From: "Stauffer, Don (MN65)" Subject: Two questions on seismographs Date: 02 Jan 1997 15:16:07 -0600 Just subscribed to the list, so if these questions are frequent (did browse some of the FAQ, but didn't find them) please bear with me. I am intending to build a Lehman sensor. Why the double pipe upright? I am planning to use a single pipe. I've got to believe resonance of a couple of feet of galvanized water pipe is way above frequency range of interest, or am I missing something. Second, I am thinking of an EO readout. This is something like the Hall effect device in that it reads amplitude directly. Here is my idea. I will have a blade with a notch cut out of a blade of copper. ___________________________/-----------I______________ Well, that isn't the best drawing of a notch, but for ascii, gee whiz. Anyway, a photo interrupter will straddle the notch with the angled edge centered in the beam at rest. If the copper blade moves one way, it uncovers more of the beam and the phototransistor collector current will increase. Movement in the other direction will cover more of beam, decreasing photocurrent. The response will not be real linear, but can be mapped, and in these days of Excel spreadsheets, I can easily linearize the results. The angle of the notch on the operating edge will be optimized for max amplitude. If I make it a straight 90 degrees, the max amplitude will be about 0.3 mm. If I need more, the more I cant the edge the larger the full scale amplitude. Now, the question is, what max amplitude should I design the system for? I have elected to try the photo-interrupter rather than the Hall effect device because I have 25 years of EO experience, and only a couple of magnetic sensor experience, so I feel more comfortable with the EO method. Don Stauffer in Minneapolis _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: John Lucas Subject: Re: Two questions on seismographs Date: Thu, 02 Jan 1997 20:48:43 -0500 Stauffer, Don (MN65) wrote: > > Second, I am thinking of an EO readout. This is something like the Hall > effect device in that it reads amplitude directly. Here is my idea. I will > have a blade with a notch cut out of a blade of copper. > > Don't know if this will be sensitive enough to the very small dislacement but you could double the sensitivity by reflecting the beam off a mirror on the arm. Come to think about it what about bouncing a low power laser beam between 2 roughly parallel mirrors. One fixed, the other on the beam? Beam==> -------------------------- / \ /\ /\ / \ / \ / \ / \ / \ / \ / \/ \/ \ Fixed==> / ------------------- \ / \ Laser Photocell _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Hannon Subject: Re: Two questions on seismographs Date: Thu, 02 Jan 97 21:06:52 -0500 Don, A discussion of EO sensor just occured on the list. Check the PSN web site for the archives http://psn.quake.net/ My instrument is made of 1.5 inch square steel tubing welded. It has only one upright column. The resonances are way up there but when you are dealing with microinch movements things get a little rubbery hence the overkill on stiffness. The EO sensor is what I am working on also. You will need all the sensitivity you can get out of the thing. So a straight notch would be best. I also reccomend using a differential sensor. Two phototransistors or whatever one being covered while the other is being uncovered. Both half covered when centered. Feed this into a differential OP amp. You get 6 dB more sensitivity with only about 3 dB more noise with this arrangment. Karl Cunningham on the list has a EO machine running and is having trouble with noise from the sensor. We are dealing with microvolts for output here. Someone else suggested using a CDS cell for the detector. Having worked with EO sensors do you have any suggestions for low noise? PIN diodes? Both Karl's and my instruments use magnetic feedback to keep things centered. Because you are reading displacment rather than acceleration or velocity long term drift will peg out things if you don't have some way of correcting for the drift. Good luck! Jim Hannon _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: Two questions on seismographs Date: Thu, 02 Jan 1997 23:45:54 -0800 Greetings, Don I didn't understand all of the details of your detector, but IMHO the most important thing to concentrate on is signal-to-noise ratio. So long as you're in the right ballpark, sensitivity-wise, you can play games getting the sensitivity to what you need with optical techniques. For instance, with an analog output range of 20 volts, a noise level of 5mV will give you 4000:1 signal-to-noise ratio. This is 12 bits, which is ok but nothing to write home about. Remember, the earthquake magnitude scale is logarithmic with respect to amplitude, and if you want to cover much magnitude range, things go from big to small *very* quickly. I too believe that a differential approach is the only thing that will give you a chance of overcoming temperature effects of these detectors. I am currently struggling to do better than 12 bits (without a lot of success, I might add). I've played with photo transistors, CdS cells, high-power LED's, various photo masks/slits, and am about to order a laser diode. (I'd welcome any other ideas). My current breadboard has a full-scale (20-volt) output for about a 0.004" movement or so. I'd like to get to a signal-to-noise ratio of 16 bits, with a noise floor of 10 nanometers or less. Right now I'm at about 40-50 nanometers. I like John Lucas' idea of bouncing the laser off off parallel mirrors. I've also thought about having a lens on the moving arm with the laser beam going through the lens. WRT the Lehman mechanical system, keep in mind the order of your low-pass filter and the natural frequency of the mechanical system. Even if your vibration is 50db down in the passband of interest, there can be a lot of vibration at higher frequencies that could get amplified if your mechanical system has a high Q. Several years ago, I had a high-frequency (5Hz to 30Hz) seismic detector running at work just for fun (it was originally made to detect people walking), and during the day it would record a lot of noise. I finally tracked it down to a forklift working across the street in a warehouse about 200 feet away. I don't think it would have contributed much in the range of 0.1Hz to 1Hz, though. Keep us informed on your progress (or even lack of it). Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Larry P. Thomas WA0GWA" Subject: Interesting IC for EO sensors Date: Fri, 03 Jan 1997 08:51:56 -0600 I say this on another SIG dealing with PIC microprocessors and thought it might be of interest to those working on EO type sensors Documentation inf form of pdf files is at http://www-s.ti.com/sc/psheets/soes012/soes012.pdf >The other interesting chip is the TSL23x series of light to frequency >converters. I do not have an appropriate data book so the question is >what is the difference between the TSL230 and TSL235? What TI book contains this information? The 235 is a fixed output version of the 230. The 230 is pin programmable for output frequency scaling and light sensitivity. The 235 is 3 lead 230 device with the sensitivity and output scaling preset by TI. Vcc, Gnd, and Fout about 10Hz to 500kHz 50% duty square wave. All of TI optic sensors are described in the TI Intelligent Optic Sensor databook. I don't have the Book # with me but I cat get it at work if you need it. TI's web site also contains the information. ----------------------------------------------------------------------- Larry P. Thomas wa0gwa voice : 1 913 888-0282 Krell Technologies fax : 1 913 782-9359 8960 Bond pager : 1 816 989-HELP Overland Park, KS 66214-1764 e-mail : lpthomas@................ www : http://www.krell.com ----------------------------------------------------------------------- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: Interesting IC for EO sensors Date: Fri, 03 Jan 1997 08:51:01 -0800 Larry -- Thanks for the tip. I've got a few samples of these devices (both IR and visible) and have played with them a little, but haven't tried to make a detector out of them. The noise seems to be quite good on the bench, but I haven't been able to come up with a good idea for a frequency counter to operate in the range of these devices (0-500kHz) that will have the dynamic range, signal-to-noise ratio, and bandwidth (I'd like to get at least 20Hz update rate). Actually I've got one idea, but it's not very practical. Count frequency and accurately measure the time it takes to count that frequency. Divide the number of counts by the time period and you have frequncy. That sounds like a large project all by itself. Any ideas on this? Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Hannon Subject: Re: Two questions on seismographs Date: Fri, 03 Jan 97 19:20:04 -0500 Don, The major source of drift is mechanical. I can leave my instrument balanced and overnight it will drift a 1/4 inch or so. This all somewhat depends on the period you set the instrument up for, anything longer than 15 seconds and things get real touchy. With a differential EO sensor almost all of the drift associated with the sensor is canceled out. The other problem with the digital HPF is that having the headroom for the drift in the A/D converter range limits the dynamic range the overall system can have. Jim Hannon -------- REPLY, Original message follows -------- CdS cells would not make a very low noise detector, in my opinion. I want to use the photo interrupter because the path is very short. Thus, the phototransistor is working at a high SNR. For drift I plan to use a digital high pass filter with about a minute or so time constant. This will not compensate for physical drift of the edge, but will zero out shifts due to shift in operating point of photo- transistor. Do I need to worry about zeroing out mechanical drift in the unit? One of the reasons I have elected to go with the transistor rather than the photodarlington is to reduce such semiconductor drift. -------- REPLY, End of original message -------- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie & Terri Thompson -- Buda MicroRanch ) Subject: Photodiodes and noise Date: Fri, 03 Jan 1997 22:14:01 -0600 I wanted to comment on electronic noise analysis using photodiodes. Several years ago I designed a 20-bit A/D converter chip for direct connection to a precision photodiode. The application was medical CAT scanners and the CAT scan imaging guys love dynamic range and LOW noise. In my opinion, Hammamatsu makes some of the best precision photo diodes on the planet. They are well specified for noise. Try to get a Hammamatsu photodiode catalog and you will learn at lot about theoretical noise limits in photo diodes. This approach is the lowest noise way to detect modest levels of light (detectable by the human eye). The key items to consider for theoretical best case noise analysis are: 1) Shot noise RMS current - this is fundamental to silicon photodiodes and goes as SQRT(2*Q*I*BW) where Q is the charge on an electron I is the photodiode current and BW is the bandwidth in Hz. The higher the operating current the better the signal-to-shotnoise ratio. Why?... the photo current goes directly as light level but shot noise only increases as the square root. 2) Input amplifier thermal noise current - JFET inputs help keep thermal noise current to a minimum. MOSFET input devices would be better but their 1/f noise is BAD NEWS! A chopped MOSFET input circuit can avoid the lousy 1/f noise problem. 3) Transimpedance resistor thermal noise current - photodiodes exhibit minimum drift when they operate into a virtual ground. (i.e. no voltage across them). This implies they normally would be connected to an op-amp summing node with the op-amp operating in the "trans- impedance" mode (also called a current-to-voltage converter). The resistor from the output back to the inverting input sets the current to voltage conversion scale factor (called the transimpedance). This resistor (all by itself) produces an input-referred RMS thermal noise current equal to: SQRT(4*kTR*BW) where k is Boltzman's constant, T is about 300K at room temp and BW is bandwidth in Hz. The goal is to "RMS sum" the noises in items 1-3 above. On item #1 you determine the shot noise current when you set the amount of light shining on your photo diode for the resting state of your seismometer. On item #2 find an op-amp with LOW thermal noise current...make sure it's specified in the .01Hz to 5 Hz range and make sure input-referred 1/f thermal noise voltage doesn't swamp your noise current contribution. On item #3 you set the thermal noise current by choice of the current-to-voltage conversion scale factor. Sum up items 1-3 as the square root of the sum of the squares. If you are wondering about 1/f noise VOLTAGE...you can input-refer it as a current by dividing the 1/f noise spectral density by the transimpedance resistor. This will give you a 1/f noise current spectral density "at a given frequency" ..i.e. like 1Hz. You can then get the total input referred noise current due to 1/f noise voltage by doing a simple integral over the bandwidth of interest. The TLC2202 has a very low input referred 1/f noise spectral density (for a MOSFET op amp at least!). It's about 60 nV/rt-Hz at 1 Hz. This is plenty low for most applications if your photo current is in the microamp range. So there you have it...you can now predict the electronic noise and compare it to your change in current for a 1 micron RMS pendulum displacement and figure out your signal-to-noise ratio. I have designed photo diode systems that come within about 1 dB of predicted using the above analysis. All-in-all the coil and magnet ain't bad! -Charlie Thompson _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Hannon Subject: Re: Photodiodes and noise Date: Sat, 04 Jan 97 08:23:20 -0500 Thanks Charlie for the photodiode information. Low noise is also important in fiber optic systems where you are