Overview of seismic instruments presented at the WORKSHOP

Overview of seismic instruments presented at the WORKSHOP High Quality Seismic Stations and Networks for Small Budgets Volcan, Panama 8 -13. March 2004 by Jens Havskov, Department of Geoscience University of Bergen Norway

Before: -Seismographs were specially made -Few standard components were used - Very specialized software Now: -Stations and networks are mainly made with standard industrial components -Digital technology used throughout -More standardized software -Sensors currently the most specialized element -Now possible to build a seismic station with mainly off the shelf products

SENSORS -Trend is to use more broad band sensors (BB), even when overkill, however BB sensors now have a similar price as 1 Hz sensors - 1 Hz sensors will go out except when used with feedback technique -4. 5 Hz geophones the cheapest sensor, now used by several, either directly or with a feedback technique - FBA based sensors will probably dominate the market in the future Typical geophone

Spring Force coil Displacement transducer Mass R Volt out ~ Acceleration C Simplified principle behind Force Balanced Accelerometer. The displacement transducer normally uses a capacitor C, whose capacitance varies with the displacement of the mass. A current, proportional to the displacement transducer output, will force the mass to remain stationary relative to the frame. The FBA can have the digitizer integrated in feedback loop

13 cm The Kinemetrics 3 -component Episensor, an FBA accelerometer

Kinemetrics Episensor internals

Left: The internals of the Güralp CMG-3 T BB sensor. Right: Sensor with digitizer. Photo’s supplied by Nathan Pearce, Güralp.

-------- 2 mm ------Principal elements of a MEMS (micro electro mechanical systems) accelerometer with capacitive transducer. The mass is the upper mobile capacitor plate which can rotate around the torsion bars. The displacement, proportional to acceleration, is sensed with the variance in the capacitance. For high sensitive applications, a feedback circuit is added which controls a restoring electrostatic force, thus we have a FBA. The size of the sensor above is about 2 mm. Figure from www. silicondesigns. com/tech. html.

Signals band pass filtered with different filter widths. The signals have been corrected for instrument response to show displacement. The maximum amplitude in nm is shown to the right on top of the traces.

Noise curves in a rural environment. The 3 dotted lines correspond to the maximum, mean and minimum levels published by Brune and Oliver (1959), the dashed lines give two extreme examples observed in the US and the full line curves give the limits of fluctuation of seismic noise at a European station on bedrock in a populated area 15 km away from heavy traffic (figure from Wilmore, 1979).

Different sensors at Univeristy of Bergen vault

Raw traces for different sensors A small window of the common traces for Z-channels. The numbers above the traces to the right are max amplitude in counts and the numbers to the left, the DC offset in counts.

Displacement 1 -20 Hz A small window of the common traces for the Z-channels. The traces have been corrected for instrument response and show displacement in the frequency band 1 -20 Hz. The numbers above the traces to the right are max amplitude in nm and the numbers to the left, the DC offset in nm. Notice that the last 3 traces are not from the same time window.

Displacement 0. 2 -1. 0 Hz A small window of the common traces for the Z-channels. The traces have been corrected for instrument response and show displacement in the frequency band 0. 2 -1. 0 Hz. The numbers above the traces to the right are max amplitude in nm and the numbers to the left, the DC offset in nm

Acceleration and displacement. The seismograms in the figure show the first few seconds of a P-wave of a small earthquake. On the site there is also an accelerometer installed (A) next to the seismometer (S). The top traces show the original records in counts. The signal from the seismometer is similar to the accelerometer signal, but with higher frequency contents for this later, and the amplitudes are different. The middle traces show the two signals converted to accelerations and the bottom traces, converted to displacement (frequency band 1 -20 Hz). The signals are now very similar and of the same amplitude.

Instrument sensitivity of several Geotech seismometers ranging from the short period S-13 to very broad band 54000 seismometer. The curves show input acceleration equivalent to sensor internal noise in d. B relative to 1(ms-2)2/Hz. NLNM is the Peterson New Low Noise Model (Peterson, 1993). Slightly modified from Geotech home page, www. geoinst. com.

