layout of drive system (top)

• 8 motors drive the dome
• the motors are named : m11,m12,m21,m22,m31,m32,m41,m42
• pairs of motors are mounted together on a drive truck.
• (m11,m12),()m21,m22 are mounted on the uphill half of the dome (with m11 highest)
  
• m11,m12 are mounted together in a truck that rides on the rack gear beam
  
• m21,m22 are mounted together in a truck that rides on the rack gear beam
• m11,m12 are connected via a drag link to one side of a connection bracket
• m21,m22 are connected via a drag link to the other side of the same connection bracket.
• The dome is then connected to the bottom part of this connection bracket.
• m31,32,41,42  are mounted similar to the m11 thru 21 about 20 feet downhill
  

• There is 1 amplifier for each motor. These amplifiers are located in the vertex shelter (on top of the azimuth arm). They are called ampxx where xx is the motor number (eg amp11,amp12...)
• There are two power supplies that power the 8 amplifiers.
• power supply 1 powers amps 11,12,21,22
• power supply 2 power amps 31,32,41,42
• The dc bus of the amplifier ranges from 250Volts to 360 volts (regeneration can raise the bus voltage).
• More info on amps,motors,power supplies

12jul13: drag link comes undone, connection plate bent.

• The plots show 14:00 to 24:00 for 12jul13
• The dashed lines mark the servofail ato 15:30 and 22:00
• page 1 has the positions (az,za)
• pages 2-6 show various status values
• page 7 has the torques for the 8 motors 
• the drag link probably came undone a little after 22:00 (although we have no was of actually knowing this).
  

• Connection plate bent. view1 , view 2
• Drag link on pig disconnected from motor. view 1
• motor shield hits hydraulic brake mounting
• motor shield and hydraulic brake mounting   view 1, view 2
• motor shield (view 1)
• hydraulic brake plate (view 1)
• motor power cables into transition box, connector pulled out (view 1)
  

jun13:When the failures occurred (top)

• 15jun13 sat @ 19:11
• motor 22 torque increases 2 seconds between shutdown
  
• 16jun13 sun @ 21:34
• Failure:
• amp22 not ready
• amp 11,12,22fault (21 remained ok)
• motor 22 torque was oscillating for 6 minutes  prior to failure. changes 3 times larger than other amps.
  
• 17jun13 mon @ 15:34
• failure:
• amp 22 not ready. everything else remained normal
• Torque for all motors was increasing. mot 22 was the highest. at value43 ft-lbs.
  
• .. amp for motor 22 replace 16:00 17jun13
  
• 18jun13 tues @ 02:18
• Failure:
• amp 22 not ready. All others ok
• Torques
  
• Torque for all motors was increasing. mot 22 was the highest. at value 35 ft-lbs
• 18jun13 tues @ 10:14
• maintenance crew driving dome with pcu (in rate mode.. velocity loop only)..
  
• Failure:
• amp 22 not ready
• amp 11,12,21,22 amp fault
• dcbus1 fault
• power supply 1 drives motors 11,12,21,22.
  
• ac dc bus fault occurs when the amp dc bus goes above 360 Volts.
• this is probably why all 4 amps had the amp fault.
• Torques:
• motor 22 torque rose 1 or 2 secs before the failure. to about 33 ft lbs
  
• motors 31,32,41,42 torques went up to around 32 foot lbs 1 sec after the failure.
• They are on power supply 2. these are probably the torques used to stop the motion
  
• the power supply for the other motors had the dc1bus fault so they couldn't apply any torque. 
  

Plots of the system status during the failures  (top)

    The below plots were made using the status info recorded around the various failures. This status info is sampled once a second. The pages contain:

• Page 1: status Gregorian status.
  
• Page 2: control status, power and brake status
• Page 3: Amplifier status
• Page 4: motor torques.
  

The next set of plots has the position, velocity, and torques for all 5 failures (.ps) (.pdf):

• The vertical dashed line shows when the failure occurred.
• Page 1: Gregorian zenith angle:
• The failures did not occur at the same zenith angle.
• Page 2: Gregorian velocity.
• 15,16, and 18 2am were moving uphill
  
• 17  and 18 10am were moving downhill
• The black line is the velocity computed from the encoder
• the red line is the velocity coming from the amplifier monitor output.
• Page 3: the motor torques for the 5 failures.
• 15 and 18 10am had m22 torque increases prior to the failure
• 17 and 18 2am had all torques increasing prior to the failure
• 16 had the torque for m22 oscillating for about 6 minutes prior to the failure.
  

processing: x101/130618/azprob.pro

damage found and corrections after 10am 18jun13 failure  (top)

• amp 22 was changed on 17jun13 after the 15:00 failure.
  
• Maintenance personal reported a large noise at the failure
  
• The drag links did not look aligned for motor 21,22
• after moving the dome, the drag links realigned themselves.
• the brake on motor 22 was inspected. When applying the brake, you could still turn the motor with a wrench.
• motor 22 was replaced.
  
• A pin connects the drag link to the connecting plate and a second pin connects the other side to the drive truck.
• One of the cotter keys (pins) that keeps this pin from slipping out was broken.
• When trying to remove the pins for inspection, it was very difficult to remove them.
• spherical bushing are used where the drag link connects to the drive truck and the connection plate.
• Some of these spherical bushings were damaged and had to be replaced.
• Inspection of the uphill connection plate and dome connection showed no deformations.
• later tests showed that one of the wheels on a drive truck was worn down. It was replaced.