Tag Archives: infrasound

The Infrasound Bucket - Part 1 - Hardware

I'd like to write a short series of posts describing my setup of the infrasound unit I've written about before.  This is the same unit we used to look at traveling acoustic energy from the Russian meteorite and will soon use to examine earthquakes! Placing the unit inside my office or even inside the apartment proved to be very noisy as I saw every time someone opened or closed a door!  The makers (Infiltec) suggested that I put it outside, maybe in a drink cooler to shield it from the weather.  I did exactly that (photos below), but the cooler turned out to not be water proof and had about 2 cm water standing in the bottom when I checked it after a small storm.  The data quality while the instrument was outside was amazing though, with seismic signals coming through very clearly.  It was time to design a new system that would: 1) Be safe to leave outside in the weather, 2) Not have thick data cables running inside to a computer, 3) Would not require an inside computer, and 4) Would automatically post the current data online.

For the first post we're going to talk about the casing setup and mounting of all the vital hardware.  One weekend we decided to go wandering about Home Depot to find a suitable shell for the instrument as well as pickup a few other essential supplies.  Lendi had the flash of inspiration that we should use a 5-gallon plastic bucket... the ones at the Home Depot "Homer's All Purpose Bucket" even have an O-ring seal on the lid.  Perfect.

A built in O-ring seal on the bucket.

Now to figure out how to hold the hardware up off the bottom of the bucket.  In an ideal world this isn't needed, but in reality water may get in and I don't want it covering electronics thrown in the bottom of the bucket.  We used 1/4" plywood cut to a keystone shape that just fits the vertical profile of the bucket.  Adding two "L" brackets from the shelving section meant for ~$15 we had the shell and left over plywood.
Test fitting the plywood into the bucket.  Notice the cooler in the background that formerly housed the instrument.

I bolted the infrasound unit to the wood by using "plumber's tape" or metal strap with holes down its length.  This isn't the most elegant solution, but it meant no drilling the infrasound case which is semi-sealed on its own.  It is also very easy to get the unit out for any maintenance.   My RaspberryPi ended up having problems on the circuit board, so I've bolted a Beagle Bone Black to the board as well.

Front of the mounting board.  Infrasound unit (right), Beagle Bone (left), and power plugs (top left).

Rear of the mounting board with power passthrough.  

With no tall standoffs handy I made use of locking nuts, washers, and other assorted 4-40 hardware.

Two holes were drilled in the side of the bucket: one for the power and one for the air tube to the infrasound instrument.  I passed the power cable through (outdoor zip cord) through as well as clear plastic tubing and sealed it with bathroom silicon sealant.  I'd recommend sealing on the inside and outside of the bucket bulkhead.  Make sure to leave extra cable and tube for drip loops. A drip loop like structure was fashioned on the outside of the bucket to ensure no rain would blow up the tube into the unit.  We taped the tube down and then ran beads of silicon to secure it to the bucket.  After the sealant dried we moved the tape and secured the rest of the tubing.

Power and air tube sealed into the bucket and loops to prevent water flow.

Inside the bucket: notice the power plug.

In later posts we'll talk about how the power is actually provided and such, but the part that pertains to the hardware is the mounting of two binding posts on the plywood at the standard 3/4" spacing.  This allows us to power the board from a banana jack on the bench for testing or operationally in the bucket.  I drilled a passthrough hole to send power from the back of the jacks to the front of the panel.

Initially I built a 5V regulator to power the computer with from an LM7805 linear voltage regulator, but this was indeed a poor choice.  Even with a decent heat sink, the chip still got blistering hot when I was drawing 700mA (of the 1000mA rated power).  Considering this would be outside in the summer heat and the fact that I didn't want the failure point of a mechanical fan I decided to use a buck voltage converter.  Linear regulators dissipate all extra power as heat.  For example: I was feeding 12VDC to the converter with a 700mA load running at 5VDC.  That means that (12V-5V)*0.7 = 4.9 Watts of power was being turned into waste heat! No wonder, remember we think of watts as energy/time (Joules/second actually).  That's a lot of wasted electricity and really just a complication to our design, but it was very clean power.