trying to cram as much signal bandwidth into as long a cable as possible. A number of years ago I did read up on this in the HP photodiode catalog when looking into the design of a fiber system. Now all we have to do is figure out if the LED contributes any to the noise or should we use an incandesent lamp? Jim Hannon _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Hannon Subject: EO Sensors Date: Sat, 04 Jan 97 19:20:09 -0500 Another idea that might help with the sensor noise problem. That is to use a chopper stabilized amp. but do the chopping right in the sensor by turning the LEDs on and off alternatingly. A lot of noise could be stripped off with a bandpass filter before the demodulator. Jim Hannon _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Don Gentry Subject: Re: Interesting IC for EO sensors Date: Sun, 5 Jan 1997 06:44:05 -0800 (PST) Most off the shelf frequency counters have a function to measure period directly. This is the recipical of frequency. Don't know if it would be usefull at low frequncies or not. This function is normally used when looking at high rate signals and output ranges would be in micro-seconds or milli-seconds. If your output is nearly a square wave it may be able to do what you want. This would give a nice rise time to use to start the period count. I'm out of ideas at this time. Jose' On Fri, 3 Jan 1997, Karl Cunningham wrote: > Larry -- > > Thanks for the tip. I've got a few samples of these devices (both IR and > visible) and have played with them a little, but haven't tried to make a > detector out of them. > > The noise seems to be quite good on the bench, but I haven't been able to > come up with a good idea for a frequency counter to operate in the range of > these devices (0-500kHz) that will have the dynamic range, signal-to-noise > ratio, and bandwidth (I'd like to get at least 20Hz update rate). > > Actually I've got one idea, but it's not very practical. Count frequency > and accurately measure the time it takes to count that frequency. Divide > the number of counts by the time period and you have frequncy. That sounds > like a large project all by itself. > > Any ideas on this? > > > > Karl Cunningham > La Mesa, CA. > karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: Photodiodes and noise Date: Sun, 05 Jan 1997 10:54:11 -0800 Hi Jim -- At 08:23 AM 1/4/97 -0500, you wrote: >Now all we have to do is figure out if the LED contributes any to >the noise or should we use an incandesent lamp? When I had two LED's and two phototransistors in a differential configuration, I tried spraying most of the parts (one at a time) with freeze spray. The most change was when I sprayed either of the LED's. That's when I decided that I should be using one LED and two phototransistors (differentially connected). Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "George A. Harris" Subject: Re: Novel feedback seismometer Date: Sun, 5 Jan 1997 14:03:36 -0800 (PST) For some time I have been investigating seismometer mechanisms. I am just finishing one which is similar to those being discussed recently. Following is a brief description of the construction. If there is enough interest, I could make up a packet of info. to be provided at some nominal fee. The unit makes use of a novel suspended mass which I have successfully used previously. The complete force balance two axis sensor is contained in a cylinder 1 1/2 inches in diam. and 5 inches high. Starting from the base, the construction is as follows: A base supports a short length of brass or other small tubing. It also supports a 5 inch length of 1 1/2 inch plexiglas tube surrounding the small tubing to enclose the assembly. Crimped into the top of the small tube is a length of fine wire. I use .011 in. hardened beryllium copper, but music wire of similar diameter will work. One inch up the wire, another two inch length of small tubing is crimped. To the top of this is attached a ceramic magnet 1/2 inch in diam/ and 1/4 inch thick with a flat upper surface. A one inch disk of thin bright aaluminum with a flat black target pattern on its upper surface is cemented to this flat magnet surface. When the wire length and weight at the top are properly sized, the assembly is barely stable, and will have a natural frequency near 1 Hz when accelerated or tipped in any direction. It is very tilt sensitive, and must be partially supported until vertical (the plexiglas tube does this). The plexiglas tube supports a plastic cylinder one inch high above the target. This cylinder is wound with two coils at right angles from its top to bottom. My coils have 600 turns of #38 wire each. When these coils are placed within 1/8 inch of the magnet, they will support the magnet when it is tipped about 12 degrees with 60 Ma flowing. Thus the range of the unit is approximately + or - .2G. Two sensors are mounted in holes in the plastic coil cylinder. They are a electro-optic unit manufactured by Omron, their EE-SY124. The LED source and transistor are side-by-side in a 4 lead dip which is approximately 1/8 inch square. When used with a black edge on shiny surface 1 mm away it produces a very linear signal for 1/4 mm motion. The output is near 3 volts into a 100K resistance in an emitter follower configuration when looking at the shiny surface with 4 MA drive to the LED. The sensors are stocked by Digi-Key (800-344-4539 their part OR520-ND) for less than $2 each. The present amplifier consists of a type LMC660 quad op-amp followed by type lm386 low voltage audio power amplifiers. Each axis uses two of the op-amps followed by two of the 386's in a push-pull circuit to drive each coil. A simple RC lead network provides damping. With proper gain settings, the output of the first amplifier has a 0-5 V. output corresponding to + and - .2G. George Harris harris@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Mark Wilson mark@.............." Subject: Laser and fiber optics Date: Sun, 5 Jan 1997 14:30:40 -0800 (PST) I have some plans for a fiberoptic pickup unit using a fiberoptic thread coiled around 4 posts. The output is the attenuation of the light through the cable. Has anyone else tried this? I have airblown fiber cable ready to try but was wondering if it worth the effort. ********************************************************************* --- mark@.............. --- http://www.markwilson.com --- wilsonm@............ --- wilsonm@.............. --- mark.wilson@....... --- General Networking Dude and Routing Rugrat ********************************************************************* _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Robert L Barns <75612.2635@..............> Subject: Ground motion amplitude Date: 06 Jan 97 10:22:26 EST Hi gang, On Jan 2, Don Stauffer asked about what amplitude of ground motion should be considered when designing a seismometer. I found a highly enlightening bunch of info on this in 'Quantitative Seismology' vol.1, by Keiiti Aki and Paul Richards, Freeman & Co. 1980, p 493--. Barnes & Noble (a deviant spelling) report that these books are no longer in print. My comments in [ ]. I can't improve on Aki and Richards wonderfully clear exposition so I'll just quote them. "Sec 10.2.1 Surface waves with periods around 20 secs. "The most prominent signals recorded by a standard long-period seismograph, especially classical mechanical ones, for a distant shallow earthquake are surface waves with a period around 20 sec. Body waves show smaller amplitude because of stronger geometrical spreading than surface waves. The surface waves with periods less than about 10 or 15 sec suffer from scattering due to shallow heterogeneities. The attenuation of surface waves is minimal at about 20 sec, because those with periods longer than 25 sec begin to lose energy into the asthenosphere. Besides, the magnification of standard seismographs decreases with periods of more than 20 sec. It was quite natural, therefore, that a magnitude scale Ms was introduced by Gutenberg and Richter (1936) based on the surface wave amplitude at a period of 20 sec. (See Appendix 2, p. 533.) "Fig. 10.8 shows the amplitude of surface waves from a shallow earhquake with magnitude Ms = 3 as a function of epicentral distance. The curve is obtained from the table given by Richter. [Fig 10.8 shows a curve of the log of amplitude in microns vs epicentral distance in degrees. It is almost linear. ] The most sensitive long-period seismometer can detect a distant earthquake with Ms ~3, which will show an amplitude of 100 millimicrons (1e-5 cm) at dist = 20 deg and 10 millimicrons at dist = 80 deg. On the other hand, the largest earthquake (Ms = 8.5) will show an amplitude of several cm at dist = 20 deg and several mm at 80 deg. This large signal dynamic range (1e-6 to 1 cm) imposes heavy demands on seismic instrumentation and recording media (See Box 11.1). "As described in Sec. 10.1, there are three basic sensors that can be used in seismology: the inertial sensor, which is sensitive to acceleration; the strainmeter, which measures the strain in the Earth beneath it; and the gyro, which has the potential for measuring rotation. It is instructive to give the value of acceleration alpha, strain epsilon, and rotation theta associated with the 20-sec surface waves from a small earthquake, say Ms ~3. For plane Rayleigh waves with period 20 sec, phase velocity 3.5 km/sec, and displacement amplitude 0.1 micron, we have alpha ~ (2*pi/20)^2 * 1e-5 cm/sec^2 ~ 1e-6 gal = 1e-9 g [ ~ is 'approx. equal' ] epsilon ~ (2*pi)/(3.5*1e-5*20)*1e-5 ~ 1e-11 theta ~ epsilon = 1e-11 ~ 2*1e-6 arc-sec "Large explosions, such as an underground nuclear test, can also generate 20-sec surface waves with amplitude approx. proportional to the yield. A 1-megaton shot in hard rock roughly corresponds to Ms ~ 5 3/4 and thus the most sensitive long-period seismometer will detect a distant shot of several kilotons (corresponding to Ms ~ 3). "Sec 10.2.2. P-waves for dist 5 deg to 110 deg. "The signal level of P-waves from a distant earthquake may be found from Gutenberg's calibration curve (see Richter, 1958, p 688) for determining the body-wave magnitude mb. Fig. 10.9 shows the value of A/T as a function of epicentral distance, where A is the amplitude in microns and T is the period in sec for a shallow earthquake with mb = 4. This curve can be used to find mb for any shallow earthquake, as mb = log(A/T){obs} - log(A/T){mb=4} + 4 where (A/T){obs} is the observed value of A/T at a certain epicentral distance (which must be known) and (A/T{mb=4} is the value obtained from Fig. 10.9 for the distance. [The curve in 10.9 is wiggly especially below 30 degrees.] For P-waves recorded by standard seismographs, T is usually around 1 sec, and the amplitude is about 10 millimicrons at dist = 20 deg and 1 millimicron at dist = 90 deg for mb = 4. These signals may be detected by the most sensitive short-period seismometers. The greatest earthquake (mb ~ 8) will show A/T of 1mm/sec at dist = 20 deg. For such large earthquakes, T may be about 10 sec, and the amplitude on the order of 1 cm. Again, we see a requirement for large dynamic range from 1e-7 to 1 cm. [Even a 16-bit A/D converter has a dynamic range of only 65,000. Vol. 2, p 573 describes seismology as used in oil exploration. They use seismographs which have electronic binary gain controls on their amplifiers, i.e., when the signal rises above a certain level, the gain is reduced by a factor of 2 until another level is seen when the gain is again reduced by a factor of 2, etc. When the signal level falls, the gain in increased. They get a dynamic range as large as 170db! The electronics to do this are simple. You Californians and New Zealanders might benefit from this scheme] "The 1 millimicron displacement at T =1 sec corresponds to an acceleration of 4*1e-10 g. and to rotations and strains of around 1e-12." The next section (10.2.3) discusses amplitude as a function of frequency but I'm getting tired of typing so get your local library to do an inter-library loan for you. Sec. 10.2.5 is "Ambient seismic noise." This is of special interest to me since I see far more of that than quakes--New Jersey being a seismic desert. Curves are given which show rapid changes in spectral noise density with frequency and a factor of 100 between 'quiet' and 'noisy' locations. There is a fairly sharp minimun around 0.03 Hz and a sharp max. around 0.13 Hz. Wind and esp. ocean waves are the main causes of the noise. The first part of Chapter 10 is an excellant discussion of the mechanics of seismometers and, of course, detailed mathematical analysis of these things. An interesting statement here: "Since the sensitivity of a pendulum to acceleration at low frequency is proportional to the square of its free period,...". Well, I'm still digesting all this and trying to relate the quantitative aspects to what my calibrator and Lehman give. Bob Barns Berkeley Heights, NJ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Cristiano Subject: Re: Laser and fiber optics Date: Mon, 06 Jan 1997 08:25:23 -0700 Mark Wilson mark@.............. wrote: > > I have some plans for a fiberoptic pickup unit using a fiberoptic thread > coiled around 4 posts. The output is the attenuation of the light through > the cable. Has anyone else tried this? > I have airblown fiber cable ready to try but was wondering if it worth the > effort. Mark, I've seen reference to this fiber optic seismo device a while back in a mag called Science Probe!. I think Forrest Mims' son used it in a science project. I also have plans that came from a science project book on how to make one. I think it would be hard to get usefull information from this device because of the differing resonant frequencies of the separate fiber optic loops. The frequencies would also be very high. Maybe if you had ONE fiber being bent by the business end of a horizontal boom? Jim Cristiano cristiano@........... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Blatchley, Brett" Subject: Re: Laser Seismology... Date: Mon, 6 Jan 1997 13:00:27 -0500 > I have some plans for a fiberoptic pickup unit using a fiberoptic thread > coiled around 4 posts. The output is the attenuation of the light through > the cable. Has anyone else tried this? > I have airblown fiber cable ready to try but was wondering if it worth the > effort. There is a chapter entitled "Experiments in Laser Seismology" in a book called "The Laser Cookbook: 88 Practical Projects" by Gordon McComb, ISBN 0-8306-9390-4, TAB Books, $19.95. It walks you though building a laser/fiberoptic seismometer. This is a great book to have on hand, but the author does not go much into how he calibrated this unit. As I understand it, the magnitude of acceleration is proportional to the audio frequency of the change in interference fringes (at the output of the fiber as "read" by a photo diode/transistor). I'm not concerned with the sensitivity of this design as much as its apparent omni-directional view. I.e. I think you can get useful data from it. >>think it would be hard to get useful information from this device because >>of the differing resonant frequencies of the separate fiber optic loops. The >>frequencies would also be very high. If you want more accurate data, I think a ring-laser is the thing to build. Here you are actually counting phase changes (interference fringes) which are measurably stable because the laser light remains coherent throughout its trip (unlike the fiber optic unit above). I.e. every light to dark transition at the output is one wavelength, and since you know the laser's wavelength, the rest is arithmetic. Here is a good URL to get you started on ring-lasers: www.phys.canterbury.ac.nz/research/ring_laser/ring_laser.html This is my first contribution, I'm enjoying your thoughts :-) ------------------------------------------------------------------------ -- Brett Blatchley, Senior Computer Engineer INMAR Enterprises, Inc. 910-631-2825 voice 910-631-7703 fax ------------------------------------------------------------------------ -- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Stauffer, Don (MN65)" Subject: Baseplate material? Date: 06 Jan 1997 12:59:34 -0600 In working on the design for my seismograph, I am looking for suggestions on the baseplate. One mechanical guru here at work is trying to convince me to use steel plate. My problem with that is getting steel plate locally without a hundred buck minimum. I was thinking of particle board with sheet metal glued top and bottom to make a sandwich, but he talked me out of that. I am making a two-axis sensor with each arm just shy of 2 feet in length, so I need a stable baseplate about 24 x 24 inches. Or, an L-shaped plate 24 inches on a leg. Any suggestions on a material that is readily available to the hobbyist? Don _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: Baseplate material? Date: Mon, 06 Jan 1997 11:22:06 -0800 Hi Don -- Look in the yellow pages under Metals. In cities of moderate size, there are usually surplus dealers that are willing to cut pieces for you (for a small fee) and have small (<$10) or no minimums. Some places have cutoff bins full of all shapes and sizes and you save the cutting charge. Another option is to look for surplus electronic equipment and find something with a chassis similar to what you want. I've used aluminum ($1.50/lb surplus) for most of my projects. It has a worse temperature coefficient of expansion that steel, but it doesn't rust and is easy to drill and tap. :) Just some thoughts. Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie Thompson) Subject: Re: Baseplate material? Date: Mon, 06 Jan 1997 14:22:57 -0600 Don't use ANYTHING that will be affected by humidity. I tried a wood base plate and the unit became a hygrometer as well as a seismometer. Steel or other humidity insensitive material would be the best material for the baseplate and upright. Tiny changes on the 'shape' due to humidity will cause you to have to readjust your sensor with each passing weather front. My steel baseplate stays adjusted nicely. Steel gets my vote. -Charlie Thompson .BUE > In working on the design for my seismograph, I am looking for suggestions >on the baseplate. One mechanical guru here at work is trying to convince >me to use steel plate. My problem with that is getting steel plate locally > without a hundred buck minimum. I was thinking of particle board with >sheet metal glued top and bottom to make a sandwich, but he talked me out of >that. > >I am making a two-axis sensor with each arm just shy of 2 feet in length, so >I need a stable baseplate about 24 x 24 inches. Or, an L-shaped plate 24 >inches on a leg. > >Any suggestions on a material that is readily available to the hobbyist? > >Don > > > >_____________________________________________________________________ > >Public Seismic Network Mailing List (PSN-L) > >To leave this list email listserver@............. with the body of the >message: leave PSN-L > > _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: George Bush Subject: Re: Baseplate material? Date: Mon, 6 Jan 1997 12:29:24 -0800 (PST) Don- At 12:59 PM 1/6/97 -0600, you wrote: > In working on the design for my seismograph, I am looking for suggestions >on the baseplate. One mechanical guru here at work is trying to convince >me to use steel plate. My problem with that is getting steel plate locally > without a hundred buck minimum. I was thinking of particle board with >sheet metal glued top and bottom to make a sandwich, but he talked me out of >that. > >I am making a two-axis sensor with each arm just shy of 2 feet in length, so >I need a stable baseplate about 24 x 24 inches. Or, an L-shaped plate 24 >inches on a leg. > >Any suggestions on a material that is readily available to the hobbyist? > I am planning on casting a simple slab from concrete for the base of my 3-axis (Amatuer Scientist) seismometer. You may want to go this route also. It is inexpensive, uses readily-available materials, is easy to cast, can be made any shape and can be made as massive as you want. George _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: bob@................. (Bob Hammond (USGS-AVO Geophysicist)) Subject: Re: Baseplate material? Date: Mon, 6 Jan 1997 11:30:49 -0900 I would suggest you look at thick PVC plastic. I used to be able to buy PVC plate as thick as 2 inches. Maybe putting two 2-inch PVC plates together would avoid the humidity problems and be less expensive than steel. The PVC was very stiff but easy to work with ordinary tools. Bob Hammond _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Re: Baseplate material? Date: Mon, 6 Jan 1997 22:35:12 -0800 At 12:59 PM 1/6/97 -0600, you wrote: > In working on the design for my seismograph, I am looking for suggestions >on the baseplate. I would like to suggest two pieces of railroad rail mounted on three bolts each mounted in concrete, in the form of an "L". The three bolts (2 at one end separated as far as possible and one at the other will let you adjust each sensor separately for mechanical balance. Rail is often available from the railroads as surplus, sometimes lying around their yards. Some people like to use it for fireplace andirons or truck bumpers. I have seen short lengths lying around in lots of places I have been. btw 3/8 threaded rod would be fine for the bolts. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie & Terri Thompson -- Buda MicroRanch ) Subject: USGS missed two events! Date: Mon, 06 Jan 1997 22:08:08 -0600 The first event missed by USGS was on 1/4/97 at 3:36 UTC. I just ignored the fact that they didn't report the approx. M5.3 event located 1300 miles from Buda, TX (probably near Chiapas Mexico). Then USGS missed a second event approx M5.1 on 1/4/97 at 18:44 UTC also about 1300 miles from Buda, TX. (probably Chiapas again). I know I'm closer to Mexico than the USGS sensors but normally they have detected and reported events of this size in the past...what gives? Was USGS "down" part of the day on 1/4/97?? I have detected 3 additional events with the same epicenter in Mexico and they correlated nicely with past USGS reports. All of the events were approximately 1300 miles from Buda, TX near Chiapas, Mexico. Any clues appreciated. -Charlie Thompson .BUE _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Dott. Roberto E. Pozzo" Subject: seismic equipments used in Italy Date: Wed, 08 Jan 1997 03:59:54 +0100 Hello to all the list ! I'm on psn-list from three months (+or-). Until now, I was a 'good listener' of your very interesting (& useful) exchange of ideas and experiences on amateur seismology. Here in Italy, although here seismicity is one of the highest in Europe, amateurs are not many ( and more or less unconsidered by officials) . A dozen of us, in last four years , have developed a three period seismographic station , based on a kit published years ago by a magazine of electronics and more and more times improved and changed in circuitry by us. Essentially it is, a long-period horizontal seismograph with electromagnetic coils (with oil damping in the simplest version) linked on a op-amp circuit that drives analog signals (+or- 5v) to a first AD converter . The same thing makes our vertical sensor ( t= 1 sec.) Another equipment allows triggered or continuos printing on a high speed thermal printer at a max speed of 120 cm/hour ( 48 inches /hour )and ,finally, drives signals to a datalogger based on an AD-card and dedicated software that shows on pc screen three traces at a time and allows storage on hard disk or floppy disk ( unfortunately, using an average rate of 30 samples per second , we have a file of about 2 Mbytes in 24 hours, and so our life goes on changing diskettes !!) . Recently ( thanks, Larry Cochrane and Ted Blank ) we have reached the winquake-compatibility on our recordings , so final result is good. We preconize to use EMON or SDR on our equpments as soon as possible ( sending one of our pc-cards to U.S.A. for compatibility tests is not simple like drinking a cup of water !!). Our precison clock (connected on serial port of pc and synchronizing pc clock) is based on DCF-77 standard emission from Mainflingen (Germany) and works so and so ( LW reception is affected by a lot of disturbances, more part of them 10 minutes before a quake ! The radiocontrolled clock stops, jammed by some propagation noise, the recording shows strange times (like 34:21 am :-) ) and earthquake comes a while after !!! Mr. Murphy lives in Italy.Don't you know ? If someone of you is interested , we can mail some schematic plan and other infos about our complete system. We also collect ,naturally, complete reports on seismicity in Mediterranean area. Final remark: I live in Piedmont region,near Turin , on January weather is snowy and moderately cold. Another cold front is passing and my long period shows one-inch waves..... regards Roberto _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: cougercat@........... Subject: Re:seismic equipments used in Italy Date: Tue, 7 Jan 97 21:04:18 PDT Greetings Roberto; In regaards on details on your seismometer, I'd like a copy of the details. Currently, I'm working out the bugs (lots of them...;} on a XYZ detector using the Analog Devices ADXL05 series of tri-axial accelerometers. Thanks; Jeff Chang 2235 22nd Ave East Seattle, Washington 98112 U.S.A. FAX # [206] 325 0307 P> If someone of you is interested , we can mail some schematic plan and P> other infos about our complete system. We also collect ,naturally, P> complete reports on seismicity in Mediterranean area. ******************************************* * mail from Jeff - cougercat@........... * ******************************************* _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Dott. Roberto E. Pozzo" Subject: Re: seismic equipments used in Italy Date: Thu, 09 Jan 1997 01:45:04 +0100 Jeff, yes, I'll send to you some material dealing with our seismic system. Please, be patient....I'm preparing it . A few days again and I hope it will be ready to send .Many compliments for your good sperimentation on tri-axial accelerometers. Let me know how your work goes on. Here in Italy we can obtain some accelerometer (u.s.a.f. surplus) for a cheap price.Some time ago, I was tempted to purchase one or two of that devices . What do you think about aeronautical accelerometers ? You live in Seattle and Boeing Co. is not too far ! :-) Roberto _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: cougercat@........... Subject: Re: seismic equipments used in Italy Date: Wed, 8 Jan 97 20:43:45 PDT P> Organization: osservatorio meteosismico del monferrato Hi Roberto; Thanks for your reply. Yes Boeing's is here but I have not been able to pick up any decent accelerometers from them. The accelerometer that I am experimenting around was made for the airbag industry. They have a span of +- 4 g's, Sensitivity is 500 mV/g, Bandwidth is DC - 100 hz or anything between, Supply voltage/current 5/24 ma. Almost ideal for portable use. The August 8, 1991 issue of Electronic Design carried an good artical on them. These were sampled for another project that didn't get off the ground from Analog Devices. The bochure says the European phone # is +89 57005 0 .....Perhaps you can get ahold of them and see if they will send you a sample or 2. I'll keep you posted. --Jeff ******************************************* * mail from Jeff - cougercat@........... * ******************************************* _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ACole65464@....... Subject: Nova TV Program: EQ Prediction and Preparedness Date: Thu, 9 Jan 1997 02:28:29 -0500 (EST) Last night here in the Seattle area the local PBS station aired an old NOVA program regarding earthquakes, and compared the quakes of Kobe Japan and Los Angeles. But what I wanted to point out, to some in our group, was the type of seismometer that they showed in an underground vault in California, contained in a glass jar. It was a Streckeisen STS-1 vertical seismometer. I only recognized it after recently coming across a description of it in an old "Bulletin of the Seismological Society of America", Dec 1982. The TV program's narrator described the seismometer only as a Swiss made instrument. The BSSA's paper described the instrument well but seeing a video of it was good because it showed an actual installation. The instrument, one for each axis, is only 180mm x 180mm x 120mm. The two horizontals have the usual styrofoam covers etc but vertical is mounted within a glass bell jar that is evacuated down to 10 mbar (that reduced the internal thermal problems). The silver container seen within the glass jar is an aluminum and "Permalloy" cover over the instrument to form both an RFI and magnetic shield. The designers used an LVDT for the displacement transducer (so when we get tied of trying to get our photo-optical type devices to work lets experiment with an LVDT). Their instruments are capable of resolving ground noise from .1 to 3000 seconds and displacement to 10xE-10 meters peak-to-peak, WOW! Also on this same program they showed the small accelerometer used to activate automobile air bags, another device that we have heard people talk about on this network, that many of us have not seen. I hope you have, or had, a chance to see this NOVA program. Regards, Allan Coleman _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Robert Mathieson Subject: Re: Nova TV Program: EQ Prediction and Preparedness Date: Thu, 9 Jan 1997 08:22:33 -0600 (CST) At 02:28 AM 1/9/97 -0500, you wrote: > Last night here in the Seattle area the local PBS station aired an old >NOVA program regarding earthquakes, and compared the quakes of Kobe .... > > The designers used an LVDT for the displacement transducer >(so when we get tied of trying to get our photo-optical type devices to work >lets experiment with an LVDT). gram. > > Regards, > Allan Coleman > What is an LVDT? Thanks. Robert Mathieson Culver-Stockton College Canton, MO 63435 (217)233-6000 rmathieson@.......