Mars Express Bistatic Radar Experiment Operations Plan and Report 16 April 2010 Summary ======= DSN Antenna: 63 Pass: N/A Orbit: 8056 Orbit Start Time: 2010-04-16T20:51:08 HGA Pointing: Specular Start Mid Point End -------- ---------- -------- Specular Condition (ERT): 21:43 22:25 23:06 Target Latitude (deg N): -14.84 -22.46 -32.09 Longitude (deg E): 326.05 317.42 310.27 Rp (km): 3394. 3394. 3394. Incidence/Reflection Angle (deg): 31.96 39.04 48.23 Slant Range (km): 10621. 9928. 8455. Slew Angle (deg): 116.08 101.92 83.54 Doppler (carrier, fd; Hz): -27693. -35688. -40782. Doppler (echo, fr; Hz): 1000. 8952. 15765. Doppler Difference (fdd; Hz): -28693. -44639. -56547. Earth-Mars Distance (m): 1.732E+11 1.732E+11 1.732E+11 NB: The DSN is no longer routinely using pass numbers; so this entry in the summary will not be maintained in this and future reports. Experiment Set Up ================= This experiment was conducted using the Madrid DSS 63, the first experiment at that complex since 2009/246. Danny Kahan and Gene Goltz were in the Radio Science Support Area (RSSA) at JPL. This data acquisition parts of this report are based largely on notes provided by Kahan and Goltz. Performance Problems and Notes ============================== The station reported only 'overcast' skies during both the pre-cal and post- cal, but there were large fluctuations in the X-Band noise power, especially during the final 15 minutes of observation (starting about 22:50), MINICAL #2, and the post-cal. X-RCP system temperature was 79K during the post-cal, compared with 20K during the pre-cal; X-LCP system temperature during the post-cal was 62K (lower than X-RCP, for unknown reasons) compared with 22K during the pre-cal. There were strong correlations between X-RCP and X-LCP sample powers when the front ends were connected to the antenna, suggesting that the problem was external to the station microwave and electronic systems. The ambient load measurements were very consistent; when the RSRs were not connected to the antenna, the RSR performance was nominal. S-Band variations follow those at X-Band starting about 22:50 except for being much less pronounced. The noise power in S-RCP and S-LCP peaked during MINICAL #2, then flattened out. The S-Band post-cals were comparable to the pre-cals. It is possible that the degradation was caused by very broad-band, off- boresight radio frequency interference (RFI). That would at least be consistent with the relatively sudden onset of the interference and the fact that X-RCP was more strongly affected than X-LCP. The intensity of the interference and its extent (both S- and X-Band) remain puzzling. The amplitude calibrations of the data were carried out using the measured values; but the results should be regarded skeptically because of the unexplained noise levels and variations. The RSR tuning predictions were inaccurate because Simpson calculated the spacecraft velocity incorrectly during experiment planning; the X-Band echoes moved off the upper edge of the 25 kHz pass band several times during the course of the observations. Kahan and Goltz tried to adjust the frequency offsets but with only limited success. The 25 kHz S-Band echoes were partly off the edge at the end of the observations. The 100 kHz data should be processed to get the science from these observations. The FGAIN setting was too low for this experiment; there was significant clipping of the directly propagating X-RCP signal during the spacecraft calibration (BCAL) periods. The pre-cal, post-cal, minicals, and surface observations were not affected. Station personnel neglected to issue the 'MFQ=ON' command to DC01 before starting the pre-cal. As a result, the noise diode was operating at a duty cycle of about 50 percent during the