Tempe Goes Seismic (V1.0 archived page)

current page here

This page is experimental and provides a place where I can post data from a homemade seismometer. In development, artifacts are expected and the data is meant for personal use - to learn. As real events are detected I'll have a better idea of what the sensitivity is.

Details: The seismometer transducer is a torsion balance consisting of a copper penny suspended between two Nd-Fe-B disk magnets by a thin vertical wire. Since copper is diamagnetic, the magnets provide passive damping. Transverse motion causes the penny, which is suspended at an offset from its center, to twist the line. The motion is registered by a pair of photodiodes connected differentially to an amplifier/filter. Photos of the seismometer, and recordings of a few earthquakes which this instrument has detected are posted below the plot of current data.

This instrument went offline on the morning of 3/13/2010 , and was replaced by Version 2.0, a mechanical re-design, on Sunday 3/14.

Seismometer data sample (archived)

Observations:

1. 6/10/09 The main artifact is a response proportional to house temperature (transient when the A/C kicks on). Ideally a differential detector would reject this common-mode signal, but there is enough signal gain that the artifact is visible. Prime suspect is the temp dependence of radiant output of the IR emitter. Biased at constant forward current, the LED output decreases at about 0.5% / deg C. Solution is to design compensation into the LED current source.

2. 6/17/09 Adding a positive temp. coeff has reduced the temperature related artifact. The main issue now is to get a proper material for the torsion balance wire. Nylon line would not hold the penny in a horizontal position without epoxy, which I want to avoid. Solder is easier to work with, so I'm inclined to work with wire instead. I'll be trying some 6-mil dia. (about 34-ga) tungsten steel wire soon. This wire (in a thinner gauge) was used in some balances built in the 1920's, so I'm anxious to try it out. Bonus is that it has a low thermal coefficient of expansion.

3. 6/24/09 The thin tungsten wire has increased the sensitivity of the torsion balance, and gives a good horizontal hold of the penny. Resonant frequency remains a bit high - about 2 Hz. It can be lowered by increasing the offset at which the penny is supported, and by making the support wire thinner. When I take both measures, the resonant freq should approach 1 Hz.
Also, tonight relocated the instrument to a more isolated area. The DAQ device is now connected by USB-over-IP using a Belkin F5L009 network USB hub and a wireless bridge. The data should now have less non-seismic interference.

4. 6/29/09 First detection of a natural seismic event, a mild M4.3 earthquake in Baja California, 372 km away. Plot shown below. Its peak intensity was about 18% of full scale for the instrument.

5. 7/03/09 The seismometer doesn't have a dedicated server. As a result, I see occasional dropouts in the data due to contention on the USB bus when I use a USB HDTV tuner to record or watch ATSC programs off-air. The RF link is robust, but the wireless USB driver loses the contest when hard-wired USB demands bandwidth.

6. 8/16/09 Reduced the tungsten wire diameter from six to five mils (.005 inch). This lowered the natural resonance freq from 2.4 to 1.8 Hz.

7. 10/04/09 Raised the gain of the ADC preamp by 18 dB. This provides enough gain that the sensitivity is no longer quantization noise limited. Also stuffed openings at four edges of the enclosure with foam for thermal isolation. An area that could use further work is the geometry that the penny and flag (light block) present to the photodetector. This can be hard to adjust, and the resting angle of the penny/flag seems to have an impact on sensitivity.

8. 10/24/09 Made a dramatic reduction in 1/f noise by replacing U1 and U3 with OP284 low-noise bipolar op-amps. The unit price for these is almost $9, but the payoff is about 40dB lower noise floor between .01-.1 Hz, and 20dB less integrated noise from 0.1-10 Hz. There is now an output offset of 120mV due to input bias current at U3A (see schematic); this will be addressed at a later date.

9. 11/14/09 Replaced the copper penny in the pendulum with a 40 gram copper block 1 in. sq. * 1/4 in. thick. The noise floor now shows more amplitude below 1 Hz.

Pictures

The completed seismometer board. To the right you can see the IR emitter, magnet assembly and differential photodetector. From right to left, in order:
1) Photodiode transimpedance amplifier/filter (dc-coupled, abbreviated "TZA" in the dataplots)
2) Instrumentation amplifier and HPF
3) Sallen-Key LPF
4) SE-to-balanced output buffer to DAQ board.

The response characteristic is bandpass, with the highpass corner at 0.008 Hz and lowpass corner at 3 Hz.

Seismometer glamor photo, in its enclosure. It's ready to be connected to the data acquisition board.

Seismometer FFT
FFT of the seismometer noise floor. The broad resonance at 2.4 Hz is that of the damped torsion balance. The null at 4.17 Hz is the result of a 3-point moving-average filter applied to the data, which is sampled at 12.5 Hz.

Baja M4.3 The instrument's first detection of a natural seismic event - a magnitude 4.3 quake 50 mi south of Mexicali, Baja Calif: about 370km from my site. Here you can see the onset of the subsurface P-waves first, followed by the much stronger S (shear) waves at 14:41:30 PDT. The event took 60 seconds to propagate 370 km.

Travel Time The USGS travel times to my site pretty much nail what was recorded above - within a few seconds.

Gulf of Calif M6.0 A larger quake, M6.0 in the Gulf of California 07/03/09, 1100 UTC, distance 902km. It is interesting that this was detected at all. The seismometer is oriented for E-W quakes, but this one was at bearing 165, due south from Tempe. It would have registered 4X stronger if it had been due west.

The shear wave component of this quake is lower in frequency and longer in duration than the smaller Baja quake.

Travel Times M6.0 Gulf of Calif. Seismic wave travel times for the M6.0 quake, 7/3/09.

