Joint Astronomy Centre
Show document only
JAC Home
JCMT
UKIRT
About the JAC
Contact Info
OMP
Outreach
Seminars
Staff-only Wiki
Weather
Web Cameras
____________________

Observing at JCMT
OMP Observation Manager
Telescope
Instrumentation
Schedule
Data Archive
Future Developments
Legacy Surveys
Newsletter & Publications
RxW Tuning Instructions
W logo! Receiver W Motor Error Recovery Instructions



Version 16 December 2008 

eh?


Occasionally the Martin Puplett interferometer drive motor servo systems get stuck.
  • This is indicated by the letter "E" for error appearing on the servoposition display for the affected servo.
Once an error occurs, the firmware in the RxW microprocessor will prevent any further drive to the motor.

  • If in doubt about the instructions below (they are complex), call KEN BROWN and get him to talk you through any part of the process you have questions about.
  • The only way to reset the error status is to reboot the microprocessor. 
The microprocessor can be rebooted using one of the two spring loaded toggle switches on the microprocessor board, behind the hinged panel in the upper left hand corner of the RxW electronics rack.

It takes about ten or fifteen  seconds for the process to finish. There are some LEDs on the board that will wink at you until the reboot is done.
  • Datum the servos.

Depending on the condition that has caused the error status, a reboot of the microprocessor may allow a new datumming of the servos, clearing the problem. This process can take about a full minute, because it is not possible to datum only the servo that needs it. All four servos are datummed whenever it is done. In many instances you will not be so lucky and the error status may reappear soon after the datum process for that servo begins.

When the error state can not be cleared with a reboot and re-datum, the affected motor can be driven with a DC power source and a special cable that has
been made up just for this purpose. Typically the motor will have been driven to one of its limits, either fully retracted or fully extended, and is jammed such that the current supplied by the driver electronics is not sufficient to unstick it. A power supply with more current capacity can be used to drive the motor out of the limits.

     The two upper interferometers, the SSB ones, are arranged such that the motors push downward against the return springs, and away from the zero datum position. The lower two interferometers, the LO diplexers, are arranged such that the motors push upwards against the springs, and away from the zero datum position. In all cases when the motor is fully retracted the position is near zero encoder units, and when it is fully extended it is at the high encoder unit values. If a motor gets stuck and you intend to drive it out of being stuck using the lab power supply, it is important that you determine whether it is stuck in the fully retracted position or the fully extended position, so that you will know which way the motor needs to be driven. When you connect the lab supply, you do not want to use the higher current capability to jam to motor even harder.

  • Is the motor at its fully retracted position?
    • YES

     When a motors is at its fully retracted position the datum switch should be actuated and the yellow LED lighted. If the motor is jammed in the fully retracted direction, it will be possible to manually push the mirror mount slider away from the motor shaft, against the force of the return spring, and observe the datum indicator LED turning on and off. The datum indicator LEDs can be viewed by opening the front panel. This is the panel which has all the servo select buttons and the control knob. There is a printed circuit card, to the far right hand end of the card frame, with four sets of three LEDs. Each set of three LEDs corresponds to one of the interferometer servos. In each set there are two green LEDs and one yellow LED. The green LEDs represent the two bit quadrature shaft position encoders, and the yellow LEDs represent the datum switch. The four sets of LEDs arranged top to bottom on the front of the card are for:

Channel B LO diplexer,
Channel B SSB interferometer,
Channel A LO diplexer,
Channel A SSB interferometer.

On a working servo channel you can rotate the knob to move the motor, and observe the two green LEDs turning on and off as the motor moves. If you drive the motor all the way in the negative direction towards zero on the readout, you can also make the datum LED turn on.

     If a motor is in the fully extended position, it will not be possible to push the mirror mount slider away from the motor shaft, and the yellow datum LED will not be lighted. In the fully extended position the spring force against the motor shaft is quite high, increasing the friction at the motor screw threads.

  • Remove the spring to make recovery easier. In some instances I have had to remove the motor to get it unstuck from the fully extended position. The motor is mounted with a hex nut. I believe it is a ½ inch nut, and it is very had to get a wrench (spanner) in the cramped space around it.


      A special cable has been built to drive the motors with a lab power supply. The cable has two 9 pin D connectors, one male and one female, and a red and black twisted pair of wires with banana plugs. This cable is being kept in the RxW cabinet in the Receiver Prep Room, in a box with other goodies used in servicing the interferometer servos.

  • The cable can be inserted between the motor and the normal wiring. This is done at the front of RxW near the motors. When this cable is in place the datum switch and motor shaft rotation encoders remain connected to the interface electronics in the electronics rack, but the motor power can only be supplied through the red and black wires. A lab power supply can be used to drive a motor out of being stuck. It is important that you first determine whether the motor is stuck fully retracted or fully extended. Connecting the red wire to positive and the black wire to negative supply voltage will drive the motor in the positive direction, away from fully retracted, towards higher encoder numbers. Reversing the polarity does the opposite.


I have found that 6 or 7 volts is usually sufficient to get the motor unstuck. Do not use higher than 10 volts. Driving a motor too fast can damage it. Once you hear the motor moving, you should be able to see the green LEDs on the interface card in the electronics rack blinking on and off.

  • Reduce the voltage so that the LEDs blink on and off slowly enough that you can see it. If the motor was stuck in the fully retracted position, drive it until the yellow datum LED goes out. If the motor was stuck in the fully extended position, drive it in the negative direction until the datum LED comes on, and then drive it back in the positive direction until the LED goes out. Be careful to stop driving the motor as soon as the datum LED comes on, so as not to jam it.


 (I have always used the Tektronix lab supply for this purpose. These power supplies have a current limiting function, which could be used to prevent damage to the motor. I have not yet determined what current is the maximum that should be used. So far I have just used the voltage control and have had no trouble. I will update this procedure to include use of the current limiting function, after I have determined the maximum current value.)

  • After the motor has been un-jammed, the servo will need to be datumed again. Replace the return spring if necessary. Datumming will not proceed for any servo with an error status. Error status can only be cleared by rebooting the microprocessor.



Contact: Per Friberg. Updated: Wed Dec 17 10:50:18 HST 2008

Return to top ^