The old linear regulator.  It's now awaiting a new use in the parts bin.

The buck converter is a switching type regulator.  I don't want to get into how switching regulators work current, but it's an interesting topic and you should have a read on the theory if you like.  I bought a small unit (P/N 1385) from Adafruit that is rated to 3A (though it gets warm there).  The power isn't quite as clean from this switching supply, but it's fine for out use here.  It works great with the Beagle Bone and provides lots of extra power for 5V accessories.  Don't want to order and ship from Adafruit? You can get the exact same thing from a model shop.  They are called "battery eliminator circuits" and allow modelers to plug their airplane, car, etc servo electronics (5VDC operation) into a 12V battery they already have in their kit.  Just clip the 3 pin servo plug off the end and you are ready to go.  Don't forget good soldering practice and to use heatshrink tube! Shorts could spark a fire, which we don't want.

The "battery eliminator circuit" or my 5V buck converter to supply 5VDC to the Beagle Bone.

So there it is! Next time I'm going to talk about setting up the power and network infrastructure.  Maybe even the serial communications! We're going to try to avoid using a serial-USB converter since the Bagle Bone has only one USB port (that I'm using for a WiFi adapter), I don't want to use a hub, and it's a chance to learn about signal level shifting and wire into that temping header on the board.

Everything fit into the bucket nicely and powers up from the bench power supply.

Chelyabinsk Meteorite - Infrasound, Seismic, and Satellites oh my!

Just as Earth was about to have a close encounter with asteroid 2012 DA14, the people of Chelyabinsk, Russia had a personal experience.  Before we talk about both 2012 DA14 and the Chelyabinsk event some terminology needs to be set out.  A meteoroid is a small chunk of debris in space, generally anything from a fleck of dust to a small boulder.  A larger space bit of debris is termed an asteroid.  A meteor is when some of this debris enters our atmosphere, heating up due to friction.  A meteor is called a meteorite if it actually reaches the surface of the Earth and survives impact.  Everyday we are pelted with many tiny meteors, but few reach the surface.  Most meteorites are never discovered as they are statistically much more likely to land in the ocean due to it's coverage of Earth's surface.  Sometimes meteorites are found on land, in fact it is common for scientists to go to Antarctica to look for the dark rocks on the surface of a white sheet of ice.  There are many pages on hunting meteorites  and a book as well, it's worth reading about if your curious how we find rocks that landed a long time ago.

It's worth saying that there are different kinds of space debris, some more stony, some made of almost solid metal, and some of ice.  While it's worth discussing, I'd rather focus on the current events in this post, so if you are curious there is a nice page at geology.com that gives the basics.

To begin, lets talk about 2012 DA14, or the non-intuitive name that we gave a near Earth asteroid that is about 160 feet in diameter and weighs a massive 190,000 metric tons.  This asteroid could do some serious damage and was scheduled to have a close call with Earth on February 15th.  How close? Well, it would be about 17,200 miles from the surface, which seems like a long way.  It's not.  The moon is 250,000 miles away (roughly) and we've been there and back in a matter of a few days.  In fact, the geosynchronous satellites that beam TV and weather data down to Earth orbit about 22,236 miles above the surface to rotate at the same rate the Earth does.  As shown in the figure below, 2012 DA14 passed between us and the geostationary satellite band; a very close call.

Why talk about 2012 DA14 in a post about a meteor over Russia? To say they are not related in any way.  They approached from entirely different directions and it just happened to be a coincidence of space and time.

Now for the event in Russia.  At 3:20:26 UTC on Feburary 15th a large meteor about the size of a schoolbus entered the atmosphere.  The 49-55 foot estimated diameter object probably weighed about 7000-10000 tons.  While heating up upon atmospheric entry the meteor "detonated" or exploded in mid-air.  This has happened before, a list of historic airbursts can be found here.  The most famous being a large explosion (also over Russia) in 1908 called the Tunguska event.  That explosion released the energy of 10-15 million tons of TNT, leveling forests and destroying an area of about 830 square miles.  The event that just occurred was about 500 kilotons of TNT equivalent, or roughly 20 times smaller.  Shock waves from the event still managed to send around 1500 people to local hospitals with shards of glass and building materials in their faces/skin from rushing to a window too see what was happening.  Videos of the entry are all over the web, in several you can hear the detonation and shock wave.

