Interpretation of the PS1 Temperature Sensors ¶

There are about 40 temperature sensors located in various places throughout the dome, the site, and on the telescope itself. Most of these sensors have now been calibrated so that they should be accurate to about 0.1 C. A few of them have not been. In particular, sensors under the M1 mirror are currently impossible to physically get to. Based on errors that we have seen in the other sensors prior to calibration, the M1 sensors could have errors as large as 2C. So be wary when trying to interpret these readings. The table below tries to describe the location of these sensors along with an indication of how well they have been calibrated. This table also gives you which listing in the camera FITS headers correspond to each sensor and a listing of the labels on each of the sensor connectors. This information is required reading for anyone trying to make sense of the FITS header temperature listings and should help with reassembly of the sensor array in case someone decides in the future that it is necessary to take things apart.

Note that the sensor location labels use a "clock" convention. The 12:00 o'clock position on M2 refers to the top of the mirror when the telescope is pointing at the horizon. The 3:00 M2 position is toward the telescope LEFT. This same convention is followed for the M1 sensor descriptions. In this case, the 12:00 o'clock position is also the top of the M1 mirror when the telescope is at the horizon and once again 3:00 o'clock is toward the telescope LEFT. There are 8 mid-truss leg (MTL) sensors and 8 Spider Vane sensors. A map showing the locations of these sensors is given below. Since the mid-truss legs (MTL) and the M2 support vanes are at 45 degrees from vertical, consistent with the clock convention we refer to their locations as MTL 12:30, etc. There are two spider vane supports that attach to each corner of the upper telescope truss. The vanes closest to the M1 mirror are called "Inboard" vanes. Those farthest from the M1 mirror are called "Outboard" vanes.

The calibration accuracy varys in three classes. Class A means that we believe that the sensors are accurate in an absolute sense to about 0.1 C. All of these sensors were calibrated by placing them into both boiling and ice water baths while recording their readings. These data were fit for both a zero point offset and a scale. Class B means that we believe the sensors are accurate to an absolute sense to about 0.2 C. The calibrations for these sensors were done in the same manner as the Class A sensors, but the data were not as clean or repeatable. For some of the sensors this was simply because they were physically difficult to get to and we could not spend the same amount of time on trying to calibrate them as the Class A sensors. Class C means that the sensors are essentially uncalibrated. All of the sensors on the bottom of the M1 mirror are Class C sensors. Based on the variations that we found when doing the calibrations on the other telescope sensors, we believe that there can be as much as 2 C errors in the zero points of these sensors. Their scale errors are probably less than 5%.

The table below gives the status of the temperature sensors as of 12-02-2009. In April 2009 we applied the temperature sensor calibrations to the EOST software. Prior to April 2009, all of the temperature sensor readings were uncalibrated. At the same time there was a change in the names of some of the temperature sensors. This configuration change did not make it into the camera software until 12-02-09. The result of this was that all of the values recorded in the TELTEMMS and TELTEMM1 FITS header strings were bogus values. This configuration change did NOT affect the other temperature sensors (except for the change in calibration values).

PS1 Temperature Sensors