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Larry P. Thomas WA0GWA" Subject: LVDT Date: Thu, 09 Jan 1997 08:58:39 -0600 >What is an LVDT? Linear Variable Differential Transformer. Basically it is a cylinders with typically three coils on it. The outside two feed a diode detector (or better yet a synchronous detector) with one producing a positive voltage and the other producing a negative voltage. These signals are summed (i.e. differential) to produce the output. The center coil is feed with the excitation voltage. A magnetic material suitable for the frequency of the excitation voltage is allowed to slide within the cylinder this differentially will increase the coupling to one outside coil while reducing it in the other outside coil. This is a very basic explanation of its operation. I'm sure there are things that can be don to improve it's performace for very small displacements as would be found in seismic applications. ----------------------------------------------------------------------- Larry P. Thomas wa0gwa voice : 1 913 888-0282 Krell Technologies fax : 1 913 782-9359 8960 Bond pager : 1 816 989-HELP Overland Park, KS 66214-1764 e-mail : lpthomas@................ www : http://www.krell.com ----------------------------------------------------------------------- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jim Hannon Subject: Re: LVDT Date: Thu, 09 Jan 97 19:59:21 -0500 I started on a LVDT design a while back. Did some reading on them. Found out that the commercial ones are very expensive usually made for aerospace applications so I built the transformer and mounted it on my seismograph. I used a ferrite bead for the core and a delrin bobbin to wind the coils on. It has some 3/4 inch conduit for an outer shield. Linear has a IC that does all the support electronics for the transformer. I have two samples. Then I got switched over to thinking about the Opto sensor. I had got as far as feeding a sine wave into the coil and watching the output on a scope as the seismograph arm swung. One thing about the LVDT is that it should be very linear especially for small displacements. Once I get my workshop back together I suppose I will end up trying both sensors. Jim Hannon _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Request for input Date: Thu, 9 Jan 1997 22:17:26 -0800 I have recently retired (12 days ago) and have been considering interesting projects to become involved in. I am looking at the possibility of designing a seismometer kit. What do you think? I am an electronics design engineer (analog and digital) and a machinist. I have purchased equipment for a machine shop, 2 lathes, a milling machine, and a surface grinder and will be moving from Pasadena to southern Oregon in the spring. After moving and setting up the shop, and of course getting my own seismo equipment up and running, I would hope to get something started. This gives some time for planning projects so I thought I would ask for opinions on this list. I don't know of a better place. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Robert Avakian Subject: Re: Request for input Date: Fri, 10 Jan 1997 02:20:26 -0800 Al Allworth wrote: > > I have recently retired (12 days ago) and have been considering interesting > projects to become involved in. I am looking at the possibility of designing > a seismometer kit. What do you think? > > I am an electronics design engineer (analog and digital) and a machinist. I > have purchased equipment for a machine shop, 2 lathes, a milling machine, > and a surface grinder and will be moving from Pasadena to southern Oregon in > the spring. After moving and setting up the shop, and of course getting my > own seismo equipment up and running, I would hope to get something started. > > This gives some time for planning projects so I thought I would ask for > opinions on this list. I don't know of a better place. > ____________________________________________________________________________ > > Al Allworth. Pasadena,CA aallworth@........ > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L As a teacher, it sounds great IF you can get the cost down to say $200-300 per kit. That would mean a full station would cost a school LT $1000 if you made both horizontal and verticla machines. That price should turn almost any earth science teacher's head. You'll need good documenation and support literature, though. Not many of us have geophysics, EE or similar degrees. Bob Avakian Midland TX _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Stauffer, Don (MN65)" Subject: LVDTs Date: 10 Jan 1997 07:48:16 -0600 One thing to watch with LVDTs is temperature coefficient of core materials. The most likely result would be scale factor error with temperature, but if the magnetic circuit is not symmetrical, it could cause a small offset. The scale factor change with temperature can be characterized and corrected for. Some expensive ones do this. Don in Minneapolis _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Robert L Barns <75612.2635@..............> Subject: Seismometer kit Date: 10 Jan 97 10:27:27 EST Al Allworth, Your idea of offering a seismometer kit sounds good to me but I have no idea how big the market is. I think that a long period (say 10-20 sec) device would have wider appeal since short period accelerometers are avail. for less than $50. Larry Cochrane, who has lots of experience with both a Lehman and a Shackelford-Gunderson, has been saying that he gets better results with the S-G. Its smaller size has great appeal. It would probably have smaller raw material cost. The S-G is well described in The Amateur Scientist column of the Scientific American mag. in Sept. 1975, pp 182-187. Also, since you live close to Larry, he could be a handy source of advice. I would be interested in buying one to work alongside my Lehman. Bob Barns _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie Thompson) Subject: Re: Request for input Date: Fri, 10 Jan 1997 10:49:32 -0600 Al, Congrats on your retirement! I think a seismometer mechanical kit would complement Larry Cochrane's most excellent electronics boards. Especially if you could price the mechanism right at the "nuisance level" for most of us. In other words a kit priced such that if you factor in a busy schedule (like mine) your deal couldn't be beat. Also I think the Princeton Earth Physics Project could use a sensor kit (other comments regarding PEPP welcomed). -Charlie Thompson Seismic station .BUE Buda, Texas >I have recently retired (12 days ago) and have been considering interesting >projects to become involved in. I am looking at the possibility of designing >a seismometer kit. What do you think? > >I am an electronics design engineer (analog and digital) and a machinist. I >have purchased equipment for a machine shop, 2 lathes, a milling machine, >and a surface grinder and will be moving from Pasadena to southern Oregon in >the spring. After moving and setting up the shop, and of course getting my >own seismo equipment up and running, I would hope to get something started. > >This gives some time for planning projects so I thought I would ask for >opinions on this list. I don't know of a better place. >____________________________________________________________________________ > > Al Allworth. Pasadena,CA aallworth@........ > > >_____________________________________________________________________ > >Public Seismic Network Mailing List (PSN-L) > >To leave this list email listserver@............. with the body of the >message: leave PSN-L > > _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Jeff Batten Subject: Seismic recording system for under 500$ Date: Fri, 10 Jan 1997 09:27:15 -0800 Note: If you already have a seismometer, or are thinking of constructing your own. Buy a interface card from Larry. If you want a complete system consider. The AS1 seismic recording system for under 500$. It includes a 2.5 second vertical seismometer. 12 bit interface and amp box (New and improved) Logging and Plotting software for windows and dos. (PSN Format). The system is now totally plug and play, you can have it up and running in less than a hour. Records quakes up to 10000km away. If you are a teacher and have a old XT sitting around, why not put it to good use recording quakes. Setup. Find a nice site for the seismometer. Connect the interface cable to the seismometer. Plug the interface cable into a serial port. Run the software. No offset knobs or gain settings to worry about. Check out www.primenet.com/~seismo If you are a home machinist I also have free plans for a Lehman. I also have Lehman coils if you do not want to wind your own. Jeff _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Dennis Leatart Subject: Re: Request for input Date: Fri, 10 Jan 1997 13:58:42 -0800 Hi! I am an earth science teacher and would encourage you to investigate all the possibilities. I would recommend you contact Jeff Batten in this list group for some of his experiences-- he has learned a lot by designing his own software and interface for seismometers. I am currently successfully running a station with his design. -- _____ __ | \.-----.-----.-----.|__|.-----. | -- | -__| | || ||__ --| |_____/|_____|__|__|__|__||__||_____| _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Robert W. Teller" Subject: RE: Request for input Date: Sat, 11 Jan 1997 10:06:54 -0700 Hello Al, Your idea about building an economical seismometer is GREAT! Perhaps we = can get together on this, I have just finished the initial design and = have operational a 1-8 channel 12 bit AD that uses a PC serial port for = data capture. This would allow any PC with a serial port to use the = hardware. I have also written a program that displays selected channels = much like only using Windows 3.1 or 95. It also saves captured events in = the standard PSN format so veiwing/analyasis programs like WinQuake can = be used. Let me know what you think. Bob Teller Montana Research Labs P.O. Box 977 Townsend, MT 59644 (406)266-4483 ---------- From: Jeff Batten Subject: Build your own Force Balance Seismometer Date: Sat, 11 Jan 1997 09:19:32 -0800 If you are trying to build your own force balance seismometer you should check out A miniature wide band horizontal-component feedback seismometer. Journal of Physics E: Scientific Instruments 1977 Ovulate 10 It has a pretty good description of a FBS. I think that a lot of readers of this group have the skills to construct the seismometer. The article has a good drawing and photos. If you have a milling machine and lathe, give it a try. Another good reference. The design of miniature wide band seismometers M. J. Usher and C Guralp Geophysical Journal of the Royal Astronomical Society 1978 (55), 605-613 M J usher I W Buckner R F Burch If you like seeing a lot of unique designs for seismometers check out old Bull. Seismol. Soc. Am. The BSSA has a few hundred articles on seismometer design back to 1906. You will probably have to visit a university library to find the above references. Jeff Batten - Research Engineer Caltech Seismo Lab 818-395-6965 Fax-818-564-0715 _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Lucas Haag Subject: Help!! Date: Sat, 11 Jan 97 14:41:02 CST I'm 13 years old and would like to get into sismology. I live right next to many oil fields (see http://www.swnebr.net/~lhaag/nogcc.html I designed this page). We have lots of small (2.0) tremmors or microquakes or something like that. I am trying to build a very inexpensive and simple seismograph. I ve looked at plans and stuff and on your site and its much to complicated for me. I'm OK with electronics but I'm no rocket scientist! If anybody can help please do.. Thanks Lucas Haag ************************************************* Lucas Haag, Assistant Field Agent Nebraska Oil & Gas Conservation Commission HCR 66 Box 25A Bartley, Nebraska 69020-9717 lhaag@.............. http://csb.swnebr.net/~lhaag "Through da keyboard, into the chips, over the copper, through the fiber, back through the copper into the chips, to the screen... NOTHING