JKT FAULT TRACING

1994 August 10

1. INTRODUCTION:
   

   This document is intended to assist observers in tracing and rectifying common (and not so common) faults at the JKT. Some will require a Duty Tech to fix, and some are features (albeit unwelcome) of the system, and some are known long-term problems, which nevertheless require logging. Hopefully this guide will help with at least some of the problems you may encounter.

2. DEFECT DATABASE:
   ALL problems should be logged within the defect database, whether a bulb has blown or you find the primary lying in peices on the observing floor! Log in to the FAULT account on LPVS3, and complete the template presented for you. Commit this to the database by pressing PF3 and quit the database using PF4. To examine the status of any defect report, login to account QUERY, where a similar template awaits. Press ENTER, then use the keypad 7 key to select JKT in the telescope panel, then PF1 to search for all "JKT" fault reports. Use the arrow keys to search forwards or backwards, and PF4 to quit.
3. ZEROSET:
   One fairly common problem during encoder zeroset is that only one pair of lights come on. The easy escape is to press ACK, then ZEROSET again, without moving the telescope. Both pairs of lights will now come on. Drive the telescope back through the zenith until these go out, as usual.
4. NO STARS SEEN:
   Blindingly obvious, but are the dome and mirror petals open (do'nt laugh, it's been done!)? Is it cloudy? Are the position and equinox in the TCS correct? If you're using the TV have you typed TV on the ICS to put the TV flat in, and is the camera turned on, and the delay light out (if not, back the gain control off anticlockwise until it goes out)? If you're using the CCD not the TV does the shutter open and close - check the Mimic. Is the dome in the correct position: check the information display, and compare the demanded and actual dome azimuths - the dome may have lost its encoder position.
5. NO GUIDE STARS FOUND:
   First check that you have entered the correct position and equinox into the GSS, and that the APERTURE command was used if using BEAMSWITCH/A on the TCS. Are the guide probes at their requested position on the Mimic? If you need more help consult the JKT Quality Control folder in the control folder.
6. RISING FLOOR WO'NT!:
   There are a number of microswitches which can cause the rising floor not to work. One of these is under the vertical metal shutter on the NW wall - if this is even partially open, the floor wo'nt work. There is a nasty microswitch on the ladder to the floor - this is tempramental and may need to be pushed down or lifted slightly for the floor to work. Unless you've hit a limit switch, the other possible culprits are the microswitch on the gate, and the emergency stop button by the gate - ensure this is fully out.
7. DOME NOT MOVING:
   Check that the DOME on the manual controls panel is not lit - if it is then press it to return to automatic mode. Otherwise check the control panel in the dome, mounted high on the W side. The switches for both shutter and rotation must both be on REMOTE, not local.
8. IMAGES ELONGATED OR SHOW JUMPS:
   A number of possibilities exist, two main ones being mechanical. If the telescope is slightly out of balance then this can cause either of these faults. Another likely cause is unfortunately due to the worm and wheel drive arrangments used for both axes. It is possible that the worm drives on one flank of the wheel, and then on the other. There is an anti-backlash drive to prevent this but it still happens, and is very difficult to reproduce in daytime, let alone cure. One potential solution is to replace the worm with a spur gear, but this will be some way off. If images are elongated or show evidence of jumping, check the Baldwin displays, in the cream panel atop the control desk. The two displays on the RHS show the RA (top) and Dec encoder rates. When tracking these should both read less than 5 units (0.25 arcs) - if not, log in the database giving as many details as possible (telescope attitude, autoguiding? etc). Other contributing factors are wind loading on the telescope, or attempting to guide on a star which is too faint for the prevailing conditions.
9. FIELD VIGNETTING:
   If there is evidence of vignetting on CCD frames, suspect that the filter wheel is out of position, as not much else can be responsible in the A&G box. There are notes on how to check the status of the main CCD and autoguider filters in the JKT Quality Control folder in the control room. In the first instance though it's probably easier to go and check the filter wheel, and make sure that it's located securely, with the clamp beneath hand-tight. You must do a Zeroset from the ICS to initialise the filter wheel afterwards.
10. UNABLE TO IDENTIFY FIELD ON TV:
    Check that you have the correct position and equinox on the TCS. On the TV control unit there are two switches, which control field orientation. The nominal position for these switches is shown by red arrows.
11. TELEMETRY OUT OF RANGE:
    If the temperature is out of range (CCD warming up), fill immediately. TEK4 has an off-centre filler tube which means that at very high and very low ZD's, then liquid Nitrogen can syphon out, rapidly reducing the hold-time. The danger is under-filling the cryostat: continue to fill until liquid Nitrogen is seen escaping from the filler tube or side vent. If the 5v digital line goes out of range, ignore it, but log it in the database.
12. TIMEOUT ON CLEAR OR BIAS LIGHT:
    This is not the same as the old CCD Timeout error, but is often mistaken for it. It occurs during clearing of the chip, and attempts have been made to change the microcode to eliminate it, for TEK only. Please log all instances of this fault in the database.
13. LEXIDATA HUNG:
    It is possible for the Lexidata to hang if you inadvertantly do something naughty (e.g. using the Lexidata or the tape drives whilst reading out). When this happens, none of the Lexidata controls light, and both these controls and the joystick are locked up. If LEXINIT has no effect, the next thing is to reset the Lexidata controller. This is in the ICS room: remove the front RH panel - on the very bottom is the Lexidata 3400 controller, with a black on/off switch on the LHS. Turn the power off, wait 10 seconds, then power it on again. Type LEXINIT on the ICS terminal, and try again. If it's still hung, then SHUTDOWN and SO from the Perkin-Elmer, then type STOPALL on the ICS system console, repeat the power cycling of the Lexidata controller as above, then type STARTUP on the ICS system console, and when the system is configured, sign back in and bring up Adam again.
14. SERVO AMPS ALARM:
    This alarm sounds occasionally, and may or may not stop the telescope tracking. This condition is very hard to reproduce in daytime, so if the alarm sounds please do'nt cancal it immediately. Go downstairs and enter the TCS room (CLIP centre), which is opposite the main door. On your immediate left is a bank of four racks. These racks are interchangable between the INT and JKT, RA and DEC, drive and torque (antibacklash), so they're not clearly labelled. If any rack shows a red fault light, note which rack it is, then enter it in the database, once you've cancelled the alarm.
15. HYDROGEN ALARM:
    There are two gas alarms mounted on the control room wall, relating to the photographic plate Hydrogen soak plant outside. If these sound they can be muted by unscrewing the transparent covers, and pressing the mute button. Call the Duty Teck, who can gain access to the plant using the keys in the key press in my office.