plots:
Azimuth summary info during failure (.ps)   (.pdf)
Gregorian summary info during failure (.ps)  (.pdf)
Az,za positions during failure (.ps) (.pdf)

The Failures.  (top)

• 19:00. The dome axis status went to local disable. This is the switch located on top of the dome that allows maintenance to disable motion of the dome. This remained till the next day.
• 19:00 Emergency stop became active.
• 20:20 jorge opens the cabinet doors in the vertex shelter
• 20:25 emergency stop comes and goes a few times and finally goes away. Jorge may have been doing resets.
• 22:42 Both azimuth encoders fail. Az encoder 1 jumps to -505.0651 degrees, az encoder 2 jumps to 360 degrees.Failure goes away about 45 minutes later.
• 23:20 the azimuth  local disable switch was turned by jorge for testing.
• 05apr07: 15:00 azimuth encoders reset using index pointer.

---

Cause of the failures.

Dome disable bit on. Gregorian summary info during failure (.ps)  (.pdf)

• Page 2 shows the emergency status bit coming on. It slso shows the cabinet doors opened by jorge at 20:20.
• Page 3 bottom plot yellow trace shows the dome disable bit coming on at 19:00.
• The following day a broken cable was found in the dome cable drape. It was the return current cable for the dome disable switch on top of the dome. This cable must have broken at 19:00 causing the dome disable (and probably the emergency stop) ot occur. The cable was respiced 05apr07.

Azimuth encoder failures.

Azimuth summary info during failure (.ps)   (.pdf)

• Page 3 shows the first encoder failures around 22:45.
• Page 4 bottom yellow trace shows the azimuth local disable switch being turned by jorge at 23:20 (the other jumps of this bit were when the system was rebooted causing all bits to jump).
Az, dome positions during failure (.ps)  (.pdf)
• The top plot shows the azimuth positions during the failure. Black in az encoder 1 (dome side). Red is azimuth encoder 2 (ch side). The encoder 2 trace has been offset by 1 degree for plotting purposes.
• At 22:45 both encoder positions jump.
• Enc1 went to -503 degrees while encoder 2 went to 360 degrees.
• 360 degrees is the preset position when reseting the encoders. Encoder 2 is an old encoder that probably had the value 360 degrees loaded into its memory. Encoder 1 was a new encoder and my guess is that it never had a reasonable value loaded into its preset memory value. That is why it came out to be -505.0561 degrees.
• So it looks like that both encoders were told to reset when the azimuth was at 437 degrees around 22:45.

---

Fixing the problems:

• On 05apr07 the broken cable in the dome cable drape was respliced. It was not obvious what it caught to cause it to break.
• The azimuth was moved to the 270 degrees reference on the encoder 1 side and then the encoders were set to 270 degrees. The carriage house was at stow position 8.8 degrees and the dome was at za=8.0 degrees.
• A pointing test was done 05apr07 and 06apr07 to check how well the azimuth was aligned with the position where the model was made. After the pointing tests, the az*sin(za) error was about -35 arcseconds. This is an error of -.1 inches at the encoder. On 06apr07 at 18:00 the models were updated to compensate for this error.