Thermal Control Procedures

Revised 18 Feb 2010 by A. R. Walker

ACTR#002/00

DT: 18th October 1999
TO: Telops, Scistaff, ETS leaders
FM: ACTR and Maxime Boccas
RE: Revised operation instructions at the 4m-telescope for image quality program

Since the end of 1993, many improvements have been carried out at the 4m telescope to improve its image quality. The principal changes are: an active primary mirror support with lookup tables, a repolished f/8 secondary mirror, an image analyzer for the Cassegrain foci, removal of most heat sources inside the dome, active control of the primary temperature during the day, active ventilation of the primary cell at night, dome ventilation doors and active control of oil temperature.

1. Weather and thermal issues

The night assistant must check that all the environmental control functions explained below are working properly during the night.

1.1 General closure conditions:

ABSOLUTE closure when:

• wind > 45 mph on average over the last 20 minutes (the telescope might still not shake badly but the air is dusty and will bring too much dirt inside the domes)
• humidity > 92 %, as indicated by the met tower (typing "weat") and no other sensor
• T_mirror - T_dew < 2°C. The values of Tdew and Tmirror-Tdew are soon to be shown in "temp4m", which is actualized every 10 minutes. See calculating the risk of condensation (need link) for more information about dew point.

Dome re-opening:

• if wind <= 45mph on average over the last 20 minutes
• if RH < 85% for 30 consecutive minutes (in all cases, check outside that no water is dripping from the buildings).
• if Tmirror - Tdew >= 2°C for 30 consecutive minutes

In moderate wind: if wind > 25 mph, avoid observing in the wind direction as it could jitter the telescope and will bring dirt inside the dome.

BEWARE PLEASE: Don't permit observing beyond these limits! If you have trouble to be heard by the astronomer, please REPORT it (so we can politely explain the rules).

back to Contents

1.2. Dome opening:

The dome AND the mirror cover have to be opened 30 minutes before sunset, orienting the shutter eastward to make sure no direct sunlight will enter the dome. Dome and mirror cover have to be closed at the end of the night.

1.3. Lateral sliding doors opening:

These doors have to be opened just after sunset and left opened all night.

• If the wind > 15 mph on average, evaluate the image quality to decide whether to close them or not (elongated images mean you have to close the doors).
• If wind > 25 mph on average, ABSOLUTE closure (because of wind-shake on the telescope and deposit of dust).

1.4. Primary mirror cooling:

• 30 min before sunset: Observer Support has the responsibility to turn OFF the mirror cooler, turn ON the sucker and the extractors (P floor).
• End of the night: the night assistant has the responsibility to turn OFF the sucker and the extractors, and turn ON the mirror cooler.
• Remember to check the status of the mirror cooler and the extractor if you notice the seeing is very poor.
• Let's recall that a 1° temperature difference between the mirror surface and the ambient air will create additional mirror-seeing (typically up to 0.5" extra for 1° difference). Thus our goal is to maintain the mirror temperature within 1° of the predicted temperature of the following night.

 back to Contents

During the night, you are invited to type "temp4m" and check that:

• the 4 sensors of the mirror are within a 0.5-1° range
• the mirror temperature, dome temperature and outside temperature are within a 1° range
• Any problems detected should be reported by email.
• The mirror cooling should NOT be turned off for instrument change unless you are really freezing to death. If ever you do turn it off, DON'T FORGET to turn it back on!
• The 4mtemp control system automatically controls the mirror temperature according to the following logic: where T_goal is the lower of T_{low_dome} and T_outside

1/ T_mirror < T_goal - 2.5° : glycol valve closed (but ambient air is still blown onto the mirror)

2/ T_goal + 1.7° < T_mirror < T_goal -2.5°: glycol valve partially opened

3/ T_mirror > T_goal + 1.7 ° : glycol valve fully opened

4/ If T_mirror < T_{dew_point} + 7 ° : the glycol valve will close (safe limit to prevent condensation)

1.5. Floor cooling and stirring fan on main floor (M):

The cooling of the C floor and M floor in the dome is always ON. It is currently regulated manually.

The large stirring fan on the M floor (inside the dome) should always be ON when the dome is closed and OFF when the dome is opened. This fan improves greatly the air circulation inside the dome. It can be turned off during instrument change or any other works around the telescope if the noise bothers you. But turn it back on when you are done!

back to Contents

1.6. Oil cooler:

The cooling of the oil circulating under the R.A. rear and front pads is activated automatically when the pumps are turned ON by Observer support at sunset. The oil glycol valve is controlled (range is 1-2.25 V) by an equation involving the low dome and oil glycol temperature sensors. It is important that the oil cooling be working: if it is not, you can diagnostic it in "temp4m" (the "before pad" and "glycol" entries on the OIL line will quickly indicate around 30°).

1.7. Chimney fans:

These fans are located inside the chimney at the level of the primary mirror and suck ambient air into the mirror cell so as to form a laminar downward flow in the chimney. This is to prevent Cass cage heat sources from creating convection in the light path in the chimney.

