Notes on restoring cold head speeds following power failure (from session with A. Conrad and R. Campbell February 20, 2007)

After a power failure, speed heads go back to full blast (72RPM) reflected by values of CH1SPDRBV and CH2SPDRBV of around 2800 milivolts. The keyword to switch the cold heads back to remote control, will reset these to the slower speed (24 RPM, reflected by values for CH1SPDRBV and CH2SPDRBV of around 900 milivolts). Important note: setting CHPREMOTE=1 also turns off the cold head power, so must be followed immediately by mm ch1pwr=1 and mm ch2pwr=1 to turn the power back on.

Also note that if the cold heads are left running full blast for many hours, the tshield value (reported by showtemps) will get down to 120K or so, well below its normal value of around 180K).

The step-by-step procedure is as follows:

1. sm ch1spdrbv ch2spdrbv (expect values around 2800 milivolts)
2. mm chremote=1 (then do the next two steps promptly)
3. mm ch1pwr=1
4. mm ch2pwr=1
5. sm ch1spdrbv ch2spdrbv (expect values around 900 milivolts)
6. mm ch1spd=333.33 and mm ch2spd=250.0 (Recommended as of Feb 2012) (**) (related: Note on August 28, 2010)
7. Monitor showtemps for a few hours to make sure values are nominal (If the shield temp is going up, that usually means no power to cold head #2, for example)

Typical showtemps:

Wed Feb 14 12:27:52 HST 2007
                      tdetblck = 28.807 degK
                       tgetter = 15.069 degK
                      thead2lo = 10.933 degK
                       tcamera = 51.119 degK
                         tcoll = 50.546 degK
                        tbench = 50.639 degK
                       tshield = 175.530 degK
                        thead1 = 162.160 degK (*)

                      tc_setpa = 29.00  
                       tempdet = 29.000 degK

                      tc_setpb = 50.00  
                      thead2hi = 50.012 degK

                    htrdetblck = 23.400 percent
                    htrhead2hi = 12.600 percent
    

1. (*) thead1 has been reading correctly since the Service Mission in Sep 2011; the known intermittent short that can result in a mis-reading has been fixed.
2. (**) To address decreasing margin in the detector block heater, we are experimenting with running ch1 at a slightly higher speed (27 RPM). This corresponds to a setting for the ch1spd keyword of 250 mV. The decreasing margin may be a sign that the ambient temperature in the AO enclosure has increased since commissioning, or possibly that the cold heads are showing signs of wear and tear. This is under investigation.

Note regarding August 28, 2010 bench temp increase

1. Prior to a power bump on August 22, 2010 (summit CCR's off for about an hour), ch1spd had been set to 250 mV (28 RPM). The 250 value has been used off and on over the last few years, especially when the ambient temperature within AO is suspected of being higher than normal.
2. After the bump, ch1spd was left at the default of 0 (24 RPM).
3. The bench temperature crept up and, after one week, it reached a temperature high enough to trip the email alarm (about 52.5K).
4. Normally that temperature (tbench) is around 51K.
5. We set ch1spd=3500 mV (66 RPM) and ch2spd=500 mV (30 RPM), and, as of this writing, will probably leave it like that until the shield gets down to about 170.
6. These speed settings brought the bench temp back down to 51K after about 3 hours.
7. NIRC2 temperature plots for the period surrounding this event are here.
8. OSIRIS temperature plots for the period surrounding this event (these contain the best indicator of AO enclosure ambient temps) are here.