PDS_VERSION = PDS3 RECORD_TYPE = STREAM OBJECT = TEXT PUBLICATION_DATE = 1997-08-12 NOTE = "MGS RST Instrument Health Report HEA7224A.TXT" END_OBJECT = TEXT END This report is being submitted by Joe Twicken. Dick Simpson will be back in the office on Thursday, August 14. Engineering Data ================ Engineering Data Statistical Summary from CSV2STAT: First SCET: 1997-08-11T11:53:14.786 Last SCET: 1997-08-11T23:59:58.783 CHANNL CHANNEL TITLE DN-LO DN-HI EU-LO EU-HI EU-AVG/STDEV RECDS ------ ------------- ----- ----- --------- --------- ---------/-------- ----- F-0190 HGA_AZ_ANG 0 0 0.00E+00 0.00E+00 0.00E+00/0.00E+00 1340 F-0195 HGA_EL_ANG 0 0 0.00E+00 0.00E+00 0.00E+00/0.00E+00 1340 L-0034 MOT1_RANGING 0 0 6036 L-0036 MOT1_TLM_MOD 0 0 6040 L-0037 MOT1_USO_ENA 0 0 6036 L-0044 MOT2_RANGING 1 1 6036 L-0046 MOT2_TLM_MOD 1 1 6040 L-0047 MOT2_USO_ENA 1 1 6036 L-0053 RF_SW_OSC 2 2 6040 L-0060 TWTA1_FILMNT 0 0 6036 L-0061 TWTA1_HV 0 0 6045 L-0070 TWTA2_FILMNT 1 1 6045 L-0071 TWTA2_HV 0 0 6036 L-0080 USO_PWR 0 0 6036 L-0090 KaBLE_STAT 1 1 6036 L-0091 KaBLE_ENABLE 0 0 6045 L-0111 MOT1_RCV_AGC 205 255 -1.53E+02 -1.38E+02 -1.43E+02/7.61E-02 4115 L-0121 MOT2_RCV_AGC 155 255 -1.53E+02 -1.25E+02 -1.32E+02/1.62E-01 4115 L-0131 TWTA1_HLX_I 0 1 -3.84E-02 -1.85E-02 -3.83E-02/1.27E-05 4115 L-0132 TWTA1_ANOD_V 0 2 -1.00E+03 -1.00E+03 -1.00E+03/5.79E-03 2194 L-0141 TWTA2_HLX_I 39 43 7.25E-01 8.05E-01 7.64E-01/2.20E-04 4115 L-0142 TWTA2_ANOD_V 60 64 -8.68E+02 -8.60E+02 -8.63E+02/2.79E-02 2194 L-0200 USO_REG_V 203 205 4.06E+00 4.10E+00 4.08E+00/1.77E-04 2195 L-0201 USO_OVEN_V 123 126 2.46E+00 2.52E+00 2.49E+00/2.33E-04 2195 T-0302 EPC1_T 120 126 2.50E+01 2.77E+01 2.61E+01/1.76E-02 756 T-0303 EPC2_T 109 114 3.03E+01 3.24E+01 3.13E+01/1.80E-02 756 T-0310 RF_ISOLTR1_T 180 185 2.64E+01 2.82E+01 2.73E+01/5.68E-03 4119 T-0311 RF_ISOLTR2_T 171 177 2.93E+01 3.14E+01 3.04E+01/5.96E-03 4119 T-0312 TWT1_T 179 184 2.68E+01 2.86E+01 2.76E+01/1.68E-02 757 T-0313 TWT2_T 164 171 3.14E+01 3.38E+01 3.26E+01/1.76E-02 757 T-0315 USO_T 174 177 1.09E+01 1.18E+01 1.14E+01/6.26E-03 756 T-0316 KA_AMP_T 150 159 3.56E+01 3.90E+01 3.73E+01/1.81E-02 1235 L-0200, L-0201 and T-0315 remain nominal. Open Loop Data ============== Data Collection --------------- No new data. Anomalies --------- See discussion of RSSD analog-digital conversion below. Analyses -------- USO Test #13 (97/220): I have processed the USO Test #13 ODS from 97/220. This ODS was recorded at DSS 65. Telemetry and ranging were on during the test. The SNR was approximately 42 dB/Hz. This was the first USO Test ODS to be processed on our Sun Ultra 2 workstation ("magellan") running the Solaris operating system. I have only recently completed porting our software and scripts to that system and installing our QDDB MgsDB database there. All files for processing this ODS were recorded in the MgsDB. Our plan for mapping operations is that the SOPC will be used for data transfer, logging and quick-look analysis, while the magellan station will be used for data processing. One nice feature of the QDDB program is that it can run simultaneously on both machines (or with multiple users on either machine) with a single shared database, and will ensure that only one person (or script) has write permission to any particular record at any given time. We are expecting that the SOPC will be "upgraded" to Solaris around the time which mapping begins. We now know that our software and database can be ported to Solaris systems and run there with few complications. The ODS was acquired at DSS 65 from 14:12:10 to 16:12:09 at 5000 samples per second in mode 2 (1250 12-bit samples/second for each of 4 A/D's). The ODS was obtained following a retransmission and there were not any missing records. There was one duplicate record, however. That was ignored by our processing software. I am happy to report that there were no POCA errors of any kind in the ODS. The recently installed DSP-R software continues to be nearly POCA error free. I