Nomenclature for the phases of seismic waves are given here. The P waves are pressure (longitudinal) waves which travel in a linear path beneath the Earth's crust. The S waves are surface transverse waves which propagate along the curvature of the Earth's crust - a longer path.

Gulf of Calif M6.9 Four earthquakes hit the Gulf of California on Aug 3, 2009 in succession: 5.8, 6.9, 5.0 and 6.2 Magnitude. All four can be seen in this plot. Two of these events caused the instrument to register a full-scale reading. There were reports that the M6.9 was felt in Phoenix, some 900 km away.

The negative and positive transients at 11:37 and 11:40 were human-caused (garage door).

Gulf of Calif M6.9 zoom Zoom plot of the 08/03/09 M6.9 Gulf of Calif earthquake. The longitudinal, higher frequency pressure waves are seen first, followed by the lower frequency transverse shear wave. This one had sustained detectable amplitude for a half hour.

American Samoa M8.0 Detection of a M8.0 Earthquake, 09/29/09 in American Samoa - 8000 km distant.

The pressure wave begins abruptly at 11:00 am (local arrival time). Other clear features are the low-frequency shear wave arrival at 11:09 am, and the Rayleigh surface wave at 11:23:30 Phoenix time.

Travel Times M8.0 Am. Samoa 09/29/09 Seismic wave travel times to Phoenix for the M8.0 quake in American Samoa, 9/29/09.

Haiti M7.0 01/12/10 Detection of a M7.0 Earthquake, 01/12/10 in Haiti - a path distance of 4400 km.

Unlike the M 8.0 Samoa quake above, where three distinct seismic wave components were visible, only the Rayleigh surface wave at 15:14 Phoenix time came up above the instrument noise floor. I wonder if undersea propagation enhances the LR component.

Travel Times M7.0 Haiti 01/12/10 USGS seismic wave travel times to Phoenix for the M7.0 quake in Haiti, 1/12/10. Only the "LR" component rose above the seismometer noise floor.

Notes:

1. Another example of a torsion balance seismometer may be found here .

2. A torsion pendulum is a special case of a compound pendulum. The two main handles on the resonant frequency are the moment of inertia of the weight, and the torsion constant of the wire (inversely proportional to the wire cross-sectional area).

3. A schematic diagram for this design is here. The seismometer is completely powered off of the 5V USB bus from the Labjack U12 12-bit data acquisition board. Supply current consumption is under 100 mA. Note: U3A should be an AD822 and not an MC33202, since low offset is needed in the feedback path.

Back to homepage

	The completed seismometer board. To the right you can see the IR emitter, magnet assembly and differential photodetector. From right to left, in order: 1) Photodiode transimpedance amplifier/filter (dc-coupled, abbreviated "TZA" in the dataplots) 2) Instrumentation amplifier and HPF 3) Sallen-Key LPF 4) SE-to-balanced output buffer to DAQ board. The response characteristic is bandpass, with the highpass corner at 0.008 Hz and lowpass corner at 3 Hz.
	Seismometer glamor photo, in its enclosure. It's ready to be connected to the data acquisition board.
	FFT of the seismometer noise floor. The broad resonance at 2.4 Hz is that of the damped torsion balance. The null at 4.17 Hz is the result of a 3-point moving-average filter applied to the data, which is sampled at 12.5 Hz.
	The instrument's first detection of a natural seismic event - a magnitude 4.3 quake 50 mi south of Mexicali, Baja Calif: about 370km from my site. Here you can see the onset of the subsurface P-waves first, followed by the much stronger S (shear) waves at 14:41:30 PDT. The event took 60 seconds to propagate 370 km.
	The USGS travel times to my site pretty much nail what was recorded above - within a few seconds.
	A larger quake, M6.0 in the Gulf of California 07/03/09, 1100 UTC, distance 902km. It is interesting that this was detected at all. The seismometer is oriented for E-W quakes, but this one was at bearing 165, due south from Tempe. It would have registered 4X stronger if it had been due west. The shear wave component of this quake is lower in frequency and longer in duration than the smaller Baja quake.
	Seismic wave travel times for the M6.0 quake, 7/3/09. Nomenclature for the phases of seismic waves are given here. The P waves are pressure (longitudinal) waves which travel in a linear path beneath the Earth's crust. The S waves are surface transverse waves which propagate along the curvature of the Earth's crust - a longer path.
	Four earthquakes hit the Gulf of California on Aug 3, 2009 in succession: 5.8, 6.9, 5.0 and 6.2 Magnitude. All four can be seen in this plot. Two of these events caused the instrument to register a full-scale reading. There were reports that the M6.9 was felt in Phoenix, some 900 km away. The negative and positive transients at 11:37 and 11:40 were human-caused (garage door).
	Zoom plot of the 08/03/09 M6.9 Gulf of Calif earthquake. The longitudinal, higher frequency pressure waves are seen first, followed by the lower frequency transverse shear wave. This one had sustained detectable amplitude for a half hour.
	Detection of a M8.0 Earthquake, 09/29/09 in American Samoa - 8000 km distant. The pressure wave begins abruptly at 11:00 am (local arrival time). Other clear features are the low-frequency shear wave arrival at 11:09 am, and the Rayleigh surface wave at 11:23:30 Phoenix time.
	Seismic wave travel times to Phoenix for the M8.0 quake in American Samoa, 9/29/09.
	Detection of a M7.0 Earthquake, 01/12/10 in Haiti - a path distance of 4400 km. Unlike the M 8.0 Samoa quake above, where three distinct seismic wave components were visible, only the Rayleigh surface wave at 15:14 Phoenix time came up above the instrument noise floor. I wonder if undersea propagation enhances the LR component.
	USGS seismic wave travel times to Phoenix for the M7.0 quake in Haiti, 1/12/10. Only the "LR" component rose above the seismometer noise floor.