ACIS Detector Assembly Heaters

Bob Goeke
Latest update: Sun Sep 20 10:45:22 1998 EST

Instrument Status

ACIS is currently integrated into the Science Instrument Module which, in turn, in integrated into the AXAF Observatory. CAT-2 Long Form Testing is complete.

For the purposes of this discussion one needs to know that all of the A side power going into the PSMC is knotted together, and the B side likewise but separately. Many lines, each separately fused, then go off in pairs (certainly for the heater board) to power the individual functions; e.g. Detector Assembly Heaters. At the Heater board, one line is led through a diode and then the two are knotted back together. In the absence of failure or noise (of which the RCTU provides plenty), one would expect to see the same voltage appearing at the input to all A side boards, and likewise all B side boards (but since the S/C provides A and B sides from a knot located all the way at the other end of the Observatory, A and B are rarely the same number). We have three telemetry points to observe these voltages on each side: the DPA board input, the DEA board input, and the Heater board input. The later, one hastens to add, is only available to the telemetry if that board is enabled, which says that the telemetry point is not quite on the input, but close. In looking back through the historical data -- like the Observatory TV test in May and June -- we see that all three points read exactly the same number on both A and B, irregardless of how ACIS is configured. In particular, the difference in voltage between the Heater board input and that of the DPA and DEA boards (on both A and B sides observed separately) is zero when the Heater board is operating either regulating at -60C or (in bakeout) at 25C. Certainly this was true on 6/18 when the B side heater was drawing 0.7A (we did a bakeout using Side A later that day).

Realization of an Anomalous Condition

On August 23 (234:67000 GMT) we were finishing up pieces of a Long Form Functional. The instrument was then in the ISIM thermal/vacuum test and had been regulating the Detector Assembly at -60C for some days using the B side of the PSMC. There was no difference in input voltage readings amongst the B side boards. For a period of 10 minutes we switched the DA temperature control to the A side to demonstrate stability and then switched back to the B side to let science testing continue. Unnoticed at that time, but clear in the data, is the first degradation of the B heater control board. With a output current of 0.5A, the telemetered input voltage was now 0.2V below that of the DEA and DPA inputs; since the DA was at a stable -60C as we switched from the A side to the B, no current surge to the output heater would have been expected, and none was seen in the data. We stayed in this mode for the next 3 days until August 26 (237:87500 GMT) when we stopped doing science and went into bakeout mode (for our 3 day soak before opening the door). The following plot shows the input voltage to the B side DA heater board (1DAHHBVO) as well as the input voltages to the DEA and DPA.

B Side PSMC in August 1998

August 98 Bus Voltage Plot

The 0.2V drop in the heater input voltage is still present (and was steady for the preceding 3 days). Upon going into bakeout mode, however, the gap increases over a period of perhaps 300s to around 0.7V.

Reviewing Historical Data

The only Side B bake done during the Observatory thermal-vacuum test (in May and June) happened at the beginning when we stayed at 25C while the cold shrouds were being filled; the heater currents were in the 0.7A range. During this limited bake, and during subsequent control at -60C, no statistically significant voltage differences existed between the heater board and the rest of the PSMC.

During the Ball T/V testing in January, however, we find some very interesting data for the B side. In the plot below we see the telemetered bus voltages as we came out of the eclipse simulation. For an initial 1500 seconds we were slewing back to -60C and were drawing an average of 1.1A; following that we commanded a slew to 25C at an average current of 2.0A. Using some data averaging one determines that during the first phase the voltages matched to 0.02 volts; during the second phase the voltages differed by 0.42 volts, with the same characteristic several hundred time constant on the leading edge.

B Side PSMC in January 1998

January 98 Bus Voltage Plot

At XRCF we had two B side bake out examples. The first on 4/17/97; the looks fiarly normal -- though the truly paranoid might see a ripple in the heater input voltage values. The second was on 5/16/97 which shows thecharacteristic voltage drop.

B Side PSMC in April 1998

April 1998 Bus Voltage Plot

B Side PSMC in May 1998

May 1998 Bus Voltage Plot

Numeric Historical Data

The following data have been massaged out of the data files; averaging has been applied to let us see through the instantaneous noise.

4/17/97 MF 2180 B 0.95 A 27.42 V
27.38 V
MF 9900 B
0.53 A 27.40 V
27.43 V
4/25/97 MF 11800 B 1.83 A 26.33 V
26.34 V
5/16/97 MF 24300 B 0.51 A 29.38 V
29.44 V
5/16/97 MF 24600 B 1.89 A 28.03 V
28.46 V
1/17/98 17:37000 B 1.14 V 23.56 V 23.58 V 23.59 V
1/17/98 17:39000 B 1.92 V 21.93 V 22.35 V 22.34 V
5/17/98 136:12500 A 0.9 A 27.58 V 27.55 V 27.50 V
B 0.33 A 32.38 V 32.52 V 32.51 V
6/18/98 168:42000 B 0.67 A 29.68 V 29.66 V 29.68 V
6/18/98 168:58500 A 1.96 A 28.97 V 28.61 V 28.63 V
6/19/98 169:7400 A 1.53 A 30.80 V 30.76 V 30.79 V
8/15/98 226:63000 A 0.7 A 30.80 V 30.76 V 30.79 V
8/22/98 233:72500 A 1.5 A 29.43 V 29.25 V 29.21 V
8/23/98 234:40000 B 0.5 A 30.57 V 30.57 V 30.59 V
8/23/98 234:64000 A 0.5 A 30.88 V 30.90 V 30.93 V
8/23/98 234:67000 B 0.5 A 31.00 V 31.18 V 31.18 V
8/27/98 238:30000 B 1,4 A 29,85 V 30,39 V 30.41 V
8/29/98 240:72000 B 1.4 A 27.51 V 28.05 V 28.06 V
8/31/98 241:70000 B 0.7 A 30.32 V 30.90 V 30.90 V
9/20/98 262:37500 B 0 A 30.42 V 30.93 V 30.93 V
9/20/98 262:38000 B 1.18 A 29.19 V 29.67 V 29.68 V
9/20/98 262:40000 A 1.18 A 30.47 V 30.40 V 30.40 V

Current Status

The instrument has stayed stable in its new mode. If one simply turns on the DA heater control boards with their set point of -60C, one draws no sensible power in the DA heaters. All three A telemetry input voltage points read the same number; the B Heater board input reads 0.3V lower than the DPA/DEA inputs. Putting both boards into bakeout, draws about 1.7A (DA at 20C trying for 25C). All three A telemetry inputs remained at the same reading (within the normal RCTU noise margins); the B Heater board input reads 0.5V lower that the DPA/DEA inputs.
Return to ACIS at TRW page.
There have been 78 accesses to this page since 9/23/2022.