BUT NET!" ************************************************* _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Steve Hammond Subject: Re: Ignore the glitch in my data Date: Sat, 11 Jan 1997 14:49:44 -0800 (PST) Michoacan, Mexico 6.8Ml -- Please, will somebody in San Jose area that recorded this tell me if they also recorded a smaller event at 20:33:37 about 145KM from South San Jose? The arrival of the P wave in the Michoacan event is much too large and I think is a smaller local event that happened within seconds of the other. Regards, Steve Hammond PS: as I have six files, I have ZIPPED them and put them on the PSN BBS in San Jose. Anybody know how to send zip files to Larry's WEB site? On Sat, 11 Jan 1997, Charlie & Terri Thompson -- Buda MicroRanch wrote: > A BIG quake just occurred about 900 miles from me...about Mag 7.0.. > probably in Mexico. Peak signal was at 10,000 counts...the largest > I have ever detected. I didn't believe it and took the lid off > the sensor....hence the glitch in my uploaded data...please disregard > the 32,000 count gliches after the S-wave. > > Charlie Thompson > Seismic Station BUE > Buda, Texas > > ct@....... > > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L > _____________________________________________________________________ From: ct@....... (Charlie & Terri Thompson -- Buda MicroRanch ) Subject: Mexico Quake - now showing! Date: Sat, 11 Jan 1997 18:07:00 -0600 I added a link to my web page that allows anyone with a web browser to view today's Mexico Mag 6.8 quake..as recorded with Larry's SDR and plotted with Larry's Winquake program. http://www.onr.com/user/ct/ct.html Here's how I turned a Winquake image into a JPEG image: Basically I viewed the quake using Larry's Winquake program. Depressed the ALT-PRINT SCRN keys to capture the Winquake image to the Clipboard. I then pasted the results using Edit -> Paste into PC Paint program that comes with Windows 3.1. Then saved the file as a ".bmp" file. Finally, I used a nifty program called L-View to convert from .bmp to JPEG and added the JPEG to my web page. This technique is a good way to get any Window's program output to a JPEG file. -Charlie Thompson Buda, Texas USA _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Steve Hammond Subject: Re: Mexico Quake - now showing! Date: Sat, 11 Jan 1997 17:24:21 -0800 (PST) Charlie, I looked at the J-PEG you put on your WEB site in TX and can clearly see that you also recorded the same very large arrival waves that I recorded here in San Jose. Maybe John Lahr or one of the other folks from the USGS can jump in here, but it looks to me like there were two events; one smaller at 20:30 followed by a the second event at 20:31 time frame. Anybody care to comment... When I first looked at the data I thought that there had been one local event in California and one distant event in Mexico but, now I think I was wrong. It appears that the initial P wave had lost all it's energy by the time it reached us and we triggered on the S waves of the smaller first event. BTW, my San Jose files are on Larry's WEB site now under the FSX extension. One is three data sets spliced together and the other two are the raw data from the first two event records I captured. Regards, Steve Hammond PSN BBS San Jose. _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Darrell Collins Subject: Re: Help!! Date: Sat, 11 Jan 1997 18:51:46 -0700 (MST) At 02:41 PM 1/11/97 CST, you wrote: >I am trying to build a very inexpensive and simple seismograph. I ve >looked at plans and stuff and on your site and its much to complicated for >me. I'm OK with electronics but I'm no rocket scientist! If you have a Home Depot or other home center in your town, you can get everything needed to build a very workable seismometer. The electronics, which you should be able to handle, are not difficult, although you will need to spend some cash. You posted to the net, so you must have a computer of some kind. That is the expensive part. Software (PC kind) is available from the PSN sites. All you need is some PVC (plastic) pipe, glue, a coil, and a magnet. If you are interested return my E-Mail and I'll help you. Darrell _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Steve Hammond Subject: Re: Help!! --- GIF and instructions Date: Sun, 12 Jan 1997 14:51:14 -0800 (PST) Lucas, I emailed you the VERTICAL.GIF drawing which is the same file as VERTICAL.ZIP found on all the PSN BBSs. It is a simple to build but works great seismograph developed by Terry Frost in Pasadena, Calif. This gets kind of long so if you don't want to know how to build it stop reading here-- All you need is some aluminum plate 8" x 14" and 24 inches of 3 inch aluminum channel for the boom and post. (The boom is 12" long as is the post.) The spring is a common screen door screen that you can buy in any hardware store. The coil can be made from a 2 1/2" plastic spool, such as the one Radio Shack uses on their #22 wire which has a center hole of 1" and is 3/4" thick, and #32 - #34 magnet wire (Allied Electronics 1-800-433-5700 stock number 214-3586 (#32 - $22.64) or better 214-3588 (#34 - $23.87). The magnet-- is a cow magnet bought from a farm feed and grain store for about $2.50 each. Last, a bit of hardware such as an aluminum L bracket, BRASS nuts washers and screws. You don't want anything that is steel because the magnet mounts on the free moving boom and could be drawn to it which would interfear with the seismograph operation. Building the device-- you need a hacksaw, file, and hand drill. You may have to make the L bracket from a larger length of aluminum L stock. Cut it as wide as the inside of the channel (2 3/4") so it will bolt to the base plate inside the channel at the rear of the device. Take the 24" x 3" channel and cut it into two halfs 12" long. Use the brass screws and drill and attach one of the sections, becoming the post, to the base plate. Mount the post so that the bottom part of the U channel is facing the center of the base plate. See the GIF. ------------------- I base plate I I- I U II------------- I boom channel II------------- I post I- I I I ------------------- I I \ door spring post I \ I---------M- boom M= magnet L I O C L= "L" bracket, O= oil damping, C=coil ------------------ base plate The other half becomes the boom with one end resting on the post and the other suspended by the spring. Use the file and cut a V in the post filing a grove about a 1/16" deep across the 3" inside flat face, 2" up from the base plate. File a 45 degree knife edge on what will be the left end of the boom (you will need to use the hacksaw and trim a bit of the two U parts of the channel leaving just the boom's 3" face) so it will have the minimal contact with the post when set in the V cut. It is important to file the V at a 90 degree angle to the post sides so that the boom is level when suspended and hinges freely without interference from the boom's channel sides that you trimmed off. File the boom's knife edge so it is as sharp as possible. I built one with an edge that had just two dimples --V------------V-- that came into contact with the V cut in two spots and it worked well for me. But I suggest that you start with the simple knife edge until you get the unit working, then make improvements. The magnet is held in place with a force fit. Drill a hole near the right end of the boom and then split the channel cuting down the center of the 3" channel from the right edge to the center of the magnet hole. This will allow the channel to expand a bit when you force the magnet into the slightly smaller magnet hole. The cow magnet I used was 1/2", so I drilled a 15/32" hole. Drill a small hole at the center top of the post and about a 1/2" to the left of the magnet hole (EG 1 1/2" or 10 1/2" from the ends of the 12" boom) so that the spring will be free and clear of the magnet top poking up through the boom. Use some #22 solid brass wire (if you buy the Radio Shack solid wire for its plastic spool you will have plenty after you pull it off the spool) and attach a length of the door spring to the post and boom so that the spring supports the boom at a level position with the cow magnet mounted in its hole. I later added some small screws with holes drilled in them and a small aluminum turnbuckle to adjust the spring tension (boom position) making this step much simpler than trying different length wires. A thought about the spring is appropriate at this point. Select a spring that is soft and has a slow recovery. You don't want snap-- You want soft gentle spring action because it will affect the period of the boom. I was able to cut a maximum length spring off of the longer door spring and mount it between the post and boom giving me a device with a 2.5 second period. (This is the amount of time it takes for the boom to swing up, down past center, and up again to the starting position. A device with a period of 2.5 seconds will be excellent for the oil field area you described and will be able to sense small and large regional events. Damping-- is this important-- yes, because it keeps the boom from oscillating after being set into motion. Damping is accomplished by the use of oil in this design. Other designs may use magnets. Take a long 1" brass screw, an aluminum tuna fish can, and some 30 weight motor oil and make the damping unit. Make sure the can is not steel by testing it with the cow magnet. Place the can under the boom with brass screw bolted through the boom. The end of the screw should be centered in 4 directions inside the can. You want to keep the boom as light as possible so I cut a 1" square out of the tuna lid and punched a hole in the center of it. I then bolted it with brass nuts to the bottom of the screw, centered inside the tuna can, thus becoming an oil baffle after the motor oil was added to the can; creating an oil damping system. Don't cut yourself, it's edges are sharp-- and will leave a nasty gash. You can re-mount the damper up or down the boom which will increase or decrease the amount of damping if needed. Coil-- drill a hole in one side of the Radio Shack spool near the center. Make sure there are no sharp edges, by sanding the plastic, which will catch as you wind the magnet wire. Start out buy putting 2 or 3 inches of wire through the hole and then tape it down to protect it. Caution, if this wire breaks off where you can not connect to it you will have to start over... I go to the trouble of gluing this wire in the hole and up the side a bit with clear silicon glue so I can repair it if needed. Wind the #34 magnet wire into the plastic form (spool) by hand. This takes about an hour and you don't want to break the wire. Good luck-- Fill the with the magnet wire. The part numbers from Allied I gave you is for magnet wire that can be soldered without removing the insulation. This will make solding the connecting wire simpler. I also like to have a solid wire mount so I use super glue and glue a plastic tie onto the side of the near the wire exit hole. I then loop #22 wire to it and solder the magnet wire to the larger wire. I also always add a tie down to the base plate of the seismograph and loop the #22 wire to it. This in turn stops me from accidentally ripping the wires out of the coil... when installing the seismograph. Tape the coil to the base plate so the cow magnet can is centered in the middle of the 1" spool hole. Adjust the height of the cow magnet after the boom spring is adjusted so that the cow magnet is about 1/4" above the base plate. You may want to lift the coil up with small wooden blocks so that it is 1" above the base plate. This will assure maximum induction. So at this point, you have a vertical seismograph with a 1 - 3 second period that will be great for recoding local events. I recorded a 4.5 at 400km near Bishop with this unit as well as several 1.0 events in my area. The PSN BBS in San Jose is at 37 14 43.33N - 121 48 54.52W within a few kilometers of the split of the Hayward and Calaveras and 16km from Loma Preata and the San Andreas. A simple amp circuit can be found on any of the PSN BBSs in the file PENDULUM.ZIP which contains the plans for the Lehmann sismograph. Larry C. also builds a great interface board that will interface this to a PC and there are several other circuits talked about here on the PSN forum. I'll be happy to discuss this and answer questions. Regards and best of luck, Steve Who am I? I'm SYSOP at the PSN BBS San Jose 408-226-0675, SHAMMOND@........... SHAMMON1@.............. SHAMMOND@........... and Steve Hammond at WWW.CHIPANALYST.COM. It just depends on the hour and day of the week. On Sat, 11 Jan 1997, Lucas Haag wrote: > Thanks for the reply. All of that I can handle. We have about all that stuff > laying around here (I live on a farm) I would appriceate some plans. > > At 06:51 PM 1/11/97 -0700, you wrote: > >At 02:41 PM 1/11/97 CST, you wrote: > >>I am trying to build a very inexpensive and simple seismograph. I ve > >>looked at plans and stuff and on your site and its much to complicated for > >>me. I'm OK with electronics but I'm no rocket scientist! > > > >If you have a Home Depot or other home center in your town, you can > >get everything needed to build a very workable seismometer. The > >electronics, which you should be able to handle, are not difficult, > >although you will need to spend some cash. You posted to the net, > >so you must have a computer of some kind. That is the expensive part. > >Software (PC kind) is available from the PSN sites. All you need is some > >PVC (plastic) pipe, glue, a coil, and a magnet. If you are interested > >return my E-Mail and I'll help you. > > > >Darrell > > > > > > > >_____________________________________________________________________ > > > >Public Seismic Network Mailing List (PSN-L) > > > >To leave this list email listserver@............. with the body of the > >message: leave PSN-L > > > > > ************************************************* > Lucas Haag, Assistant Field Agent From: "Dott. Roberto E. Pozzo" Subject: Re: seismic equipments used in Italy Date: Mon, 13 Jan 1997 01:33:39 +0100 cougercat@........... wrote: > > P> Organization: osservatorio meteosismico del monferrato > > Hi Roberto; > > Thanks for your reply. Yes Boeing's is here but I have not been able to > pick up any decent accelerometers from them. The accelerometer that I am > experimenting around was made for the airbag industry. They have a span of > +- 4 g's, Sensitivity is 500 mV/g, Bandwidth is DC - 100 hz or anything > between, Supply voltage/current 5/24 ma. Almost ideal for portable use. > > The August 8, 1991 issue of Electronic Design carried an good artical on > them. These were sampled for another project that didn't get off the > ground from Analog Devices. The bochure says the European phone # is > +89 57005 0 > ....Perhaps you can get ahold of them and see if they will send you a > sample or 2. > > I'll keep you posted. > > --Jeff > > ******************************************* > * mail from Jeff - cougercat@........... * > ******************************************* > > > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L Hi Jeff, many thanks for info about accelerometers. Please read my reply to Edward Cranswick about this matter. By now, the only data I can obtain on those types of accelerometers are the ones that I posted to Edward. Isn't a sensibility of +- 20 G too high for a seismo use ? What do you think about ? greetings roberto _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Dott. Roberto E. Pozzo" Subject: Re: seismic equipments used in Italy Date: Mon, 13 Jan 1997 01:28:36 +0100 Edward Cranswick wrote: > > Roberto- > Interesting what you are doing. > What are the specs on the USAF accelerometer below? > -Edward > > > tri-axial accelerometers. Let me know how your work goes on. Here in > > Italy we can obtain some accelerometer (u.s.a.f. surplus) for a cheap > > price.Some time ago, I was tempted to purchase one or two of that > > devices . What do you think about aeronautical accelerometers ? You live > > -- > Edward Cranswick Tel: 303-273-8609 > US Geological Survey, MS 966 Fax: 303-273-8600 > 1711 Illinois St cranswick@........ > Golden, CO 80401 USA E.M. Forster said, "Only connect". > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L Many thanks, Edward for your nice reply. I alway read what you write her in PSN-L because it's very interesting and well documented. Ok, here follows the only techical data that I can obtain (by now) from a catalogue issued by an italian electronics surplus company about those accelerometers: - made by ETHER Co. , Type LA2 - 2 sensors 2300 ohm (positive and negative) - sensibility ( three diferent available types) +- 20 G ,+- 30 G , +- 50 G dimension: 45x60x25 mm - new in original case The cost is about 25,000 Lire here in Italy ( about 17 U.S. dollars) I don't know much more about those devices , but I think that perhaps the sensibility is too low for seismological use ( isn't true ? ) What do you think , Edward ? many thanks again roberto Dr. Roberto Ezio Pozzo Osservatorio Meteosismico Del Monferrato Via Crova,3 14049 Nizza Monferrato (AT) ITALY tel. (039) 0141 793026 - 0337 240912 ---------------------------------------- - _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: LEDs in Accelerometer, wavelength question Date: Sun, 12 Jan 1997 22:01:03 -0800 Greetings -- To fill everyone in about my experiments with LED temperature coefficients ... I really didn't do much other than to spray parts with freeze spray. This was in a circuit with one LED for each of the two phototransistors. The phototransistors were in a differential configuration and right next to each other, so I figured that any temperature change in one should be matched in the other fairly closely. Those changes would be canceled electronically. The same was true for the LEDs. |----- interrupter connected to beam | LED --> || -- -- || <-- phototransistor LED --> || --|-- || <-- phototransistor | Top View After seeing some amount of low-frequency noise, I wondered how much might be due to differential temperature changes in the LEDs or the phototransistors. So I sprayed each in turn with freeze spray and noted the amount of output change. I don't have the exact numbers, but I think the change was equivalent to about 1E-5 meter of movement of the beam, when I sprayed each of the LEDs. The phototransistors showed a change too, but about 10:1 less. The phototransistors were operated saturation, with the output being derived from the difference in saturation voltage between the two. At that point, I decided to try only one LED rather than two. I mounted the one farther away from the detectors (about 1"), so it would shine clearly on both. The overall noise didn't go down because the light level reaching the phototransistors (and hence the current) went down a lot, due to moving the LED farther away (one of those 1/d^2 things, you know). After that I switched to a much brighter LED (see below), but there still wasn't any drastic improvement in noise -- perhaps 2:1 or so. Since I still had about 10:1 more noise than I wanted, I didn't pay too much attention to the details of the difference between two LEDs and one. I just figured that one has got to be at least as good as two (and ought to be better). The LEDs I was using at the time were part of a General Electric H23A1 (IR wavelength) matched emitter-detector pair, and were only specified as a coupled pair. I changed to a "super-brite" red LED (unknown manufacturer, from local surplus store). This LED is VERY bright. At 80mA it could work as a flashlight in a pinch. Even though the detector was only characterized for IR, it seems to work fine with the red LED. In fact, using the same setup geometry, the phototransistor current is about 100x what it was using the IR LED. By the way, does anyone have any idea whether using IR vs red LED vs blue LED vs a UV source for the EO detector is going to make a significant difference? I'm trying to get noise levels down to 1E-8 meter or better at the interrupter, and the wavelength of red light is 60-70 times this. It seems that what the phototransistors see is really the sum of MANY photons, and the wavelength of each one is not that significant, but I'm guessing. Comments? Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Peter Styles Subject: Re: seismic equipments used in Italy Date: Mon, 13 Jan 1997 21:39:34 GMT Dear Edward/Roberto We purchased a dozen of the Analog Devices ADXL05 accelerometers and a student of mine built up the tri-axial seismometer described in last years Scientific American. We weren't too impressed by its sensitivity and don't really think they are usable even they do have a good bandwidth. They do however make great digital inclinometers and we have built some to fit in a heat flow probe we will be using in Lakes in southern Chile next month. > I don't know much more about those devices , but I think that perhaps > the sensibility is too low for seismological use ( isn't true ? ) > What do you think , Edward ? > > many thanks again > roberto > > Dr Peter Styles Home: 6 Swan Lane Bunbury Tarporley Cheshire CW6 9RA 01829 260808 Work: Department of Earth Sciences University of Liverpool Brownlow Street Liverpool L69 3BX (44) 0151 794 5174 _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: watson@................ (Charles Watson) Subject: Re: USGS Missed Mag 5.0 quake Date: Mon, 13 Jan 1997 14:13:50 -0800 Charlie & Terri Thompson -- Buda MicroRanch wrote: > > I recorded a M5.4 (approx) quake at about 6:43:35 1/3/97 UTC about 1000 miles > from Buda, Texas. It appears the USGS missed this one. Did anyone > else record this event? > > Thanks, > > Charlie Thompson > Buda, Texas > ct@....... Charlie, Just a matter of record, the NEIC doesn't miss much. It is more are they are understaffed and underfinanced to perform quick analysis outside the US borders. It will probably come out on the QED and most certinly in the PED. You can sometimes get quick information from IDC: http://www.cdidc.org:65120/web-bin/recentevents You can find IDC's URL and other sources of current earthquake information at: http://www.seismo-watch.com/EQSERVICES/CurrentActivity/CurrentActivity.html --- Charles P. Watson, Consulting Geologist Advanced Geologic Exploration Seismo-Watch / Earthquake Information Services Voice: 702-852-0992 / Fax: 702-852-3226 watson@................ http://www.seismo-watch.com _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: "Dott. Roberto E. Pozzo" Subject: Re: Ignore the glitch in my data Date: Mon, 13 Jan 1997 01:56:27 +0100 Charlie & Terri Thompson -- Buda MicroRanch wrote: > > A BIG quake just occurred about 900 miles from me...about Mag 7.0.. > probably in Mexico. Peak signal was at 10,000 counts...the largest > I have ever detected. I didn't believe it and took the lid off > the sensor....hence the glitch in my uploaded data...please disregard > the 32,000 count gliches after the S-wave. > > Charlie Thompson > Seismic Station BUE > Buda, Texas > > ct@....... > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L Hello Charlie, for your info , the Mexico Quake made here in Italy a good recording on my home built (near Lehman type...) long T seismograph . P and S phases were confused by a weather disturbance (a cold front was passing just yesterday above us) , but Rayleigh L phases performed a fine 15 minutes recording. Unfortunately, Winquake elaboration was not available , because just in this period , my data-logger is 'explosed' on my tech-deck due to mainteinance works. Mr Murphy lives in Italy , as I always say ! regards roberto Dr. Roberto Ezio Pozzo Osservatorio Meteosismico Del Monferrato Via Crova,3 14049 Nizza Monferrato (AT) ITALY tel. (039) 0141 793026 Cell. 0337 240912 ------------------------------------------ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Mike Lozano Subject: Re: Steve Hammond & Lucas Haag Q&A Date: Sat, 30 Nov 1996 10:51:42 -0600 > Who am I? I'm SYSOP at the PSN BBS San Jose 408-226-0675, > SHAMMOND@........... SHAMMON1@.............. SHAMMOND@........... and > Steve Hammond at WWW.CHIPANALYST.COM. Lucas Haag > Hi Steve & Lucas, I've never written anything to this newsboard, but I've been listening for a few months. In reference to your 'cheap and easy' vertical seismograph for use in local events: I too was going to build a Lehman type instrument, but (at the time) I couldn't get the magnets specified from Edmund quickly enough for my impatience. So, I visited the local farm and ranch store and purchased a set of 3 one-half inch diamenter alnico cow magnets for $11. After staring at them for the better part of a day, I came up with a slightly different approach. Since the current induced in the coil is largely dependent on the velocity of the movement, the number of turns in the coil, and the strength of the magnet; why not increase the number of magnetic lines per unit length by putting similar poles together? I took some 1/2 inch PVC, stuffed two cow magnets into it, separated by a thin piece of rubber sheet (to prevent 'clanking'), and capped the two ends with PVC caps. The coil had been placed on the PVC tube before capping: _________|***|_________ | (magnet 1)|(magnet 2) | | ______________________| End cap |***| Coil To make a long story short, this 'packing' of the magnetic lines of force really seems to increase the current produced by a substantial amount. So much so, that I had to put my Fluke 87 meter on the Volts range, because the Mv scale went right off the page with very little movement. I've yet to locate the one-inch thich 12" X 30" aluminum plate I want to use as a base plate. I know it's somewhere in storage, but I've forgotten exactly where! I'm totally new to seismology. I'm a meteorologist and electronic engineer by profession. Is this avenue worth pursuing? Or, should I just be patient, and wait for the Edmund Scientific magnets to arrive? Thanks for your help and tolerance of dumb questions. Mikel Lozano, mikel@....... -0- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Lucas Haag Subject: Re: Steve Hammond & Lucas Haag Q&A Date: Mon, 13 Jan 97 20:29:16 CST I think your're probaly on to something there. It's a good idea. Now to I need to find a amplifier circut of some type. I'm going to record data on cassetes not paper using plans in a Radio Shack book. Do you think this will work?? At 10:51 AM 11/30/96 -0600, you wrote: >> Who am I? I'm SYSOP at the PSN BBS San Jose 408-226-0675, >> SHAMMOND@........... SHAMMON1@.............. SHAMMOND@........... and >> Steve Hammond at WWW.CHIPANALYST.COM. Lucas Haag > >Hi Steve & Lucas, .... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Mike Lozano Subject: Re: Steve Hammond & Lucas Haag Q&A Date: Sat, 30 Nov 1996 12:50:57 -0600 Lucas Haag wrote: > > I think your're probaly on to something there. It's a good idea. Now to I > need to find a amplifier circut of some type. I'm going to record data on > cassetes not paper using plans in a Radio Shack book. Do you think this will > work?? Hi Lucas, Thanks for the reply! I, myself, plan to use one of Larry Cochrane's a/d boards into my computer, but I suppose, you could build your own from Radio Shack plans. As you're planning to use a cassette as a storage medium, you'll have to turn the varying voltages into a frequency. That's assuming you want an analog system. In my design, for example, the magnets are fixed, and the coil moves. I'm sure that will present some problems with wire flexure, and the like; but I plan to have the coil wires go up the stainless steel horizontal pendulum support. I found a piece of solid brass 1/4" rod in my junk box - so, tentatively, I plan to make the pendulum about 30 inches long. As I mentioned in my first message: I'm just getting started in this, and lots of the things I try may not be according to Hoyle, but at least I'll have fun doing it. By the way, Lucas, I see you're in Nebraska. I'm in the next state, East - Iowa. Matter of fact, the television station where I work as a meteorologist can be found under http:\\www.kcci.com Be sure and check out our weather page. Anyway - got to run. Oh - by the way! I just noticed that the clock and date on this computer is screwy. Please disregard the date and time on the message. This is being written on the 13th of January, 1997 at 9:57PM CST. Again, thanks for responding! As we say in my native Basque language: 'Mila eskerrik asko - lagunak', or 'Thanks a bunch, friend' Mikel Lozano -0-> > > ************************************************* > Lucas Haag, Assistant Field Agent > Nebraska Oil & Gas Conservation Commission > HCR 66 Box 25A > Bartley, Nebraska 69020-9717 > lhaag@.............. > http://csb.swnebr.net/~lhaag > > "Through da keyboard, into the chips, over the copper, > through the fiber, back through the copper into the chips, > to the screen... NOTHING BUT NET!" > ************************************************* > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Dual Cow Magnets (was Re: Steve Hammond & Lucas Haag Q&A) Date: Mon, 13 Jan 1997 21:30:02 -0800 To Mikel Lozano: One question on your arrangement -- is the coil going to move back and forth on the outside of the PVC pipe, or are the magnets going to move inside it. No matter which, I'm sure that the more voltage (or current) you get out of the coil, the better. I think the objective is to have the moving coil cut across the most magnetic lines of flux with the smallest movement. Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie & Terri Thompson -- Buda MicroRanch ) Subject: Re: USGS Missed Mag 5.0 quake Date: Mon, 13 Jan 1997 23:52:49 -0600 You are correct...it did show up on the link you suggested below. It was a Mb4.1 of the coast of Jalisco Mexico. This is the smallest quake I've detected in Mexico...I estimated it the quake as a M5.5 event. I'm still having trouble downloading a USGS report to append to the Winquake file...will try again tommorrow. thanks! -Charlie Thompson Buda, Texas >Charlie & Terri Thompson -- Buda MicroRanch wrote: >> >> I recorded a M5.4 (approx) quake at about 6:43:35 1/3/97 UTC about 1000 miles >> from Buda, Texas. It appears the USGS missed this one. Did anyone >> else record this event? >> >> Thanks, >> >> Charlie Thompson >> Buda, Texas >> ct@....... > >Charlie, > >Just a matter of record, the NEIC doesn't miss much. It is more are they >are understaffed and underfinanced to perform quick analysis outside the >US borders. > >It will probably come out on the QED and most certinly in the PED. You >can sometimes get quick information from IDC: >http://www.cdidc.org:65120/web-bin/recentevents > >You can find IDC's URL and other sources of current earthquake >information at: >http://www.seismo-watch.com/EQSERVICES/CurrentActivity/CurrentActivity.html > >--- >Charles P. Watson, Consulting Geologist >Advanced Geologic Exploration >Seismo-Watch / Earthquake Information Services >Voice: 702-852-0992 / Fax: 702-852-3226 >watson@................ >http://www.seismo-watch.com > >_____________________________________________________________________ > >Public Seismic Network Mailing List (PSN-L) > >To leave this list email listserver@............. with the body of the >message: leave PSN-L > > _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Re: Dual Cow Magnets Date: Mon, 13 Jan 1997 21:57:04 -0800 This looks like a win-win arrangement to me. More flux from the two magnets and the lost flux from the ends pushed around to where it can do the most good. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Re: Request for input Date: Mon, 13 Jan 1997 22:18:57 -0800 List, Thanks for the input re: Seismo kits. I'm getting very anxious to get moved so I can begin prototyping some of my ideas. The more I think about it the more ideas come and all must be tried. I have several, both electronic and mechanical that I will discuss when I work out some details. I have some discussions I will have privately with some of you in the near future. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Ron Westfall Subject: Lehman magnet mod Date: Tue, 14 Jan 1997 11:02:45 -0800 I noticed Mike Lozano's contribution about putting several cow magnets in a PVC pipe, so I thought it was time to mention the modification I made to the Lehman design. I had noticed that using a single magnet with the pickup coil was problematic because of the limited clearance both in width of the opening between poles and depth before the coil ran into the back of the magnet. I used two magnets and their keepers to form a much larger C shaped magnet: __________ ___________ __________ | N ||___________|| S | | ____N_||___________||_S____ | | | | | | | |_________| | | | | | | | | | |______ | | ______| | | S | | COIL | | N | |________S_| | | |_N________| | | | | The advantages of this approach in order of importance are: coil dimensions are no longer critical, there is greater depth of travel, and magnetic field strength is greater. With the original design the coil was already fairly narrow to fit in the gap of the magnet. Coil form side wall material had to be carefully chosen to keep the side walls thin yet achieve stiffness, so there would be enough wire in the dense part of the magnetic field. With the above approach the side walls can be thicker opening up the range of materials that can be used. The greater depth of travel is a minor advantage, because any motion that would exercise the greater travel would probably result from a local quake. Motion from a local quake would probably saturate the electronics. The only benefit is that you now don't have to worry about the coil running into bolt heads that are fastening the magnet to a support. For those of you who have not built the seismograph yet, the magnets have a hole through the back of the magnet (base of the U shape). Its natural to use a bolt to fasten the magnet to a support. In the original design this leaves either a bolt head or nut sticking into the interior of the U shape. Given the smallish dimensions of the magnets this is an intrusion that can't really be afforded. You can also glue the magnets or use some other form of attachment, but a brass bolt is darn convenient. The two paired magnets lastly offer the advantage of greater field strength. Like Mike experienced, swinging the 10,000 turn coil through an unoptimized gap (the original keepers result in slightly too large a gap) gave voltage transients anywhere from 0.1 to 0.5 volts with 0.25 volts being typical. The greater field strength may or may not be an advantage, because you will have to turn down the gain on the preamp a corresponding amount to avoid saturating the A/D electronics. I have not installed the electronics yet, so the greater field strength could turn into a disadvantage if the preamp gain can not be turned down enough. The disadvantage of using paired magnets is that now you need two magnets instead of one, so cost goes up. On the other hand, I found a local supplier of the same magnets that Edmund distributes. They cost about $25 Canadian (about $17.50 US) each. It sure beat Edmund's price which is about $48 - $50 US each. I encourage all of you to just look in the phone book under magnets. The local supplier I found is a small hole-in-the-wall engineering company that specializes in industrial magnetics. Amongst more sophisticated magnetic gadgets, they sold the horseshoe magnets used in the seismograph. For those of you in locations where there are no suppliers close by, Mike's contribution and mine should suggest that the magnet configuration is not particularly critical. Use your imagination and find some combination of permanent magnets which will provide an overall pull roughly in the range of the 25 lb. mentioned in the original design. Ron Westfall westfall@...... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Mike Lozano Subject: Re: Dual Cow Magnets , continued . . . Date: Sun, 01 Dec 1996 01:31:33 -0600 Karl Cunningham wrote: > > To Mikel Lozano: > > One question on your arrangement -- is the coil going to move back and > forth on the outside of the PVC pipe, or are the magnets going to move > inside it. > > No matter which, I'm sure that the more voltage (or current) you get out of > the coil, the better. I think the objective is to have the moving coil cut > across the most magnetic lines of flux with the smallest movement. > > Karl Cunningham > La Mesa, CA. > karlc@......... > Hi Karl! The design I have in mind at the moment keeps the magnets stationary, with the coil being attached to the boom. Because the gap between the coil and magnet housing is about 1/16", the boom should be able to move an appreciable distance. Unlike you lucky? guys in California, I'm not too worried about saturating the amplifier. A fews months ago, I had a design project which called for an abnormally large dynamic range in the input data. I solved it with a transistorized amplifier with a very non-linear output curve. Almost like a logarithmic amplifier in reverse. Of course, the idea behind putting the magnets in opposition is to maximize the flux lines over a very small area. I wish I knew how to submit .DXF drawings, but a visualization of the fields opposed to each other will reveal that the packing of lines of force dies off exponentially from the line of max opposition. Using that logic, a deflection in either direction from the center point would produce less current as the coil encountered less and less flux lines. In a nutshell, this seems (at least on paper) that deflection per unit time would be self limiting. In the case of the horseshoe magnets, on the other hand, it would seem that current would be mostly a function of velocity, since the flux lines between unlike poles would be (at least in theory) uniformly distributed. As I said earlier, here in the MidWest, unless the New Madrid fault, or the Mid-Continent Rift Zone decide to both get frisky, I'm not too worried about saturation. As a matter of fact, Raymond Anderson of the Iowa Dept. of Geololgy & Resources instrumented the Rift zone for five years, and got nothing more than a couple of microseisms. I'd be deliriously happy if I could get even so much as a squiggle from a distant earthquake. I do intend to use the Lehman design. Incidentally, the magnets encased in PVC fit very nicely into a PVC floor flange which I am going to mount at a 90 deg angle to the base, so that the magnetic tube will be parallel to the base. Please stay in touch! Mikel Lozano -0- > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Re: Lehman magnet mod Date: Tue, 14 Jan 1997 20:25:08 -0800 At 11:02 AM 1/14/97 -0800, you wrote: > >I noticed Mike Lozano's contribution about putting several cow magnets in >a PVC pipe, so I thought it was time to mention the modification I made to >the Lehman design. > Ron and List, I would be surprised if this approach produced much of an improvement in output. One thing to keep in mind with these coil/magnet combinations is that it is NOT how much flux you can get through the coil on axis, it's how much flux is cut by the movement of the coil. With opposite poles facing each other the flux lines will be pulled toward the opposite pole and not as many will turn outward through the coil. Any increase in output would be caused by the increase in total flux in the area of the coil. This can be demonstrated with iron filings on a piece of cardboard over the magnets. The ideal would be if you could get all of the flux to turn 90 degrees and go outward through the center area of the coil. Then any movement would cause all of the flux to cut turns on the coil. A coil twice as long as the diameter of the hole might also capture more of the flux and give more output. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Karl Cunningham Subject: Re: Dual Cow Magnets , continued . . . Date: Tue, 14 Jan 1997 22:53:23 -0800 Hi all -- I have a couple of theories about magnets in coils, but since my knowledge in magnetics is sorely lacking, I thought I'd ask for more knowledgeable comment... 1. The VOLTAGE induced in a coil cutting across magnetic lines of flux is proportional to the number of lines cut by the coil per unit time (speed of the coil). (Assuming the voltage is measured by a high-impedance amplifier.) 2. The CURRENT induced into a coil (assuming it is feeding a current shunt -- low impedance) is proportional to the amount of force applied, and that the speed of the movement will be proportional to that force. This force over a distance will be doing work, dissipated as heat in the resistance of the coil. The force is there because the coil's current will be opposing the movement by generating an opposing force proportional to the current in the coil. Not a very elegant description, but am I even close to what actually goes on? Comments appreciated. Karl Cunningham La Mesa, CA. karlc@......... