TC Test with SUBKTEMP control

Data were downloaded from ops_1@chesterfield on 2009 Jun 13 with the export utility from HIPE (BUILDNUM 229). Converted format was selected.

bias voltages in NHK: 31.1, 31.3, 31.4, 127.6 mV (PSW, PMW, PLW, PTC)
bias frequency in NHK: 131.1 Hz
phases in NHK: 180.7, 180.7, 180.7, 180.7 deg (PSW, PMW, PLW, PTC) -- different settings from PTC headroom tests earlier in commissioning phase

Not much more than adding PHOTOFF/PHOTOFFxxx in to the PHOT/PHOTFARRAYxxx, using PHOT/TIME and PHOTOFF/TIME to time coordinate.

00002068a.pdf, first part of test starting at 11:58 UT; "HCSPTCTE_50976753_0002.IPF" control starts at ~800 seconds
00002068b.pdf, second part of test starting at 13:17 UT; "HCSPTCTE_50976753_0003.IPF" control starts at ~1000 seconds; "HCSPTCTE_50976753_0004.IPF" control starts at ~2000 seconds; "HCSPTCTE_50976753_0005.IPF" control starts at ~4100 seconds

Here is a summary of key parameters for the various settings:

HCSPTCTE_50976753_0002.IPF:
- set point: *change* of -150 ADU
- Kp = 30.0
- Ki = 0.8
- Kd = 0.0025
- low-pass filter gain, b1, b2: 1.565, -1.143, -0.413
HCSPTCTE_50976753_0003.IPF:
- set point: *change* of 0 ADU
- otherwise same as HCSPTCTE_50976753_0002
HCSPTCTE_50976753_0004.IPF:
- set point: *change* of 0 ADU
- low-pass filter gain, b1, b2: 74.855, 1.647, -0.701
- otherwise same as HCSPTCTE_50976753_0002
HCSPTCTE_50976753_0005.IPF:
- set point: *change* of 0 ADU
- Kp = 7.5
- Ki = 0.2
- otherwise same as HCSPTCTE_50976753_0002

All control settings in this test are stable against runaway. However, the low pass filter test "0004" showed some oscillation.
L0 temperature drift (approx. 0.5 mK) dominates the "a" part of this test. While SUBKTEMP is stable, the other temperatures drift in a manner consistent with increased power on the BDAs.
For the "b" part of the test, the cause of the slight warmer drift in the BDAs is not clear.
Once the integral term has built up to the proper value, stopping and restarting control stabilizes quickly -- except when Ki is changed.
The nominal control parameters ("0002"):
- stabilize SUBKTEMP in approximately 250 sec.
- but inject noise into the PTC thermometers which is about 10 times greater than with control off.
Test "0004" (change in low pass filter):
- induced ~5 sec oscillation in control, visible in PTC thermometers and TCHTR.
- (can't judge stabilization time)
Test "0005" (lower Kp and Ki):
- stabilizes SUBKTEMP in as little as 250 sec (but perhaps more in a fair comparison)
- substantially reduces the injected noise in the PTC thermometers, now about 3 times the noise with control off.
Note how cosmic-ray-free the PTC thermistors are.
There are at least two PTC thermometer signal levels jumps in these data, cause unknown:
- PTC-3, 11:54 UT, before my plots
- PTC-2, 12:01 UT, 200 sec. into "a" plot
- PTC-2, 12:04 UT, 350 sec. into "a" plot -- large
- PTC-2, 13:22 UT, 300 sec. into "b" plot

CDD, 2009 Jun 15