EPOS-N - European Plate Observing System
Seismometers / vault construction
Stations: Upgrade/install notes
Ideas for improvements
Installation / Inventory
Last update: February 17, 2020, at 01:26 PM
From page 8 in Amphenol datasheet (click to open):
NOTE: Front face of pin inserts illustrated.
According to advice in Trillium 120P manual, the general interconnection between sensor and digitizer should be like this:
In order to determine optimum earthing/shielding method, we performed a series of measurements. The aim was to determine the configuration (if any) that provided the lowest 50 Hz interference from mains power supply.
With regard to cross coupling between the three seismic channels X, Y and Z, we adhere to advice of sensor manufacturer by terminating the shields of the three twisted, shielded signal pairs at one point only - on the digitizer side.
In addition, it is useful to establish a standard for use of Güralp "Scream!" spectrogram software which can be used during sensor/digitizer installation, in order to ensure that seismic signals have a reasonable shape, and that mains 50 Hz interference is at lowest possible level.
Sensor was placed on pier in "Seismo-lab", basement of "Realfagbygget".
Measurement setup, sensor in foreground, placed on pier in "Seismo-lab".
Based on measurements outlined above, we will use this interconnection between Trillium 120PA sensor, Güralp CMG-D24 digitizer and power supply. Note that only the three seismic components, power supply and grounding/shielding is shown - see complete wiring diagram and -list below.
Interconnection of Nanometrics Trillium 120PA, Güralp CMG-D24 digitizer and power supply.
Click to see details of rightmost signal cables:
Link to Excel part list with hyperlinks to spesific part:
View of part list-link
It's not very tempting to perform Trillium 120PA recentering using laptop in adverse weather conditions, if the sensor enclosure is exposed to the elements. Trials with a RS-232/Bluetooth adapter were successful - this means you can have remote terminal access to the 120PA from approx. 100 meter distance (the two Bluetooth devices are both of Class 1 category = long distance).
LM Technologies Bluetooth - RS232 adapter, mod. LM-058.
The adapter must first be configured by connecting to it's serial port from a PC, and using a terminal program (e.g. like TeraTerm).
Important: During software setup, place the DCE/DTE switch to the DCE position (towards the DB9-connector).
To obtain longest range, use a Class 1 USB/Bluetooth adapter. During testing, we used an older adapter manufactured by Billionton (P/N: UBTCR3C1A-B) which is obsolete now; any current device (e.g. this SDM USB Bluetooth 2.1 Adapter Class 1) should work fine.
After installing the adapter, scan for Bluetooth devices in the neighborhood, and select "Trillium 120PA".
The adapter on the Trillium side will automatically announce it's serial port capability, and you will be notified on the specific COM port assigned during the Bluetooth session.
Note that connection could be a two-stage process: First the "bridge" between the two Bluetooth devices is established, and then you have to actually make the connection.
The "SENSOR A" connector on the Güralp CMG-DM24 digitizer is furnished with 12 V output, intended to power the manufacturer's own brand of sensors. In our case, using sensor and digitizer from different manufacturers, we must ensure this power supply meets the sensor requirements.
The CMG-DM24 User's Manual does not provide any specifications regarding the 12V power output on the SENSOR A connector. Instead, clues must be sought after in the specifications of those sensors that are intended to be used with the digitizer.
The CMG-3T compact three-component broadband sensor is clearly a candidate for DM-24 partnership, as sensor output and digitizer input matches. This sensor has worst case current consumption during locking and unlocking; value is 490 mA. However, a footnote states that
"Because centering, locking, and unlocking consume varying amounts of power, it is recommended that you use a power supply capable of delivering 1 A at 12 V."
From this we must conclude that the voltage output of the DM-24 sensor connector is capable of supplying at least 1 A at 12 V - assuming, of course, that voltage supply feeding the digitizer itself can provide this amount of current - and then some for the digitizer itself, worst case.