X-RCP pre-cal. The command was issued before the X-LCP noise diode was turned on; and, from that point forward, the X-Band noise diode performance was nominal (100 percent duty cycle) in both polarizations. At step 5 of the pre-cal, Goltz and Kahan learned that station personnel were using the old version of the BSR procedure, not the version which has been written into the Network Operating Procedures (NOP) document. This represented a breakdown in communication but should not affect results. A narrow band signal is present in both 100 kHz X-Band channels about 20 kHz above the surface echo at the beginning of surface observations and about 6 kHz at the end. Its source is not known; the fact that it has a frequency trajectory very similar (but not identical) to the surface echo is puzzling. There is an unexplained loss of gain in the S-LCP channel between 22:17 and 22:32. Power versus time plots for S-LCP show an inverted J signature with an initial voltage drop of about 5 percent (22:19), nearly full recovery, then a deeper drop of about 33 percent (22:31) before a nearly complete recovery by 22:32. Data Acquisition ================ RSRs were configured as in Table 1. Table 1 ---------------------------------------------------- RSR Channel Mode ATT FGAIN Operator dB ----- ------- ---- ---- ----- ---------------------- RSR1A S-LCP 1-W auto 60 UNK RSR1B X-LCP 1-W auto 60 UNK RSR2A S-RCP 1-W auto 60 UNK RSR2B X-RCP 1-W auto 60 UNK RSR subchannels (SCHAN) were defined as follows: Table 2 ---------------------------------------------------------------------- Subchannel Sample Rate Comments ---------- ----------- --------------------------------------------- 1 2 ksps Occultation bandwidth (not recorded for BSR) 2 8 ksps Occultation backup (not recorded for BSR) 3 25 ksps Primary recording bandwidth 4 100 ksps Backup recording Table 3 lists ADC amplitude levels read from RSR displays during the experiment. Times are in UTC and should be considered approximate. "Steps" are as defined in the briefing message. RSR ATT settings are in units proportional to dB. Acronyms and abbreviations are explained after Table 3. Table 3 ----------------------------------------------------------------------------- Activity Time Step # S-LCP X-LCP S-LCP X-RCP Notes / Comments 2010/106 (new NOP) RSR1A RSR1B RSR2A RSR2B -------- ----- --------- ----- ----- ----- ----- ------------------------ FGAIN 17:30 60 60 60 60 Set FGAIN Pre-Cal 18:32 1 -5.9 -10.0 -9.0 -9.8 ADC amplitude (dB) 18:35 att auto -9.9 -10.1 -9.9 -9.9 17.5 18.0 16.5 27.0 Attenuator settings (dB) 18:36 2 -9.8 -0.6 -9.8 -0.2 18:38 3 -0.6 -0.5 -9.8 -0.2 18:39 4 -0.6 -0.5 -0.5 -0.2 18:41 att auto -9.7 -10.2 -10.0 -2.8 Reset attenuators 28.5 20.5 28.0 31.5 Attenuator settings (dB) Amb load physical temps S1=15.38 S2=18.50 X1=14.00 Local weather: T=12.7C H=76.3 percent sky=overcast 18:45 rec 3 e Begin 25 kHz recording 18:49 5 -9.7 -10.1 -10.0 -2.8 63 using old procedure 18:55 6 -9.8 -10.0 -9.8 -2.7 19:00 7 -9.7 -21.6 -9.9 -13.1 19:05 8 -9.8 -20.6 -9.7 -14.5 19:10 9 -9.7 -19.6 -9.9 -14.5 Confirmed 12.5K ND 19:15 10 -9.7 -9.8 -9.9 -2.9 Amb load phys temps: S1=16.12 S2=18.25 X1=14.25 19:20 11 -9.8 -10.0 -9.8 -2.8 19:25 12 -9.8 -21.6 -9.9 -14.5 19:30 13 -9.7 -21.5 -9.8 -14.5 19:35 14 -9.7 -21.6 -19.5 -14.5 19:40 15 -9.7 -21.6 -21.3 -14.5 19:45 16 -9.6 -21.6 -21.3 -14.4 19:50 17 -18.9 -21.6 -21.3 -14.5 19:55 18 -20.9 -21.5 -21.2 -14.3 Amb load phys temps: S1=16.44 S2=18.12 X1=14.50 20:00 end -20.9 -21.0 -21.4 -12.5 Stop 25 kHz recording BOT 20:00 -20.9 -21.0 -21.4 -12.5 20:18 X-band AOS 21:50 S-band AOS 21:06 rec 3 e -21.1 -20.9 -21.3 -13.1 Resume 25 kHz recording 21:21 -21.0 -21.3 -21.2 -13.9 DSS 63 switch to Mars pointing predicts 21:23 sfro 3/SX +5K +24K +5K +24K Rcvr tuning offset (Hz) sfro 