Predicted total noise equivalent acceleration of a standard electronic circuit with a 4. 5 Hz geophone. The contribution from each element is shown and the Peterson noise curves are shown for reference. Thermal noise is the noise due to Brownian thermal motion of the mass, Johnson noise is caused by thermal fluctuation of the electrons within any dissipative element in the electronic circuit, voltage and current noises are generated within the amplifier. Figure from Barzilai et al (1998).

The equivalent ground acceleration noise spectrum for a digitizer with its input shorted, using a low-sensitivity virtual sensor of 4. 5 Hz and G=30 V/(m/s). The record of the digitizer noise has been reduced to the equivalent ground motion using its response combined with an imaginary low-sensitivity sensor. The spectrum then shows the worst-case sensitivity for ground motion that can be achieved with this digitizer and sensor. Of course, a more sensitive sensor would give a lower equivalent ground noise. The smooth curves on top and bottom are the Peterson New Low-Noise and New High-Noise models, for reference.

Sensor name C f 0 Dam Mas Rg CDR Rc K Dyn Mov G Geodevice JC-V 103 1 1. 0 0. 01 1000 100 5. 0 mm Geodevice JC-V 104 3 1. 0 0. 7 350 70 Geodevice JC-45 -3 3 4. 5 0. 02 300 yes 2. 0 mm Geo Space GS 1 1 1. 0 0. 54 0. 70 4550 280 Geo Space GS-11 D (1) Geo Space HS 1 1 4. 5 0. 34 0. 023 380 32 1. 8 mm 1 4. 5 0. 28 0. 028 1295 45 Geotech S-13 1 1. 0 ~0. 02 5. 0 3600 6300 23 629 0. 198 164 3. 0 mm Geotech S-13 J 1 1. 0 0. 9 6400 20 344 140 1. 5 mm Input/Output SM-6 1 4. 5 0. 26 0. 016 375 --- 28 --- 4. 0 mm Kinemetrics Ranger SS 1 Mark products L 4 C 1 1. 0 0. 07 1. 45 5000 6530 100 345 0. 40 1 1. 0 0. 28 1. 0 5500 8905 yes 276 6. 2 mm Mark products L 4 A 1 2. 0 0. 28 0. 5 5500 8905 276 Mark products L 22 1 2. 0 88 Mark products L 28 B 1 4. 5 0. 48 0. 02 395 35 2. 0 mm Sprengnether S 6000 3 2. 0 0. 5 280 45 0. 44