At Cassegrain foci (f/8), Observer Support has to turn ON the chimney fans before sunset. The switch is on the power supply on the old console room desk. The night assistant must turn it OFF at the end of the night.

1.8. Air conditioning:

The current thermal plan in the building is not changed. Always leave the air conditioning at full power. Do not use heat sources at level M, MZ and C. Always maintain the doors closed in the passageways (especially the glass door in front of the lift at GR floor, the cryocooler compressor and pump rooms). In general, respect the signs in the building.

1.9. Temp4m and weat

Until we have a graphical web interface, type "temp4m" and "weat" to check the temperatures and weather conditions. Type, "/ut02/seeing/temper/tlog4m" for graphical analysis (the script should work at least on ctiot0 and ctioa2). We are working on a web page for thermal related issues.

back to top

2. Brief reminder of known optical problems

The image quality of the telescope can suffer deterioration mainly from 3 optical problems:

2.1. Astigmatism

 This is due to tiny slippage of M1 and/or deformation under its own gravity. Astigmatism will show up as elongated images, perpendicular on each side of the focus, especially for large H.A. (typically more than 2 hours). We compensate that with a lookup table, which controls the pressure in the air bags under M1. Thus the TCP toggle "CORR" should always be ON, at all foci. Note that if the F/8 secondary mirror lose its vacuum, strong astigmatism will immediately show up (typically 2-4 microns as measured by iman).

2.2. Coma

This aberration shows up in 2 cases: at Prime Focus if the primary mirror axis is not aligned with the PFADC and in the 2-mirror configuration (f/8) when there is a misalignment between them. Coma produces images with a flare (a tail like a comet) which is identical (same amplitude and direction) on each side of the focus.

At Prime Focus, we don't have a "quick-fix" tool to correct coma so any problem need to be reported (it hasn't happened, though).

Coma usually increases the further you go from the optical axis -this is called "field coma" and is normal (it is always supposed to be 0 on axis)- and also shows up when the optical axis of M1 and M2 are not coaligned -this is called "decentering coma" and is abnormal-. Our f/8 is a Ritchey-Chretien (RC) type, which means it is optically designed to correct the field coma. If the telescope is properly aligned there should not be any visible coma at f/8. If coma shows up anywhere in the field (it would usually be constant all over the field), there is a mirror misalignment and this can be corrected by using IMAN.

IMAN detailed instructions are on the web at Instructions for normal use and in the written manual next to the TCP ("User documentation: active optics system").

back to Contents

2.3. Defocus

Thermal expansion that can occur at night will cause the telescope to defocus. The prime focus changes at a rate of -110 microns/°C, the f/8 focus at -780microns/°C. The actual movement of He f/8 mirror is a factor 10 less (i.e. -78 microns/°C)

Use "temp4m" to check the Serrurier truss temperature and adjust the focus accordingly.

Mosaic can automatically check the Serrurier truss temperature and adjust the telescope focus accordingly before each exposure.

3. Prime Focus nights:

* CORR ON all the time.

* Temperature drift compensation:

Focus number increases when focal plane moves away from M1

Focus number is in microns neglecting the effect of the PFADC (the PFADC has a power of 1.076, i.e. multiply the focus change by 1.076 to get the real motion in microns in the focal plane).

Decrease focus setting by 110 units per 1° increase in temperature

 back to Contents

4. f/8 nights:

* CORR ON all the time.

* The toggle next to f/8 on the TCP status display should be OFF all the time (this is for an eventual coma lookup table ; if ever you need to turn it off, go in the TCP menu, and select /Active_Control/Tilt_Secondary/On-Off_Auto_Tilt/Off).

* Temperature drift compensation:

Focus number increases when focal plane (and M2) moves up

The focus number represents microns of motion of the focal plane (not of M2)

Decrease focus setting 780 units per 1° increase in temperature

*Collimation:this has to be checked according to the following procedure:

• On the first night of a f/8 run, check in the TCP menu that /Active_Control/Tilt_Secondary/Display_Tilt shows the tilt value (amplitude and azimuth) written on the white board and in the logbook. If this is not the case, choose /Absolute_Tilt and enter the right values. Next, run IMAN according to the instructions in the IMAN instruction book (a version is also available on the web). Write the new tilt values on the white board and in the logbook.
• At the beginning of each following night of the f/8 run, check /Display_Tilt as described above and correct with /Absolute_Tilt if necessary.

Back to Top

The images are automatically stored in the appropriate archive.

One image should be taken at the beginning of the night, after doing the telescope zpoint and focussing the instrument. ). The spectro mode of Osiris needs to be focused in the imaging mode (both modes are parfocal), thus it is easy to take that first seeing sequence.

A second image should be taken by the time it is 1 hour before morning twilight, preferably after the focus has been rechecked. If Osiris is used in spectro mode, you obviously need to change back to imaging mode to take this second seeing frame.

In case some problems forced you to start late (i.e. after 11PM), only a single measurement will be required.

For the time being, no seeing measurement is required with IRS. In the near future, we hope to do the seeing measurement with the Tip-Tilt camera for all IR instruments.

Back to Top