4/X +25K +25K Rcvr tuning offset (Hz) MiniCal1 21:25 1 -20.9 -21.2 -21.3 -11.4 Confirmed 12.5K ND 21:28 2 -20.9 -21.3 -19.4 -12.0 21:31 3 -20.9 -19.3 -21.3 -12.0 21:34 4 -19.0 -21.2 -21.3 -11.9 21:37 5 -21.0 -21.2 -21.4 -12.1 21:40 end -20.9 -21.2 -21.2 -12.0 BSR 21:43 rec 3 e -20.8 -21.2 -21.3 -12.0 Begin 100 kHz recording 22:01 sfro 3/X +31K +31K Rcvr tuning offset (Hz) 22:10 X-band echo out of BW 22:20 sfro 3/X +38K +38K Rcvr tuning offset (Hz) 22:30 sfro 3/X +45K +45K Rcvr tuning offset (Hz), 10 minutes early 22:48 sfro 3/X +52K +52K Extra tuning offset (Hz) end BSR 23:06 sfro 3/X +24K +24K Rcvr tuning offset (Hz) -20.5 -19.0 -20.9 -10.0 End 100 kHz recording MiniCal2 23:08 1 -20.3 -18.0 -20.9 -9.2 Confirmed 12.5K ND 23:11 2 -20.3 -17.6 -19.1 -9.8 23:14 3 -20.2 -16.5 -20.6 -8.4 23:17 4 -18.6 -17.0 -20.8 -8.1 23:20 5 -20.3 -17.3 -20.8 -8.1 23:23 end -20.4 -17.2 -20.9 -7.9 sfro 3/SX 0 0 0 0 Remove tuning offsets DSS 63 return to s/c pointing predicts EOT 23:46 -20.6 -17.4 -21.1 -8.6 PostCal 23:58 1 -9.7 -10.0 -10.1 -2.7 Amb load phys temps: S1=16.69 S2=17.62 X1=13.81 00:01 2 -9.8 -10.1 -10.0 -2.8 00:04 3 -9.6 -10.0 -9.9 -2.9 00:07 4 -9.7 -9.8 -10.1 -2.9 00:10 5 -9.5 -10.1 -10.0 -2.9 00:13 6 -18.9 -17.0 -21.5 -8.7 00:16 7 -20.9 -16.1 -21.6 -8.7 00:19 8 -21.0 -17.0 -19.4 -8.6 00:22 9 -20.9 -16.9 -21.3 -7.9 00:25 10 -20.9 -17.0 -21.3 -8.6 Amb load phys temps: S1=16.62 S2=17.38 X1=13.75 Local weather: T=9.8C H=96.0 percent sky=overcast EOA 00:28 -21.0 -16.9 -21.3 -8.7 End 25 kHz recording AMB = ambient load BOT = Beginning of Track BW = bandwidth CNR = Carrier to noise ratio CONSCAN = conical scan tracking CW = continuous wave (carrier only) EOA = End of activity EOT = End of Track FRO = frequency offset HGA = high-gain antenna LOS = loss of signal ND = noise diode No = noise power NOP = Network Operations Plan Pc = carrier power rcvr = receiver RFI = radio frequency interference S1 = ambient load for S-RCP S2 = ambient load for S-LCP s/c = spacecraft SL = S-LCP SNR = Signal to noise ratio SR = S-RCP TLM = telemetry X1 = Ambient load for both X-band channels XL = X-LCP XR = X-RCP Post Analysis Summary X-Band Amplitude Calibration: Because strong but time-variable, broad band interference appeared on both X-Band channels at about 22:50, the normal procedure for calibrating these data did not work very well. The normal procedure assumes that the background noise power varies slowly and that a meaningful average value over the length of the surface observations can be obtained. This can be compared with the noise power measured during the pre- cal and post-cal, and a scaled value of the pre-cal/post-cal system temperature can be then applied. Since the noise power increased by a factor of 3-4 during the final 15 minutes of the 83 minute observation (and continued to vary through MINICAL #2 and the post-cal), the calculated powers were suspect. We followed the normal procedure to obtain the system temperatures on each X- Band channel at zenith during the pre-cal. Since the X-RCP noise diode only ran at about 50 percent duty cycle during the pre-cal (see Performance Problems and Notes), the noise diode minicals during tracking were not useful; and we dispensed with the noise diode measurements entirely except as sanity checks. In fact, the noise diodes appear to have functioned normally after the X-RCP duty cycle was corrected; but extracting system temperatures from the MINICAL #2 data was still very difficult because the background level was so volatile. Instead, we used 10-second averages of noise power in 8 3125 Hz windows to monitor the noise power throughout the experiment in our preferred 25 kHz data. The system temperature at any time was assumed to be the pre-cal system temperature times the current noise level