NEGATIVE FEED BACK Name Geodevice FSS-3 B Geo. SIG VE 53 Geotech KS-10 Kinemetrics WR – 1 Lennartz LE-3 Dlite Lennartz LE-3 D/5 s Lennartz LE-3 D/20 s ACCELEROMETERS Akashi V 450 Eentec EA-140 Geodevice BBAS-2 Geo. SIG AC 63 Geo. SIG AC 23 Geotech PA-22 Güralp CMG-5 Input/Output SF 3000 Kinemetrics FBA-23 Kinemetrics Epi. Sensor Sprengnether FBX 23 Sprengnether FBX 26 VELOCITY BB Eentec P-123 Eentec R 1 rotational (1) Eentec EP-105 Eentec EP-300 Eentec PMD 223 Eentec PMD 103 Geodevice FBS-3 Geodevice MBS-1 Geodevice BBS-1 Geotech KS-2000 Geotech KS-54000 IRIS Güralp CMG-1 T Güralp CMG-3 T (2) Güralp CMG 3 -ESP Güralp CMG-6 T Güralp CMG-40 T Nanometrics Trillium Sprengnether S-3000 Q, Sprengnether WB 2023 Sprengnether WB 2123 Streckeisen STS-2 Streckeisen STS-1 (3) C f-range Out In W G Wt Dyn. Resolution 3 3 1 1 3 3 3 1. 0 -40 1. 0 -50 0. 05 -20 1. 0 -80 0. 2 -40 0. 05 -40 8 10 2. 5 5. 0 12 12 ± 12 12 0. 6 0. 5 0. 3 0. 1 0. 6 800 1000 500 160 400 1000 12 2. 5 3 5 2 7 7 120 140 125 120 120 3 nm/s, 1 Hz 1 nm/s, 1 Hz 2 nm/s, 1 Hz 1 3 3 3 DC-6 Hz DC-100 DCDC-100 0. 1 -50 DC-100 DC-50 DC-200 DC-50 6. 6 5. 0 10 10 4. 5 5. 0 3. 6 2. 5 10 10 10 ± 12 12 ± 15 12 ± 12 0. 4 0. 8 0. 5 1. 2 0. 4 0. 2 50. 2 5. 0 2. 5 5. 0 10. 0 2. 25 5. 0 1. 2 2. 5 10 10 8. 6 2 3 2. 5 5 5 0. 5 2 1 1 140 135 120 102 114 155 120 135 155 90 135 0. 005 μg 2 μg 10 μg 0. 3 μg 11 μg 0. 4 μg 3 3 3 3 3 3 1 0. 1 -50 0. 05 -20 0. 03 -50 0. 017 -32 0. 033 -50 0. 05 -40 0. 017 -50 0. 008 -50 0. 01 -50 0. 003 -50 0. 003 -100 0. 1 -50 0. 03 -100 0. 03 -50 0. 033 -30 1. 0 -250 0. 03 -20 0. 016 -50 0. 0330. 003 -10 10 5 10 7. 5 10 10 10 20 10 10 10 8 36 36 20 12 12 12 24 12 12 12 ± 12 12 23 12 0. 2 0. 4 0. 3 0. 5 0. 6 2. 0 1. 2 0. 7 2. 9 0. 8 0. 6 0. 4 0. 1 0. 2 1. 8 2000 50 2000 1000 2400 1500 2000 1500 3200 1500 278 1000 1500 5 1 5 9. 5 11 5. 3 12 12 14 7 66 14 12 9 3 7 11 2 5 5 130 106 135 150 146 132 120 140 160 180 170 145 (4)

Main units of a seismic recorder. There are no flow arrows between the units since all can have 2 way communication. The GPS can be connected to the digitizer or the recorder. The power supply may be common for all elements or each may have its own regulator, but usually the power source is unique (e. g. a battery).

12 bit: ± 2048 counts 16 bit: ± 32768 counts 24 bit: ± 8388608 counts The analog to digital conversion process. The arrows show the location and values (amplitudes) of the samples and the signal is thus approximated with a sequence of numbers available at time intervals Δt.

AD 7710 Crystal Rate Hz 25 ADC number of Bits 20. 0 Dynamic range (d. B) 120 ADC number of bits Dynamic range (d. B) 31 22. 1 133 50 19. 5 117 62 21. 9 132 100 18. 5 111 125 21. 6 130 250 15. 5 93 21. 1 127 500 13. 0 78 20. 8 125 1000 10. 5 63 20. 1 121 Effective resolution of 2 different analog to digiial convers as a function of sample rate. F is the output data rate (samples per second), AD 7710 is a chip from Analog Devices and Crystal is the Crystal chip set.

Unfiltered and filtered record of seismic background noise in a residential area in Western Norway on a hard rock site. The recording is made with a 4. 5 Hz geophone and a 16 -bit ADC at a sample rate of 50 Hz (Geo. SIG GBV 116). The filter is an 8 -pole Butterworth filter with zero phase shift.

A 5 Hz signal is digitized at a rate of 2 Hz. The digitization points are indicated with black dots. Depending on where the samples are taken in time, the output signal can either be interpreted as a straight line (points in middle) or a 1 Hz sine wave.

A FIR filter compared to Butterworth filters. The FIR filter has a gain of 150 d. B at the 50 Hz Nyquist frequency, while the 10 and 25 Hz Butterworth filters have gains of – 42 and -112 d. B respectively.

