1

I'm a seismic geophysicist and would like to attach a geophone - which is basically a transducer - to any type of RP. The types and costs of geophones are both variable and extensive (in some cases). The main idea to note is this: a geophone (or array of geophones, perhaps) usually records ground displacement in 1-, 2-, or 3-D and sends the signal off as either an analogue or digital output. I want to record that signal in real-time via a RP and (eventually) develop a pipeline to send that data to a cloud database.

How might I do this?

nate
  • 111
  • 4
  • 1
    Hi @Nathan Benton, Welcome and nice to meet you. Your question is interesting. So I wikied to know more: Geophone Construction - Wikipedia (1) Geophones have historically been passive analog devices and typically comprise a spring-mounted wire coil moving within the field of a case-mounted permanent magnet to generate an electrical signal. (2) Recent designs have been based on microelectromechanical systems (MEMS) technology which generates an electrical response to ground motion through an active feedback circuit to maintain the position of a small piece of silicon. / to continue, ... – tlfong01 Nov 20 '19 at 14:15
  • (3) The response of a coil/magnet geophone is proportional to ground velocity, while MEMS devices usually respond proportional to acceleration. MEMS have a much higher noise level (50 dB velocity higher) than geophones and can only be used in strong motion or active seismic applications. / to continue, ... – tlfong01 Nov 20 '19 at 14:16
  • I am not too sure if what the wikipedia says is correct and update: MEMS has a higher noise level (50 dB velocity higher). I think we can google "geophone" "MEMS" to get some updated info. I read that there are many methods to filter the noise, say using digital filtering and other methods of DSP (Digital Signal Processing). I read wiki's sections on construction and frequency and have the feeling that they were a bit out of date, and perhaps not compiled by digital signal processing experts. So there is hope that traditional "three spring coil" system is not that superior than MEMS. .. – tlfong01 Nov 20 '19 at 14:26
  • A couple of weeks ago I read in this forum about vibration detection sensors which I remember is also using " coil springs" which I think can be very accurate (because IT advances too fast that the domain experts expect. / to continue, ... – tlfong01 Nov 20 '19 at 14:29
  • You might read my question to this question on 801 vibration sensor. I happen to have a couple of such sensors in hand. There are cheap, but as other MEMS sensors using in smart phones, because of huge mass production, become very powerful. An example is the NASA space quality of miniature camera, using ZEISS lens. Other examples include very high quality audio micro phone, with professional noise filtering (OK not very loud, but very HiFI) https://raspberrypi.stackexchange.com/questions/104752/how-can-rpi3b-read-the-output-of-the-801s-vibration-sensor-module. – tlfong01 Nov 20 '19 at 14:37
  • In short, I think MEMS and smart phone technology, together with real time cloud big data base data and AI neural network non algorithmic pattern recognition should replace mechanic spring coil stuff. I do know that modern air carriers (A300!) are still using mechanical gyros. But at the same time military drones do use MEMS gyros/accelero/magneto/gps. I used to think that mechanical watches are replaced by very many times more accurate digital watches, which in turn replaced by smart phones. I am just think aloud some brainstorming ideas. Sorry for typos and unorganized thoughts. – tlfong01 Nov 20 '19 at 14:48
  • 1
    And in case what the Wikipedia says is true, ie mechanical coil spring sensors have low noise level compared to piezo MEMS gyro/accelero, we can perhaps use a hybrid approach, combining the cheapie but miniature 801 type sensor, augmented with more noisy MEMS but digitally filtered. https://imgur.com/gallery/leMnH1c. – tlfong01 Nov 20 '19 at 14:58
  • Thanks for the insight @tlfong01. For what this is worth, I'm looking to detect and analyze seismic wave-forms (specifically surface-waves) that usually have comparatively long periods associated with them (e.g. 5 - 20 seconds). Could these sensors you refer to manage that kind of low-frequency sensitivity? Also, I'd really like to know more - if possible - of HOW to connect a geophone (or a similar sensor device) to a RP. – nate Nov 20 '19 at 15:08
  • *** Errata ***: I forgot that my answer about 801 vibration sensor has been deleted. This is a copy the essential parts of the answer about the vibration sensor I mentioned in the answer of vibration: https://penzu.com/p/484bcd9a. – tlfong01 Nov 20 '19 at 15:16
  • 1
    Hi @Nathan Benton, Ah yes, you are right in pointing out that your signal frequency is VERY LOW 5~20 second period (of order 0.01 Hz). I happened to once played with low frequency brain wave (Alpha Beta etc), of range 1 to 20 Hz. Of course you frequency range is one order smaller. My brainstorming idea is to use "long mechanical coil springs or flexible blades" perhaps 12' long, with MEMS gyro/accelero attached at the end of the long spring/blade. A similar method is to use sort of long arm pendulums, also with MEMS gyro/acceleros at the end. Bed time. See you later. Cheers. – tlfong01 Nov 20 '19 at 15:26
  • This morning I skimmed some more references and found the project seems not so complicated as I expected, for the reasons I would summarize later:: (1) Geophone - Wikipedia https://en.wikipedia.org/wiki/Geophone

    (2) Seismic Data - Marine and Petroleum Geology, 2013 https://www.sciencedirect.com/topics/earth-and-planetary-sciences/geophone/pdf

    (3) Learn more about Seismic Wave https://www.sciencedirect.com/topics/earth-and-planetary-sciences/seismic-wave

     – tlfong01 Nov 21 '19 at 01:26
  • You might like to read some of my posts related to low frequency wave analysis, including EEG and ECG: How can Rpi read data from a heart rate sensor module, such as AD8232? (1) https://raspberrypi.stackexchange.com/questions/98416/can-raspberry-pi-work-exactly-like-arduino-uploding-arduino-codes-into-raspberr,

    (2) ECG Graph https://imgur.com/gallery/zbxx5cx.

     – tlfong01 Nov 21 '19 at 01:36
  • This is another project on low frequency (ECG) data acquisition noise filtering problem: (1) "Rp3B+ MCP3008 / AD8232 ECG Data Acquisition Noise Problem": https://raspberrypi.stackexchange.com/questions/96482/raspberry-pi-3b-interface-with-heart-monitor-ad8232-using-mcp3008-and-spi-pins/96498#96498 – tlfong01 Nov 21 '19 at 01:42
  • And my Micky Mouse project of EEG analysis: (1) EEG Data Analysis: https://www.raspberrypi.org/forums/viewtopic.php?f=37&t=229740&p=1408920&hilit=weight+sensor+tlfong01#p1408963. – tlfong01 Nov 21 '19 at 01:52
  • 1
    There's stuff from 2012 onwards on seismography at https://www.raspberrypi.org/forums/viewtopic.php?f=37&t=3731 – Dougie Nov 21 '19 at 08:09

1 Answers1

0

After reviewing the material provided here and conducting more research, I've decided that - as a benchmark for future advances in availability and especially price - this option is a great place to start resolving the issue/question posed here.

nate
  • 111
  • 4
  • That product looks promising. One thing to note about the RasPi is that it doesn't have an onboard ADC, so any analog transducer will need some peripheral circuitry. – BobT Dec 03 '19 at 14:34