So how do we know so much about this object considering we didn't know anything about it until it exploded overhead? Well, remote sensing helps us.  When a meteor entered over Wisconsin in 2010, I wrote about following the trail on the US Doppler Radar Network (here).  This time we could see the meteor from weather satellites (Meteosat 10 image below) as well as on seismic and infrasound stations.  Another meteosat also captured several frames that have been made into a video here.  Current estimates of the entry speed are in the area of 40,000 mph with a very shallow entry angle.

First the seismic observations.  So far I've seen reports of the Borovoye, Kazakhstan station seeing a gorgeous signal (thanks to Luke Zoet on this one). The station details, and even a photo are available at the USGS network operations page.  Below is a filtered (0.15 Hz low-pass) seismogram from BRVK.  This would be a result of the shock wave rattling the ground and seismic station.

Next, and rather interesting, are the infrasound observations.  Infrasound is very low frequency sound (below 20Hz) that we can't hear, but can record as air pressure variations.  It so happens that Steve Piltz of the Tulsa National Weather Service has a microbarograph.  Upon seeing his data from an earlier earthquake (yes, ground movement causes air pressure waves), I immediately bought a unit from Infiltec and set it up in the office at Penn State.  Below is a picture of the station.

Infrasound propagation is incredibly complex and difficult to predict over such long distances, so I've done a simple calculation that is very likely going to be revised upon some discussions with seismologists this week.  First, I wanted to know how long the sound would have to travel.  To find the shortest travel distance between a latitude and longitude set you can assume a spherical Earth (not too bad for such a back of the envelope calculation) and some math.  Remember trigonometry? Well when it's modified to work on a sphere instead of in a plane it's creatively called 'spherical trigonometry' and consists of a strange function called the haversin.  If you are curious about how I calculated the travel path of the sound waves checkout the wikipedia page on the Haversine formula, but I've included the formula below.  Below is the result of the calculation, a great circle path between Chelyabinsk and State College, PA.
d = 2 r \arcsin\left(\sqrt{\operatorname{haversin}(\phi_2 - \phi_1) + \cos(\phi_1) \cos(\phi_2)\operatorname{haversin}(\lambda_2-\lambda_1)}\right)

The distance the wave would travel would be something like 8670 kilometers.  Sound travels at 340.29 m/s at sea level, but since we're making assumptions we'll say 300 m/s is a nice number.  So the wave would take somewhere in the 7-8 hour range to reach State College (assuming it's non-dispersive and many other likely not so great assumptions).  Luckily for us, the event and the arrivals are overnight.  During the day my infrasound station is swamped by signals from office doors opening and closing amongst other things.  The meteor entered the atmosphere at 10:20 pm local time, so I've plotted the infrasound from 10 minutes before the meteor entered to well after the energy should arrive.  There is a large increase in the noise shortly after entry, but this is too soon.  Could it be seismic energy or arrival of a faster shock path? Maybe, that's a point for some discussion and revision later.  The big thing to notice is the noise increase at about 7-8 hours after the entry.  It's still early in the morning, so it is doubtful that this is people coming into work.

I've made a .SAC (seismic analysis code) file of the raw data for about 24 hours around the event available to download here.  Download the file (~19Mb) and play with it! The data is collected at 50Hz, but all that is in the meta-data (as well as location details).  I use ObsPy to do most of my analysis in Python, but you could use SAC or other codes meant for seismic event analysis.

Steve's station in Oklahoma recorded similar signatures.  His data over a slightly shorter time span (5:21-11:42 UTC) is below, showing similar signatures.

Infrasound is actually what allows us to determine the energy release from the explosion.  As it turns out seismic stations and infrasound have been used to monitor nuclear testing for years (a relevant topic currently considering the recent tests by North Korea).

Overall, I'd say stay tuned for any updates.  Eventually the infrasound station I have will be setup for live streaming.  I'm sure after discussion with some folks more versed in infrasound travel we can clean up the data and maybe do some more back of the envelope energy/rate calculations for demonstation purposes.