Seismic spectrum as a function of magnitude

Manufacturer and/or name Earth Data PS 6 -24 Geodevice EDAS-24 L Geotech DR 24 N C H 3 -6 Dy n d. B 140 Input V 8 3 -6 Sen -siti vity 1. 0 , μV 0. 5 G factor 110 Cro ss talk , 120 d. B 110 Alias DB 120 R M oh m 1. 0 Samp le rate, 1 Hz 1000 1 -500 Po we r 1. 5 W 1. 0 Out put 232 485 232 4. 0 1. 4 232 GPS 140 101000 50, 100 1 -200 Drift ppm GPS 0. 5 130 10 135 1 -6 4. 8 129 20 80 130 4. 5 1256 Geotech 49. 65 16 bit Güralp DM 24 1 -3 140 90 (84) 3 -6 3. 3 129 10 1 140 1. 0 1 -200 1. 5 3. 0 1 -200 1. 2 1 -200 1 -250 < 1 10 10 ~ 130 20 110 130 1 1 -500 < 2 130 14 1 80 140 1. 0 4. 4 0. 5 130 1 3 0. 7 142 10 130 100 1. 5 3 -6 2. 2 130 80 140 2. 0 4 305 90 10 0. 48 132 1 120 0. 4 232 3 0. 8 130 2. 5 1 140 0. 5 232 3 9. 5 108 2. 5 120 0. 4 8 10 120 10 140 1. 1 232 422 lpt 3 0. 7 142 10 116 1 -2 14 (3) 0. 0 6 1. 0 140 130 20500 25500 20200 11000 0. 01200 10200 25200 1 e-5 5000 20200 232 422 232 T 232 422 (2) Güralp DM-16 R 8, 16 bit Hakusan LS-7000 Xt Kinemetrics Q 330 Kinemetrics Q 730 BL Lennartz M 24 8 305 90 10 1 6 2. 0 135 10 3 -6 4. 7 135 20 1 -3 1. 9 3 142 Nanometrics HDR 24 Nanometrics RD (1) Nanometrics Trident Ref. Tek 72 A 3 -6 7. 2 3 -6 SARA SADC 10 SARA SADC 20 Public domain SEISAD 18 (4) Symmetric Res PAR 8 CH TAIDE enterp. TDE-324 F 3. 0 232 145 2. 5 0. 0 2 1. 0 1 GPS GPS pps GPS 0. 5 GPS 1. 6 GPS 10 pps GPS Bu f Tri N N Y Y Y Y N N N Y N N N

The current trend in the development of the different elements of the portable recorders is: Computer: Based on a standard computer and operating system: -Linux seems to be the favorite operating system, but Windows NT/2000 is also used. -Single board PC’s with low power consumption. Communication and data transfer: -RS 232 -Ethernet/TCP/IP -USB + others Sample rate, dynamic range and sensitivity: -Sample rates from 1 -1000 Hz, -Dynamic range of at least 22 bit - LSB (least significant bit) resolution of 0. 1 μV. Standard Data acquisition software Power consumption: Below 2 W.

PC 104 computer 8 cm

Field equipment made by Ui. B

13 cm x 18 cm x 34 cm

Nanometrics Taurus, the next generation handheld recorder (25 x 15 x 6 cm). Power consumption 0. 8 W. From Nanomtrics home page, www. nanometrics. ca.

Name C H Bit Rate T C O Storag e Com Trig P o E nv Wt Strong motion recorders with built in accelerometers, most can also be used with seismometers Geodev. GSMA 2400 3 22 5 100500 0. 5 M S SLT 1. 3 P 10. 0 Geotech DS-2400 36 22 3. 8 101000 0. 5 P SD ST 6. 3 P 16. 0 Geotech LLC 3 16 14. 0 200 0. 5 M S L 3. 3 P 24. 0 Kinemetrics QDR (1) 3 11 100 57 M S L 1. 0 0. 1 Kinemetrics Etna 3 18 8. 0 100250 P S L 2. 2 P 9. 0 Kinemetrics. Mt. Whi t. Geo. SIG IA-1 (2) 18 19 3. 5 P S L 18 P 68 3 16 100200 50 -300 M T 5. 7 3. 3 Geo. SIG GSR 18 3 18 5. 0 100250 20 M SD SL 1. 6 P 7. 2