CONT02 Station Tests Brian Corey, Ed Himwich, Dan Smythe, Richard Strand 4 July 2002 UT Introduction ============ In preparation for the October CONT02 observations, we request that the staff at each of the eight stations involved conduct a series of tests designed to uncover previously unrecognized problems that could adversely affect the CONT02 results. These tests, which are described below, are similar to those used before previous CONT experiments. In conjunction with these tests, we will also be examining the correlator output from some experiments, probably RD0207, R1026 and R1030. (Onsala will be requested to tag along to R1030 for 3 hours to R1030 since it cannot participate in RD0207.) If any problems are detected in this experiments or the station tests, we hope to have them fixed by early September so that the data from a follow-up experiment, probably GRAV01, can be used as a spot check on the stations' performance. (Onsala will be asked to tag along for three hours to this experiment as well.) We will be looking for problems in several areas, including fringe and phase cal stability, narrow track calibration and parity error rates, and overall system sensitivity. Should we find any problems, we will contact the affected stations and suggest further tests or remedial action to be taken. Please carry out the tests and report your results by August 15. That will leave us a few weeks before the follow-up experiment to fix any problems the tests may reveal. In addition, please inform us of *any* problems you have been experiencing with the hardware, software, communications, logistics, etc., that could have an impact on the CONT02 observations. CONT02 will require continuous operation for extended period of time, so any failure that cannot be addressed quickly during the observing could lead to significant loss of data. We are better off trying to fix things beforehand rather than during the continuous recording period. If (for example) your tape recorder has been having problems lately, or your Field System hangs up for no good reason, or a VC LO drops out of lock occasionally, or the schedules don't leave enough time to get on source, or there are any other problems, please let us know now! Logistics ========= In performing these tests, please record all measurement values either in Field System logs or on paper for later transcription to email. For information you record in logs, please include comments in the logs about what test is being recorded. Liberal use of comments to explain the various readings is very helpful. We would like to conduct these tests using the CONT02 set-up, which is the same as the IVS-RD and IVS-R1 set-ups, but that is impractical because of limitations on how we can extract phase-cal at the stations. So instead we have prepared a special set-up. We have provided special test procedure files on cddisa under the vlbigeo account under /misc. Please use the library appropriate for your system; the files are: contm4.prc (Mark IV systems), contgc.prc (Gilcreek VLBA system), contkk.prc (Kokee VLBA system), and contap.prc (Algonquin VLBA4 system). We have tested the procedures as much as we can reasonably and expect that they will work. However, if you encounter any problems please let us know and we'll fix them. Please place FS logs from the tests on cddisa under /misc in the vlbigeo account with appropriate names, we suggest using the format "testnxx", when "n" is the one or two digit number of the test and "xx" is your two character station code. Send e-mail messages to weh@vega.gsfc.nasa.gov (Ed Himwich) and bcorey@haystack.mit.edu (Brian Corey) to inform us as the logs become available. Send e-mail copies of measurements that are recorded on paper to the same addresses. Please don't wait until all the tests are complete before sending in the data. It would be better to send the results as they become available. If you discover an error or our cryptic instructions lead you astray, you can always re-submit new data. Please contact us directly if you have any questions about the tests. Tests ===== Please be sure to use the "testa=1" set-up for tests 1-6. The tape drive must be turned on and a tape mounted for tests 1-6. For each test you record in a FS log, please execute the set-up procedure after you open the log for the test. This will allow us to verify that it executed correctly in case there is any odd behavior. 1. Phase cal total power. With the "testa=1" setup, measure the fractional total power contributed by the phase cal, for comparison with the 10-kHz amplitude measured at the correlator. Do this test with the antenna pointed near the zenith and under good weather conditions (no rain). The easiest method probably is to measure the system temperature CALTSYS (SXCTS for FS version 9.5.7 or less, hereafter only CALTSYS will be mentioned) with phase cal on, and then with phase cal off. You may need to repeat the measurements several times, say 10, interleaving the phase cal on and off measurements, in order to beat down the noise. A second method is to pick one VC/BBC near the center of each band (S or X), and repeatedly measure its USB video power level (using the Field System or a voltmeter) with phase cal alternately on and off. Also record the power level with no IF signal into the VC/BBC (TPZERO). 