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Mike Lozano Subject: Re: Dual Cow Magnets , continued . . . Date: Sun, 01 Dec 1996 05:31:05 -0600 Karl Cunningham wrote: > > Hi all -- > > I have a couple of theories about magnets in coils, but since my knowledge > in magnetics is sorely lacking, I thought I'd ask for more knowledgeable > comment... > > 1. The VOLTAGE induced in a coil cutting across magnetic lines of flux is > proportional to the number of lines cut by the coil per unit time (speed of > the coil). (Assuming the voltage is measured by a high-impedance amplifier.) > > 2. The CURRENT induced into a coil (assuming it is feeding a current shunt > -- low impedance) is proportional to the amount of force applied, and that > the speed of the movement will be proportional to that force. This force > over a distance will be doing work, dissipated as heat in the resistance of > the coil. The force is there because the coil's current will be opposing > the movement by generating an opposing force proportional to the current in > the coil. > > Not a very elegant description, but am I even close to what actually goes > on? Comments appreciated. > > Karl Cunningham > La Mesa, CA. > karlc@......... > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L Hi Karl, Seems like you're right on the money! In fact, the old-fashioned D'Arsonval meter movements (which are sort of miniature seismometers when they're on their backs) used to be protected (and probably still are) by shorting the two terminals together with a piece of solder or copper wire. It makes a dandy damper for the movement when the terminals feed an extra low impedance. Of course, in the case of a D'Arsonval movement, the term would be resistance, since it's a direct current which is generated by a single movement. I suppose for the measurement of magnitude, it's best to have a low impedance input, but since I'd be tickled to even see a squiggle here in the MidWest, I'm going for the high impedance amplifier. I'm particularly interested in seeing the generated waveforms. Mikel Lozano -0- _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Doomsday David Allen Subject: Re: Does anybody know of anyone who will doe a demonstartion Date: Wed, 15 Jan 1997 15:13:23 -0700 Darrell Collins wrote: > > At 08:42 PM 1/14/97 -0700, you wrote: > >I Am senior at Dysart High School. I am Interested in siesmology > >The earth science teacher at my school asked if me if I could get some > >one to come to our school a do a demonstration with a siesmograph. > > Hello, > I am in Scottsdale, and could help. I teach at Scottsdale Community > College. Please let me know about when you would want the talk, the > subjects to be covered, and the amount of time I would have for the > presentation. I also need to know how much background your class > has had (plate tetonics, mid-ocean rifts, sub-duction/up-heaval/slip > etc). I need to get a seismometer "portable", (these things are heavy!). > Perhaps it would be good for your teacher to phone me some evening? > Darrell > Phone: 602-948-8993 > Fax: 502-998-8931 > > _____________________________________________________________________ > > Public Seismic Network Mailing List (PSN-L) > > To leave this list email listserver@............. with the body of the > message: leave PSN-L Yes I would be a good Idea if you talk to the teacher about a date and stuff. I do know how much a portable siesmic unit can Weigh, when I was a junior at Dysart I audited Asu Exploration geophysics class so I got to see a lot of the equipment used. As for the earth science class These kid are Mostly freshman Maybe one or two sophmores and the teacher will have already probably taught them the basic on earthquake and plate techtonics. I also am hopefully bringing an anclimometer in to show them too. I give your phone number to the teacher so he can call you and set up a time Thank you very much, I sure the kid will Really Enjoy it Gregory David Allen _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Al Allworth Subject: Re: Dual Cow Magnets , continued . . . Date: Wed, 15 Jan 1997 23:59:41 -0800 At 10:53 PM 1/14/97 -0800, you wrote: >Hi all -- > >I have a couple of theories about magnets in coils, but since my knowledge >in magnetics is sorely lacking, I thought I'd ask for more knowledgeable >comment... > >1. The VOLTAGE induced in a coil cutting across magnetic lines of flux is >proportional to the number of lines cut by the coil per unit time (speed of >the coil). (Assuming the voltage is measured by a high-impedance amplifier.) > >2. The CURRENT induced into a coil (assuming it is feeding a current shunt >-- low impedance) is proportional to the amount of force applied, and that >the speed of the movement will be proportional to that force. This force >over a distance will be doing work, dissipated as heat in the resistance of >the coil. The force is there because the coil's current will be opposing >the movement by generating an opposing force proportional to the current in >the coil. No.1 is correct. For No.2, you are correct except you have a situation where you can never really see a true current because of the resistance of the coil. If the coil has a resistance of 2K then you always have this 2K in series with the "generator" so that any external load can only receive the current that can be passed through the 2K from the theoretical 0 Ohm generator. Some instruments are designed with a particular winding resistance and a specified load resistance which results in critical damping of the instrument. The current through the coil produces the correct opposing force for damping. Often the signal level is reduced to much less than 1/2 what it would be using another means of damping and a high impedence amplifier. ____________________________________________________________________________ Al Allworth. Pasadena,CA aallworth@........ _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Ron Westfall Subject: Re: Lehman magnet mod Date: Thu, 16 Jan 1997 10:48:34 -0800 Al Allworth wrote: > > I would be surprised if this approach produced much of an improvement in > output. One thing to keep in mind with these coil/magnet combinations is > that it is NOT how much flux you can get through the coil on axis, it's how > much flux is cut by the movement of the coil. With opposite poles facing > each other the flux lines will be pulled toward the opposite pole and not as > many will turn outward through the coil. Any increase in output would be > caused by the increase in total flux in the area of the coil. This can be > demonstrated with iron filings on a piece of cardboard over the magnets. The > ideal would be if you could get all of the flux to turn 90 degrees and go > outward through the center area of the coil. Then any movement would cause > all of the flux to cut turns on the coil. A coil twice as long as the > diameter of the hole might also capture more of the flux and give more output. > My primary goal in coupling two magnets together with their keepers into a larger C shape was to ease the tight geometry of the original design. As I mentioned in my post, any increase in magnetic field strength is probably offset by having to turn the gain down on the pre-amp. I performed an experiment last night to try and verify whether there was any field strength increase. I used a single magnet in the original configuration and two magnets with their keepers in my modified configuration to determine what voltage was induced in one of my sensor coils. Holding the magnet(s) in my hand, I moved the magnet towards and away from the coil repeatedly in the same type of motion that would be seen when the magnet(s) and coil were installed on the seismograph. Needless to say the range of motion was larger than would be experienced by the seismograph and precision of motion did not meet any scientific standards. With the single magnet I got induced voltages anywhere between 0.1 and 0.3 volts. With the two magnets I got induced voltages anywhere between 0.5 and 1.0 volts. This implies that the magnetic field strength is stronger, but the experiment was not rigorous. I can think of a couple of aspects where the experiment was not strictly an apples to apples comparison. Whether the magnetic field strength is stronger with two magnets or not, it might be possible to increase the output with a configuration of two magnets. I reviewed my university physics book last night, and the formula it gave showed that induced current was proportional to both the magnetic flux linkages AND the number of turns being cut by the magnetic field. The number of turns parameter is why we build these coils with 10,000 turns of #34 wire rather than 100 turns of #10 wire . So what does this have to do with using two magnets you ask? Well ... by using the two magnets the gap between poles can be larger, so you can modify the design of the coil to be wider thus putting more turns into the strongest part of the magnetic field. On the other hand, the magnetic field strength will also drop as the gap increases. The million dollar question is whether the field strength drops at the same rate the number of turns increases as the gap is widened. Anybody care to wind a few 10,000+ turn coils of various widths and experiment with this ? Ron Westfall westfall@...... _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ct@....... (Charlie Thompson) Subject: Fuss over magnet design for Lehman Date: Thu, 16 Jan 1997 15:02:13 -0600 I've been listening to the discussions regarding magnet designs for the Lehman. Several observations: 1. If you have low-noise gain in your preamp it can be used instead of huge magnets and coils. 2. Once you are clearly detecting the 6 second background noise you've reached the point of diminishing returns anyway. 3. I use a cheapie blue horse shoe magnet from Radio Shack and a 240VAC relay coil and easily detect the 6 sec. background noise of the earth using Larry's most excellent preamp-filter. You can view my humble design at http://www.onr.com/user/ct/ct.html 4. Big magnets and coils are only needed if you have preamps with high input-referred noise. I can see a small amount of preamp noise but it is dominated by 6 second seismic background noise. I have the gain on the preamp cranked pretty high. I routinely detect mag 5.0 quakes 1200 miles away with this setup. 5. The Lehman design is forgiving as long as you make a low-friction pivot, adjust the period to 10 seconds or better, use a good preamp with almost any coil-magnet combination, and make the baseplate out of "non-humidity absorbing" materials like steel or aluminum. If you use a wooden base you will have to re-center the pendulum frequently. Mine stays put (bugs and spiders notwithstanding!). Lastly, you MUST have a thermal air current management system like an insulated box with a heater in the top for conditions where the floor is warmer than the air. If you don't manage the air currents properly you will LOSE 10dB of signal-to-noise or more. Basically, nothing in my design resembles the original publications but the thing still works great! The best test of sensitivity is to stand about 5 feet from your sensor and shift your weight back and forth from one foot to the other. Even on a concrete slab you should be able to record your shift in weight. If you can see your own movement you are sensitive enough to start recording earthquakes! That's my 2 cents worth and I hope everyone is encouraged to finish their designs and start recording. have fun, Charlie Thompson Buda, Texas _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: Larry Cochrane Subject: Re: Force Feedback Accelerometer Finally Works Date: Thu, 16 Jan 1997 16:52:27 -0700 Hi Allan and Others, Glad to here you got you new sensor working! Hope to see some digital seismograms from your system soon. At 11:27 PM 1/15/97 -0500, you wrote: >Hello All, > I have described the basic design in some past postings to this group. >Its a crude, small device with a 4" long pendulum hanging vertically. The >natural frequency is approx 2 Hz. A chopped light beam using LEDs and CdS >cells connected to a difference amplifier. Pendulum movement of +/- .005" = >+/- 4.25 volts at the output pin of the .05 Hz low pass filter that provides >the velocity signal from the acceleration signal of the difference amp's >output. In the past I may have gotten close to success but my misconceptions >held me back especially when it came to pendulum damping. I was wondering why you are using such a low low-pass filter on your system? I think one of the reason you are/where not getting the P wave is because of this filter. I did a FFT on the P wave of several teleseismic events and I see a lot of energy between .05 (20 seconds) and .5 hz (2 seconds). > If anyone has another way that I can use to calibrate the dampening >signal PLEASE let me know. So if you are operating a S-G type seismometer >that has a healthy output signal and yet is insensitive to P wave arrivals >then you may be experiencing the same dampening problems. The method I use is a coil (I use a relay coil) temporarily placed near the pendulum's magnet and then run a wire from the coil into the other room where I have SDR running. You can connect a battery (or variable power supply) temporarily across the coil wires, or, charge up a electrolytic cap and then connect the charged cap to the coil. This method seems to work fine for both my SG and Lehman sensors. Regards, Larry Cochrane Redwood City, PSN _____________________________________________________________________ Public Seismic Network Mailing List (PSN-L) From: ACole65464@....... Subject: Re: Force Feedback Accelerometer Finally Works Date: Thu, 16 Jan 1997 23:33:26 -0500 (EST) Larry, To answer your question as to why I used a low pass filter for the accelerometer is as follows. The acceleration output signal from the displacement transducer needs to be integrated once to obtain a velocity signal. This integration could be done digitally with software, or electronically. I choose the later as I am not storing the data on a computer/data logger etc. The simplest way to integrate the signal is thru a low pass filter with the corner frequency set a .05 Hz. The integrated signal rolls off at 6dB per octave above this corner frequency. The instrument now gives a "flat velocity response" between .05 Hz to 2 Hz (the natural period of the pendulum). Acceleration signal appears below .05 Hz and the signal rolls off at 18 dB above 2 Hz. A simple first order filter provides this integration. I could have used a state variable filter with the center frequency set to .05Hz, at the bandpass output pick off the signal with anything above .05 Hz integrated, and at the same time I can pick off the signal at this filter's low pass output that doubely intergrates the acceleration signal, rolling off at 12dB per octave to provide a displacement signal, again "flat" between .05 - 2 Hz. I found that a true electronic integrator are difficult to keep stable and for my purposes the "filter" works fine. Using a "filter" requires some additional gain to boost the signal to see action up to 5 Hz but is no big deal. Thanks Larry for