divided by the average noise level from the pre-cal. This gave the following Tsys values, where BSRn denotes the nth file of 100 kHz data: X-RCP X-LCP ------------------ ------------------ Activity Times Tsys Times Tsys -------- ----------- ----- ----------- ----- Pre-Cal 18:45-19:15 20.67 19:20-19:36 22.12 BSR1 21:43-22:04 24.39 21:43-22:04 24.94 BSR2 22:05-22:25 24.42 22:04-22:25 24.87 BSR3 22:26-22:46 25.78 22:25-22:46 25.90 BSR4 22:46-23:06 39.55 22:46-23:06 32.34 Post-Cal 24:14-24:22 79.64 24:14-24.16 62.38 X-Band Frequency Calibration: The directly propagating carrier can be seen in 100 kHz X-Band spectra throughout the surface observations. We measured its frequency at 315 second intervals in both polarizations; the true value at each time was assumed to be the average of the two measurements. We considered fitting a quadratic to the average values; but there is a cusp in the data (both polarizations) at about 22:49 that would not have been represented in such an approximation. We calculated the carrier-to-echo differential Doppler at 1-second spacings (program BSRPREDICT), interpolated the carrier frequencies to the same times, and compiled a table of steering coefficients, which was used to adjust the phase of the data samples so that the center echo frequency would appear to be at 50 kHz in a 100 kHz window. We then digitally filtered the samples to isolate the center 25 kHz and used those data to estimate echo power in each polarization. S-Band Amplitude Calibration: The S-Band amplitude calibrations were similar to those at X-Band, though the pre-cal, post-cal, and mini-cal results appeared to be more reliable. We based calibration on the changes in background noise power, assuming that changes in front end gain were negligible. The S-RCP system temperatures are shown below: S-RCP ------------------ Activity Times Tsys Notes -------- ----------- ----- ------------------------------------- Pre-Cal 18:45-20:00 23.34 Too much time between AMB and SKY Minical1 21:25-21:34 23.32 BSR1 21:43-22:04 23.62 BSR2 22:04-22:25 23.60 BSR3 22:25-22:46 23.92 BSR4 22:46-23:06 24.80 Minical2 23:09-23:15 26.78 Post-Cal 23:54-24:04 22.76 Post-Cal 24:07-24:19 22.69 Better than pre-cal; used as reference The S-LCP inverted-J gain changes (see Performance Notes) were reflected in the S-LCP GNC file by using a 15-segment linear piecewise approximation to the voltage over 80208-81144 seconds. The S-LCP system temperatures adopted for calibration at other times are shown below S-LCP ------------------ Activity Times Tsys Notes -------- ----------- ----- ------------------------------------- Pre-Cal 19:36-20:00 23.34 Used as reference Minical1 21:31-21:40 24.78 BSR1 21:43-22:04 25.05 BSR2 22:14-22:15 25.39 BSR3 22:35-22:36 25.89 BSR4 22:46-23:06 26.67 Minical2 23:14-23:23 27.80 Post-Cal 23:52-24:06 24.81 Post-Cal 24:06-24:21 24.81 S-Band Frequency Calibration: Only S-LCP carrier frequency estimates were used in calculating steering coefficients; the S-RCP carrier was weaker and only visible intermittently. Measured carrier bin positions are shown below with linear interpolation among these points. The surface echo was then taken to be fR = (bin - 1)*(100000/1024) - fdd, where fdd is the differential Doppler between carrier and echo predicted by program BSRPREDICT. Time (s) Carrier Position (bin) -------- ---------------------- 78625.5 498.0 80857.5 497.3 81802.5 497.2 82747.5 497.1 Dick Simpson Original: 2010-04-28 Edits to Performance Problems and Notes: 2010-05-05 Added Post Analysis Summary; updated Performance Notes: 2010-05-07 Edited Performance Notes and Post-Analysis Summary: 2010-05-09 Edited Performance Notes, added S-Band calibration text: 2010-05-18 Removed smart characters which would not pass PDS validation: 2011-05-13