The Trillium 120PA provides three "State-of-health" signals - U_MP, V_MP, W_MP - that indicates how well centered the three masses are. Maximum values are +/- 4.0 volt. The voltages reflect the amount of feedback coil DC current needed to keep the masses in center positions. If any voltage approaches 3.5 V, mass centering should be performed.
Principles of Velocity Broad Band seismometers are described in chapter 5 of New Manual of Seismological Observatory Practice (2002), revised version, electronically published 2009
The force-balance principle will in effect make the displacement transducer output ground acceleration. In fig. 5.17, the stage labelled "Displacement transducer with integrator" outputs Velocity - which is used as the seismic signal after transformation to X,Y,Z reference frame. Yet another integration yields Position - which reflect the amount of DC current applied to the Force Transducer in order to keep mass in centered position.
From this figure, one should conclude that the Velocity signal should NOT contain any offset that reflects the mentioned DC current in the Force Transducer. Of course, when the mass imbalance is so great that the DC current is unable to keep the mass centered, the Velocity signal will be influenced.
This argument is confirmed by observation - we can have large signals on the Mass Position output, and no offsets on the Velocity signals (keeping in mind the transformation needed to compare the two signal groups).
To summarize: We should not be able to see mass imbalance on the Velocity signals - until the imbalance is very great.
An electronic circuit placed in the junction box could automatically activate the Mass Centering line on power-up.
However, there are some problems with automatic centering when power is applied to the seismometer:
Mass centering can either be controlled by the digitizer web interface, or through the command line interface.
The command line provides both a high and low level interface. Ref page 227+ of the MAN-EAM-0003.pdf Issue E - Feb 2014 manual.
One question remains: How long is the duration of the MASS CENTERING command signal sent from the digitizer - and does it satisfy the requirement of the Trillium 120PA sensor? This must be determined by experiment - it's not stated in the digitizer documentation.
If the MASS CENTERING signal duration is not compatible with sensor requirement we can consider using the MASS UNLOCK signal instead, which is subject to direct on/off control.
The Trillium 120PA sensor outputs single-ended UVW MASS POSITION signals in the +/- 4.5V range.
Digitizer specifications are given on p. 120 in the MAN-D24-0004.pdf manual. The MASS POSITION inputs are, however, not mentioned explicitly; one must assume they are categorized as MUX CHANNELS since they are single ended and not differential, thus they have +/- 10V input range, which accommodates the sensor MASS POSITION signal output range.
Trillium / DM24 interface - calibration test. Click to enlarge.
Open DM24 web interface and examine mass position status:
Mass position status before mass centering.
Log in and click "Instrument Control" menu. You should see this:
DM24 web interface - instrument control page.
Allow some time for the system to complete mass centering - let's say 5 minutes - and check mass position status again. Notice the improvement.
Mass position status after centering.
DM24 web "system status" shows mass positions as e.g. "-0.9% -- status 99%" (if good) and "-19.7% -- status 80%" (if not so good). However, we need to "calibrate" what these figures really mean in terms of voltage, because our authoritative guide with respect to allowed mass position ranges is the Trillium 120PA "User Manual" [Nanometrics P/N: 15149R6, release date 2009-05-06], and there everything is specified in terms of voltage levels. Quoting:
: "Trillium 120P/PA seismometers have three analog voltage outputs representing the DC currents applied to each of the three channel feedback coils. These are the mass position outputs, which cover a range of ±4 V."
: "Choosing When to Initiate Mass Centering: When the voltage of any of the three axis mass positions (analog signals U_MP, V_MP, and W_MP, referenced to AGND) approach ±3.5 V."
: "As one or more of the masses in a Trillium 120P/PA seismometer approach the range limit of ±4 V, the seismometer is subject to some degradation in performance. Low frequency self-noise increases by approximately 2 dB per V of mass position. Below ±1 V, the effect on noise is negligible. However, as the mass positions approach ±4 V, the noise level goes up to –173 dB m2/s4/Hz at 100 s period. Depending on the site and the type of study being conducted, this increase in the noise level may or may not be an issue."