2. Spurious signal tests. (You *knew* you'd have to do this, right?) With the "testa=1" set-up, for each VC/BBC, measure the power level of the 10-kHz tone in the USB video output under 3 conditions: A. Phase cal on, receiver LO locked (normal) B. Phase cal off, receiver LO locked (Use switch on delay cal ground unit to turn off phase cal.) C. Phase cal on, receiver LO unlocked Use a low frequency spectrum analyzer set with a spanned bandwidth small enough, and/or an averaging time long enough, to make the noise floor at least 40 dB below the phase cal. Levels measured for 2B and 2C should be >40 dB below those for 2A. Please note in the log whether the levels in 2B or 2C appear to be consistent with the general noise level or whether they are actual signals that stick up above the surrounding noise. If you don't have a low frequency spectrum analyzer, you can make a less sensitive, but still useful, measurement using the DQA or Mark IV Decoder. Please open a new log file before performing the test. There is a procedure in the test procedure library that will measure the phase-cal in the USB channels, "usbpc". You can use this procedure for each of the three states mentioned above. Please be sure to put comments in the log so that we will which state is which. 3. Video converter image rejection. With the "testa=1" set-up and with phase cal on and receiver LO locked again, record the level of the 10-kHz tone in the USB and LSB video outputs from each VC/BBC. The USB/LSB power ratio should be >20 dB. Again you may use the DQA or Mark IV Decoder if you don't have a low frequency spectrum analyzer. Please open a new log file before performing the test. You should use the "usbpc" and lsbpc" procedures in the test procedure library. 4. Orientation dependent effects. These tests attempt to find problems that are dependent on the antenna position. Please open a new log file before performing the test. With the "testa=1" set-up, the antenna should be moved over its entire range of motion. We recommend that you move the antenna along its extremes of motion for this test, e.g., for an AZ-EL antenna, you would place the antenna at its minimum elevation, and scan the entire range of azimuth (including any cable overlap), then at one azimuth extreme, scan to the maximum elevation, then at the maximum elevation scan back around in azimuth completely, then at the other azimuth extreme scan in elevation. If time allows you are encouraged to make a few more scans in azimuth and elevation to fill in the coverage. Please enter a comment at the start and stop of each scan to tell us where the antenna is. More frequent comments are welcome. You should monitor the cable cal, and using an oscilloscope, the phase cal in one channel for each band. You may need to make more than one pass to check all the items carefully. While the antenna is being moved, please run the procedure "moving" from the test procedure library every 15 seconds with: "moving@!,15s" in a new log. If your station has the "track" command implemented, please edit the "moving" procedure and uncomment this command. In order to get at least one sample every 20 degrees please limit the antenna speed during the scans to less than 80 degrees/minute if possible. Please check for the following: A. Are there jumps or rapid drifts in the cable cal? B. Are there jumps or rapid drifts in the phase cal for either band? 5. Diurnal behavior. Please open a new log file before performing the test. Use the "testa=1" set-up with the antenna parked at zenith. For 24 hours, use the "overnite" procedure from the test library or a procedure of your own devising to record the following quantities every 5 minutes ("overnite@!,5m"): Phase cal phases and amplitudes for two X-band channels (use Decoder or DQA) Phase cal phases and amplitudes for two S-band channels (use Decoder or DQA) WX Cable cal All receiver monitor points System temperatures in all channels (CALTSYS) Please add comments describing the general weather conditions during the 24 hours. Please modify the provided "overnite" procedure as necessary for your station. In particular you will need to add any commands to monitor the receiver ("rxall" for CDP style receivers) or other useful monitor points. You might also need to change the name of the system temperature procedure in "overnite" from CALTSYS to SXCTS. After placing the logfile on cddisa, if possible please delog the data locally. The sooner a problem is found, the sooner it can be fixed. Please check the following: A. Are the variations in phase, cable cal, and temperature consistent with each other? B. Does the phase cal amplitude stay in the 3-10% voltage range? (for the Mark IV decoder, please divide the raw sampled amplitude by 10 to get the voltage in percent, the DQA command reports voltage in percent as is) C. Are the temperature changes of the receiver box, LO, and phase cal unit within acceptable bounds (5, 5, and 2 deg C, respectively)? Any temperature variations should be gradual, in step with the outdoor temperature, with no jumps. D. Are the weather data free of jumps and unreasonable behavior. 6. Short period variations: A. Please open a new log file before performing the test. Use the "testa=1" set-up with the antenna at zenith. For 10 minutes, record every 15 seconds (or as frequently as possible if 15 seconds cannot be achieved): Phase cal phases and amplitudes for two X-band channels (use Decoder or DQA) Phase cal phases and amplitudes for two S-band channels (use Decoder or DQA) Cable cal. The test procedure "sample" can be used for this. B. Examine the results for any systematic behavior. With the antenna not moving these values should be almost constant. If some systematic behavior is observed, you may want to try to record other data, such as the LO and phase-cal temperatures to locate a correlation. C. Record some general comments about the weather conditions. 7. Cable cal: A. Is the counter used in single sample mode, i.e., no averaging? B. Is the reading stable at the 1 microsecond level on time scales of 1-30 seconds? C. When a test cable is inserted, how does the change in reading compare with the calibrated length of the cable? D. Does the counter reading change between forward and reverse recording? 8. Power Supplies. For the power supplies in the following pieces of equipment, please verify that the ripple is less than 10 mV peak-to-peak and the noise is less than 100 mV peak-to-peak. The main point of this test to identify any obviously sick power supplies. If you have time you should check over several sweep rates. Two good choices are 10 milliseconds per division triggering off the line frequency and 0.1 milliseconds per division auto-triggering. If you have time you can check at other rates while auto-triggering. (Please be aware that in some VLBA racks the black and red test points are reversed on the negative voltage supplies.) A. Rack B. Tape drive C. RX 9. Met sensors. A. Is the barometer stable at the 0.1 mb level, with no jumps over a few minutes? B. Is the temperature stable at the 0.1 degree C level, with no jumps over a few minutes? C. Record the barometer offset relative to a local authority (e.g., airport). 10. Pointing and SEFD A. Verify that pointing errors are typically less than 0.1 FWHM beam at X-band. B. Measure the SEFD in each channel for a strong source. This can be done by using the ONOFF command if the pointing programs have been installed at your station. Otherwise measure system temperatures (CALTSYS) on and off source. Repeat at least once and check for consistency of results. 11. Tape recorder. Please perform checks #1,2,6,7, and 8 from the Recorder Check memo, Mark IV Memo 267 at ftp://web.haystack.mit.edu/pub/mark4/DAS/reccheck.text or ftp://web.haystack.mit.edu/pub/mark4/memos/267.pdf. The text describing these checks are included below with "11." prepended to the numbers. The text has been modified slightly from the original to make it more relevant to the CONT02 set-up. 11.1. Forward-reverse shift test. Measure the tape position by issuing a pass command after peaking on a recorded track while moving the tape in the forward direction. Repeat in the reverse direction. The difference between these two readings must be less than 50 micrometers. Occasionally (approximately monthly, or after calibrating the head positioner), repeat this test at the beginning, middle, and end of the tape. 11.2. Speed shift test. Use peak and pass to measure the shift in tape position between normal (80) and double (160) speed. This shift should be less than 20 micrometers. Check in both directions. 11.6. Read stack calibration test. Record 1000 feet with heads 17, 18, and 19 with tapeformc pass=13,13. (If you don't have "tapeformc", please use "stack=0,0,f,f") Reproduce with head 18, move the read stack with stack=,698.5,,f and stack=,-698.5,,f to find the tracks near +/-700 um. Use peak and stack to verify that these tracks are at +/-698.5 +/-10 um. 11.7. Write stack calibration test (for Mark IV and VLBA4 only, skip for VLBA). Record 1000 feet with head 18 enabled at stack=0,,f. Back up the tape 1000 feet and repeat with stack=-700,,f and with stack=700,,f. Use stack=,+/-700,,f, peak, and stack to verify that these tracks are at +/-700 +/-10 um. 11.8. Record margin check (write voltage). Record 1000 feet on a forward pass with the "4f8m12a=1" set-up from the test procedure, use 160 ips record speed. Perform a parity check using your local version of "check80f". Repeat the test on fresh sections of tape, reducing the write voltage in 1 volt steps until the error rates reach a minimum and then start to rise again. Place the write voltage at the value just above where the rates starts to increase. This test will be difficult if you do not have good reproduce. Please record your best error rates as part of the test. if you cannot get reproduce, please contact one of us so we can put you in touch with a recorder expert to resolve the problem. 12. Correlator tests. The items in these check will be performed at Haystack and GSFC. The stations don't need to take any action for these items. They are included here for completeness. 12.1. Correlation A. What problems are occurring with correlating data from these station? B. What special operational adjustments are required for these stations (e.g. deleting specific channels, using certain tape drives) ? 12.2. FRNGE Plots (both tape directions and both frequencies) A. Do phases line-up channel to channel (after offset adjustment)? B. Are there phase-cal amplitude variations, channel-to-channel? C. Are there fringe amplitude variations channel-to-channel? D. Are there time dependent variations in the phase-cal phase, amplitude, fringe phase or amplitude? In particular are there high frequency variations within a scan? E. Are parity errors nominal for all tracks F. Are any tracks missing significant amounts of data G. Are phase-cal amplitudes nominal H. Are fringe amplitudes nominal 12.3. Raw Data from Database A. Is phase-cal amplitude versus phase flat? Is there evidence of spurious signals, reflections, or ground pick-up? B. Do cable data, pcal remove, and temperature show consistent behavior? C. Items in (B) should be reasonable behaved, in particular there should be no jumps D. Are the inferred SEFDs reasonable? E. Fringe amplitude varies as expected by baseline and source, but not over the sky. F. Phase-cal amplitude does not vary over the sky. G. Are there fringe amplitude variations channel-to-channel? H. What is the average and RMS fringe phase misalignment by channel over the experiment. I. Is the data yield per channel over the experiment uneven or low?