MOBB to MARS Design Notes P. McGill * Holes in existing MOBB glass spheres are 7/16" * Voltage range for MOBB instruments Guralp CMG-1T Seismometer: 10 to 36 VDC 2D-ACM Current Meter: 7 to 15 VDC Quanterra Q330 Digitizer: 10 to 18 VDC HESC104 PC/104 Power Supply: 6 to 40 VDC TS-12W PC-104 Power Supply: 10 to 40 VDC TS-BAT3 PC-104 Battery Backup: 8 to 28 VDC * We may need low-voltage cutoff. We can use the Xenotronix DP10 for lead-acid batteries. It will cut the power when the voltage drops below 10.5 VDC. * The cable from the MOBB SIIM to the MOBB experiment is Falmat FMXCAT51806K12. The total roundtrip resistance of the four 18 AWG power wires (two each way) is the same as one wire one way. The resistance of 18 AWG cable is 25.5 Ohms/km or 0.0255 Ohms/m. 20 m x 0.0255 Ohms/m = 0.51 Ohms. The voltage drop at the maximum possible load of 6 Amps is about 3 volts. The voltage drop at the typical load of 0.36 Amps is 0.18 volts. * If the MARS cable delivers 165 Watts over 3.6 km, the voltage at MOBB is 276 VDC. This is the most power that MARS can deliver on this cable. The Vicors will shut down at 250 VDC, so this provides a 10% margin. * The original extension cable was Falmat FM022307-3CP Rev.A which cost $11.80 per meter. This cable did not lay well on the bottom and was replaced by South Bay Cable SB-46974 Rev. A which cost $7.80 per meter. The South Bay cable has two twisted pairs at its core, as opposed to a twisted quad in the Falmat cable. * The MARS 1 pps pulse is 10 ms (20 ms modulo 10), falling edge active. The Guralp and Quanterra want a 100 ms pulse, so the MARS pulse must be stretched. A Fairchild 74HC123A monostable multivibrator has a pulse width of T = RC, so R = 1 M and C = 0.1 uF yields a 0.1 sec pulse. * The accounts on the Vulcan and MOBB-Shore are: User: ncss Pass: vttBots User: root Pass: vttBots1 We should use the ncss account for normal operation as follows: % ssh 134.89.42.110 -l ncss * NTP on MARS is at 10.91.128.55 * The account for access to Quanterra Q330 documentation is: User: laurel Pass: kathmandu * The DPG amplifier board has its 10-pin connector labeled, but not its 8-pin connector, which connects to the sensor. These connectors have pairs of pins in common, so they are actually 5- and 4-pin connectors. With the 4-pin connector on the right, the pins are: top: +In -In -Out bot: +Out * These are the MOBB IP addresses: Device WNS External MARS External MARS Subnet ------ ------------ --------------- --------------- MARS SIIM 134.89.52.112 134.89.42.112 10.1.11.2 MOBB SIIM 134.89.52.113 134.89.42.113 10.1.11.3 Vulcan ETHO 134.89.52.110 134.89.42.110 10.1.11.10 Vulcan ETH1 NA NA 10.1.11.129 Q330 134.89.52.114 134.89.42.114 10.1.11.130 MOBB-Shore NA TBD 10.91.41.2 MOBB-ILOM NA TBD 10.91.41.3 NTP-MARS NA NA 10.91.128.55 NTP MBARI 134.89.12.32 134.89.12.32 NA * Power consumption measured on Wet Node Simulator: 48 V @ ~120 mA = 5.8 W 375 V @ ~50 mA = 18.8 W UPS capacitor bank on MOBB can take as much as 27 W by itself when charging, so overall power consumption will be higher when powered is restored after a power failure. * Important dates 26Feb09 - The extension cable between MOBB and the MARS node is installed. 5Apr09 - The cable is trawled, 39 days after installation, but continues to operate despite some damage. 25Sep09 - Both MBARI and UC Berkeley receive a funding supplement from NSF to replace the damaged cable ($68k to MBARI). 28Sep09 - An ROV dive at the MARS node reveals that the MOBB cable has been trawled a second time, occurring sometime in the previous few weeks. Again it continues to operate despite additional damage. 11Feb10 - At about 6am PST the MARS 1 amp current limit on our 375 VDC tripped, and the ground fault is at a maximum limit reading of 1 ma. 48 VDC is okay, and we can talk to the MARS SIIM. 11Mar10 - The MARS-end SIIM is recovered by cutting the extension cable near the SIIM. Inspection of the over-stressed SIIM connector shows that it is probably not the cause of the system failure. 25Mar10 - A dive to recover the extension cable reveals that the MOBB site was trawled a third time on 11Feb10, severing the cable at the MOBB-end SIIM, damaging the electronics module, and destroying the current meter. All equipment was recovered except for the buried seismometer, the trawl-resistant bottom mount, and the bare extension cable as its end could not be found. 15 & 16Nov10 - The cable-trenching tool for the cable-burial toolsled is tested on Ventana. It was able to bury the cable 9 to 12 inches deep in sandy sediment. 13Dec10 - The severed 3.6 km extension cable is recovered on a Ventana clean-up dive. Of all the MOBB equipment, only the buried seismometer and the trawl-resistant bottom mount remain deployed. 16Jun11 - The cable-trenching tool is tested on Ventana a second (third?) time with the following improvements: a thinner and sharper knife blade; a tensometer; six wheels to improve vehicle stability. It performs well and is ready for use. 23Jun11 - The extension cable is redeployed along with a refurbished electronics module and current meter. * Locations MOBB - N 36 41.439, W 122 09.962 * Sensitivity of recorded channels as listed in the RESP file at NCEDC LDD - Differential Pressure Gauge is 419 counts per Pascal. Typical activity of +/- 7000 counts is 7000 / 419 = 16.7 Pa, or 0.000167 atm, or 1.67 mm of water. This seems too low by three orders of magnitude. BHE +3.69026E+09 counts per meter per second at +1.00000E+00 Hz BHN +3.59330E+09 counts per meter per second at +1.00000E+00 Hz BHZ +3.59330E+09 counts per meter per second at +1.00000E+00 Hz The RESP format is documented at http://www.iris.edu/dms/nodes/dmc/data/formats/resp/ To convert velocity to displacement, multiply the gains by 2 pi times the calibration frequency, 1 Hz in this case. I think this will give you counts per meter of displacement for sinusoidal motion at the calibration frequency. The MOBB data don't show zero counts for zero velocity, and actually sit at about -5.8E+05 counts at rest. I see "sudden" shifts, taking 3 to 10 minutes, with the velocity increasing linearly (i.e. constant acceleration) and reaching a max velocity of about 1E+04 counts. This is a max velocity of 1E+04 counts+ / 3.69026E+09 counts per meter per second = 2.7E-06 meters/sec, or 2.7 microns/sec If this takes place over, say, 3 minutes, then the constant acceleration is 2.7E-06 meters/sec / 180 sec = 1.5E-08 meters/sec/sec and the total displacement is 0.5 x a x t^2 = 0.5 x 1.5E-08 meters/sec/sec x 180^2 = 2.4E-04 meters or 0.24 mm