, Apache Point keeps a stock of suitable toggle switches on hand.
There are also software limits implemented. In azimuth, these allow motion 315 degrees CCW and 140 degrees CW from the home sensor. Since the home sensor is at about 305 degrees (N=0), this corresponds to azimuth positions of -10 to 445 degrees.
Motor control is implemented through the use of some Oregon MicroSystems motor control cards in the telescope computer.
There are three control cards:
For the telescope and guider/tertiary control, we have used a OMS PC38-6E (with encoder option) for the telescope and OMS PC38-6 (without encoder option) for the guider/tertiary control. For the dome control, we only use simple I/O bits to turn motors on and off, and are using an OMS PC34 card. These all have adjustable addresses based on jumpers on the cards (see the manuals for specifics); the telescope card is at 0x310, the dome card at 0x314, and the guider card at 0x318.
These three cards are all ISA bus cards that run under DOS. Oregon Micro Systems still exists (under Pro-Dex), but they have discontinued construction of the PC3x series. At the end of 2011, we ordered a ISA OMS PC48-6E (roughly $1900, S/N 3454), which is supposed to be a drop-in replacement. We installed it in February 2012 in place of the OMS PC38-6E. Oregon Micro Systems says they will discontinue production of these (PC4x) by the end of 2012. Documentation on the PC48 is available via a PDF manual.
The PC48E did not immediately work completely: it worked for the XYZ axes (azimuth, altitude, rotator), but not for the TUV axes (secondary actuators). The card had some voltage leakage that OMS attributed to the encoder option (for UV axes), so OMS issued a RMA and repaired this issue (from OMS: per EW 138 U55 was removed and pin 6,7,9,10 were clipped off of U65). However, this did not resolved the issue. Turns out the motor controllers for the secondary actuators (Anaheim Automation model MBL-500) have 10 ohm pull-down resistors for step, direction, and enable, while the OMS cards require pull-up resistors. How it was working with the PC38 is a bit of a mystery: perhaps those cards were specially modified, or perhaps they had ``beneficial'' leakage.
With the PC38-6E we were getting intermittent "hangs" of the system which we think may have been attributable to the card (but not sure!). Nonetheless, it is probably a viable spare at some level, especially for the guider PC38-6. It has been left in the computer, configured at base address 0x300.
Because of the end-of-life issues, we also ordered another spare PC48-6E (S/N 3489).
The primary maintenance required to keep the telescope in good operating condition is routine inspection and cleaning of all drive surfaces on the telescope. This is critical to keep motor and encoder rollers from slipping. Use Kimwipes and acetone to clean the azimuth, elevation and rotator drive surfaces at minimum once a month. During moth season, this may need to be done once every few days.
Periodic rotation of the telescope over its full range of motion is advisable to insure smooth motion.
Periodic lubrication of the azimuth bearing may be advisable.