used the spk_c_970525-970907 SP-kernel file to steer the data. This file was created on 97/211. The power spectra of the USO Test #13 open-loop data exhibited 650 Hz spurs above and below the carrier frequency. We have now come to expect this when telemetry is on. There did not appear to be any other spurs in the power spectra. Amplitude fluctuations were very small, on the order of 0.2 dB peak to peak. There was no obvious period in the fluctuations. A histogram of the raw open-loop data samples (quantized to 12 bits) showed a series of discrete lines. Closer inspection revealed that the lines were all at quanitization values whose binary code ends with a string of 0's, for example lines at quantization values 1024, 1024+32, 1024+64, 1024+128, etc. Furthermore, each line was paired with a quantization value immediately to its left (i.e. next lowest value) at which there was a deficit in the histogram value. Apparently, the RSSD analog-digital conversion process does not operate uniformly across the entire range of quantization values. Values which end in a string of 0's appear to be significantly more likely to occur than those which end in a string of 1's (the next lowest neighbors). Looking back through our notebooks of USO Test and ORT data, this behavior is not new and is not exhibited at a single station only. We will further investigate the implications of the analog-digital conversion anomaly. I estimated the USO frequency to be 8423152961.21 Hz. The best linear fit to the USO frequency estimates since MGS was brought back from safe mode has a slope of 0.181 Hz/day. The standard deviation of the residual errors to this fit (for recordings since 97/162) is approximately 0.08 Hz. Prior to safe mode, the drift rate of the MGS USO was approximately 0.253 Hz/day. I have estimated the USO Allan deviation at a number of integration times from this ODS. The SNR during the test was approximately 42 dB/Hz. I did not compensate for the spacecraft spin cycle. The measurements confirm that the USO frequency stability has deteriorated since MGS entered the safe mode. The stability during the previous two USO tests (both conducted after the return from safe mode) indicated a deterioration. It was not possible to positively confirm it, however, because both of those tests were conducted at the DSS 25 BWG facility for which we had no other data to compare. The Allan deviation for this test follows: t (sec) AD ------ ---------- 0.4096 1.3630e-12 1.6384 4.0690e-13 3.2768 4.2951e-13 6.5536 4.4167e-13 26.2144 3.6724e-13 The frequency stability is unusual in this test because it does not improve appreciably with integration time. Even in the two prior tests at DSS 25, the Allan deviation was on the order of 1.4e-13 at an integration time of 26 seconds. RS ORT-4 (97/220): I have processed the six open-loop data segments acquired at DSS 45 as part of RS ORT-4. The ORT data sets were also processed on our Sun Ultra 2 workstation running Solaris. This was the first ORT data processed on that machine. Quite remarkably, the six simulated occultations were completely processed in less than 11 minutes!!! There were no missing records in any of the data sets. There were no apparent frequency spurs in the power spectra. SNR and carrier amplitude oscilllations were discussed in previous Instrument Health Reports. I estimated the USO frequency in the middle of the test to be 8423152961.1 Hz. The histograms of the raw data samples for each simulated occultations showed the same analog-digital conversion anomalies as did the the histogram for USO Test #13. Note that ORT-4 and USO Test 13 were not conducted at the same DSN station, although they were conducted on the same date. Closed Loop Data ================ Data Collection --------------- We have obtained the following files from OSCAR and EPIC: Original File Name Name on SOPC ------------------------------ ------------ ATDF: ODF: 97223C224.DAT;1 7223224A.ODF IONCAL: TROPCAL: WEATHER: UTPM: EOP: