Remember the high-altitude balloon project from last summer? We’ve finally achieved lift-off! The balloon team has completed the payload assembly, tested it on the ground, tested it tethered inside the warehouse, including radio datalink, and done a tethered test outside in the parking lot. Unexpectedly, they also did a “free-flight” test when the tether string broke, and they were able to see the balloon payload float off to parts unknown. We learned a bunch of things: get a stronger string; use a higher-powered radio link; be ready to track the system (even if you don’t think it’s going to fly away); and use cheaper batteries. Oh, well—that’s the way real science goes. The team is putting together a new payload/balloon/parachute and plans to do more tethered experiments and an actual, planned flight later in the summer.
With all we’ve learned with the Gen3 and Gen4 pixels, we’re ready to design our own single-board ERGO pixel that will have all the functionality we want. Simon is prototyping the system with a Texas Instruments development board, writing the code to make it dance, and designing a custom printed-circuit board (PCB) that puts everything, G-M detector and all, on a single board. The Gen5 system will incorporate an LCD touch-screen, so the operator won’t even need to connect a keyboard and monitor. Simon says he’ll have an operating prototype by summer’s end. More news to follow!
What if an ERGO pixel could hook to a monitor and keyboard? Our “Gen4” development project, led by Daniel, is working with the Gen3 team to create a pixel based on the popular Raspberry Pi linux microcomputer that works as an entire desktop computer, including access to the web by browser, easier user network set-up, and all kinds of other nifty things. Daniel is writing the Python code to take in data from the GPS, format it and send it to our database server. We’re in initial trials, and we’re starting to develop a special GPS + Geiger-Müller board to go with it (adapted from the Gen3 shield). We are targeting the ERGOberryPi as a large-volume production pixel with high functionality and extremely low cost. Daniel is also designing the system enclosure and desktop user interface.
Since the Raspberry Pi is such a popular platform, we’re planning to have the ERGO board available through commercial channels, so anyone can buy a Raspberry Pi and the ERGOberryPi board, build his own ERGOberryPi, send in a request for a pixel identity, and join the ERGO network.
Over the last few months, with the help of Jim Brown of seti.net, much of the code needed to create an “ERGOduino” has been completed, and we are in testing with a pixel based upon the EtherMega board a…nd our own custom “shield” that includes the u-blox GPS module, the Geiger- Müller detector, and the electronics to make it work. The entire pixel consists of an EtherMega board and our shield.
We’re calling this pixel “Gen3”, and Sloane has a team working to build ten of them. We’re planning to have those ten units completed and “in the field” around South Florida this summer. The team is also working on a rugged, instrument-grade enclosure and an AC power supply to complete the Gen3 system.
Simon has three ERGOduinos being tested online. An ERGOduino consists of an “Arduino Ethernet” board, a custom-built ERGO shield containing the GPS and detector-interface circuit, and a G-M detector. It’s a little, bitty thing, and looks like it will cost a fraction of the old ERGO beta pixels.
Right now, the ERGOduino replicates the function of the beta ERGO pixels except for datalogging. We’re going to move to an EtherMega board, a new version of Arduino, that will allow us some more code space to add us to cache and log data. You may notice the ERGOduinos (1111, 1112, 1113) coming and going from the map during the test period, but keep an eye out for them.
Meanwhile, Jim Brown is working on a next-generation ERGO pixel using an EtherMega that will work securely with our new database. Jim has solved a lot of problems and developed a lot of knowledge that’s helped us with the ERGOduino. This summer, we hope to have some students working on a Raspberry Pi ERGO pixel, and Simon is hoping to develop a totally custom board using an MSP430 or Stellaris processor and including an interactive LCD touchscreen. Lots of exciting developments!
It’ been a few months since the last newsletter, and we’d like to tell you some of the exciting things going on with ERGO–
ERGO Worldwide Leaders…
I often use the phrase, “You knew this job was dangerous when you took it.” One of the reasons we took on this work was to figure out how to engage a worldwide collaboration of students and teachers in a scientific project. We’ve found that there are all kinds of problems, but the fun lies in finding solutions. For example, did you know there are some places on the planet where you can’t just send boxes containing electronic instruments? We’ve found at least one country where the US Postal Service won’t send packages, and although Fedex accepted it, our package was returned by Pakistani customs office. There are probably many such places around the world, but right now we’re looking for someone who is traveling to Pakistan and would be willing to help a host there set up an ERGO pixel. In some areas setting up an ERGO pixel might come down to a personal connection with someone there.
As the network expands (we’ve built 119 ERGO pixels now, and over 110 of them have been sent out to hosts), we’ve found that it’s not always easy to set up and to keep a pixel on-line. People move, get promoted, graduate, and other things happen that result in an ERGO pixel either never going on-line or going off-line after it’s originally set up. The solution to these problems is communicating with the hosts, and we needed a way to “scale” beyond what one person could manage with emails and phone calls. So, we’ve divided the world into five areas, and we’ve got volunteers to be leaders for those areas:
• South Florida (we have a lot of pixels in a cluster here)—Sloane Rice
• Americas (outside S. Florida)—Jordi Orbay
• Eurasia—Alan Garrett
• Middle East—Alan Garrett
• Asia-Pacific—Alan Garrett
• Africa (well, we’re hoping to get some pixels there, but we need some help)
Sloane, Jordi, and Alan have started to make connections with all our hosts in their areas, and we’re starting to see results as pixels come on-line.
ERGOduino, the Do-It-Yourself Pixel
Last summer we realized it was time to design a new, much-less-expensive pixel. What if we could make it “open-sourced,” so anyone could build their own pixel? We looked around to see what microcontroller architectures were available, and we decided on the Arduino platform (specifically, the Arduino Ethernet). The Arduino community is a vast, creative group of people who build and program electronic control systems for all kinds of things, from cat feeders to MP3 players. The simplest Arduino costs under $US 20, and the Arduino Ethernet, which includes micro-SD storage and Ethernet connectivity, costs around $US 60. So, you can buy an Arduino Ethernet board, connect a GPS receiver and muon detector, add some software, and you have an ERGO pixel! Since most things designed for Arduino are named something followed by “uino,” we’re calling it the ERGOduino.
In order to simplify adding the GPS and detector we have designed a “shield” (an add-on board for the Arduino) that includes everything you need except the detector. The shield includes a Radionova GPS 12-channel receiver (with built-in nanosecond-level timestamping!) and circuitry that can accept signals from just about any kind of detector you could imagine. The total cost of parts for the shield is around $30. The Radionova GPS has its own built-in antenna, but you can connect an external antenna if that works better in your installation.
We have the shield hardware pretty well developed, but that leaves one important piece: the software. Jim Brown of seti.net has been working on that part, and he’s very near to a solution. The software has to read in the GPS location and timestamp data from the Radionova GPS when an event occurs and format the data properly for “posting” to our database server. The database has been moved to our own dedicated server at a co-location facility, and it’s being administered by James Riley at betaforce.com. We look forward to trial operation of the ERGOduino very soon.
If you’re interested in helping develop the ERGOduino, let us know. A follow-on project might be a similar system based on the Raspberry Pi, a very powerful, cheap, and versatile new open-source platform.
The ERGOnaut—Our Deep-sea Diver
You might have been following our experiment to place an ERGO pixel very deep underwater. Last November we took the ERGOnaut 1 on our research ship, Patent Pending, for its first sea trial. Our plan was to drop it on a tether to -850 meters, but we found the seas too rough on the day we sailed, so we re-located to a better-protected site just east of Elliot Key with a depth of 150 meters. Well, it’s still there. Our winch control failed, and the ERGOnaut 1 dropped all the way to the bottom, where it got snagged. All our efforts to pull it loose failed, so that’s where it lies. If anyone wants to dive to 150 meters to retrieve it, we’d sure like to see the data recorded on the internal memory card and the video from its on-board camera. But, unfortunately, that depth is just beyond where scuba divers can go. Does anyone have an ROV that can go to that depth?
But, we never give up, so we’re planning to build a second diver, the “ERGOnaut 2,” to try it out this summer. Stay tuned….
A Few More Beta-4 Pixels
Following a mention of ERGO in the Bad Astronomy blog we received many inquiries from people who wanted to participate in the project, but we were running out of the initial Beta-4 ERGO pixels. While we are developing the ERGOduino low-cost pixel, we decided to build ten more Beta-4 pixels to meet the demand. The ten circuit boards have been fabbed, five have been built, and three have been sent out to Norway, New Mexico, and New Jersey. Look for them on the ERGO map. We have five Beta-4 pixels remaining to build, and we hope to have the ERGOduino ready by the time those run out.
Cosmic Rays and Lightning
One of our high-school students, Troy Gonzalez, has been working on our project to see if we can establish a correlation between cosmic-ray air showers and the initiation of lightning bolts. Atmospheric physicists think the voltage gradients in storm clouds, even though they are millions of volts, aren’t great enough to initiate a lightning discharge hundreds of meters in length. Various theories have been advanced, but the leading one is that when the electric field becomes strong enough at a particular space where a cosmic-ray air shower occurs, the ionized paths created by the passage of cosmic-ray muons through the atmosphere form a path for the initial lightning discharge. The process is known as “runaway breakdown” (see the article in Physics Today: “Runaway Breakdown and the Mysteries of Lightning”.
Troy has completed building five lightning-detection circuits, which we have married to five ERGO timestampers (basically, an ERGO pixel without the Geiger-Müller detector). Lightning discharges send out radio impulses, and Troy’s detectors pick up those impulses to trigger an ERGO pixel, which reports the exact time and location where the impulse was received. Once we have the five detectors installed around South Florida, we’ll be able to determine the exact time and location of lightning discharges by a process called Time-delay-of-arrival Multilateration. Then, it’s just a matter of waiting for some lightning bolts to appear in close proximity to some of our ERGO cosmic ray detectors and see if there were cosmic-ray air showers detected at those times and locations. If we can do it, and if we can establish a statistically-significant correlation in space and time between lightning discharges and cosmic rays, it will be new science.
Troy has applied to take his project to the International Science and Engineering Fair, so wish him luck!
All the Beta-4 ERGO pixels have incorporated a Geiger-Müller (G-M) tube muon detector. Although G-M detectors are inexpensive and reliable, there are two problems with that choice. First, they are physically small, which gives them a very small “window” that is sensitive to muons. It would be great to have a larger detector area in order to catch more muons. Second, it would be better for ERGO pixels to have a detector that is sensitive to high-energy cosmic muons but not sensitive to the lower-energy terrestrial background radiation. That way, we would be counting only cosmic-ray events.
Early in the project we tried to develop “resistive-plate chamber” (RPC) detectors, such as those used at accelerator labs, but we were unable to produce them reliably. A new type of detector, the Gas Electron Multiplier (GEM) has been developed recently by scientists at the Large Hadron Collider, and it could possibly be a good path to a larger, muon-only detector for ERGO. Does anyone want to try to make one? Here’s a reference: GEM Detectors.
Simon Tsaoussis has been working on another type of detector, using a plastic scintillator (available cheaply as surplus) and a PIN diode (a sensitive kind of photodiode). Simon is working on prototyping one on a tiny circuit board. He wants to build it into a cellphone!
And That’s the Way It Is—
Thanks to all the interns and volunteers for all the work they have done. I think you’ll agree, if you’ve read through this long-winded newsletter, that we’ve made a lot of progress!
We built up the ERGOnaut diving instrument with a housing made from 12mm-thick 250mm-diameter aluminum tubing, a 25mm-thick aluminum plate on one end, and a 100mm-thick clear acrylic (plexiglas) cover on the other end. We went with the clear window so we could have a camera inside to take pictures during the dive. Lighting was provided by an external underwater LED flashlight filled with mineral oil (to prevent collapse of its housing). Battery power was from a 4000 mA-Hr 11.1V LiPo battery, which can run the ERGO for 11 hours.
During the dive, energetic-particle event data would be recorded on the micro-SD card of the datalogger, so we could analyze the data when the ERGOnaut was brought back up to the surface. The Geiger-Muller detector and ERGO timestamp generator board were exactly the same as the 115 units we have sent out around the world.
Our plan was to drop the ERGOnaut on a 200-lb-test fishing line (spectra fiber) to a depth of 2580 feet, approximately 30 miles east of Miami. When we got to that location, we found the seas too rough to attempt the drop (8-10ft seas!), so we relocated to a calmer location east of Elliot Key (latitude 25.50981889N; longitude 80.06057306W, 470 ft depth). We put the ERGOnaut into the water at that location, but a failure in the winching system caused the instrument to descend to the bottom rather than hovering above the bottom. When it got to the bottom it became snagged on something, and none of our efforts to dislodge it were successful. Our last-ditch attempt was to pull on the line as hard as possible; the ERGOnaut didn’t budge, and the line broke.
2. Make it big enough for the circuit boards to go in edgewise (parallel to the long axis of the cylinder) rather than flat, so the status indicator lights are visible
3. Completely seal up the diver prior to the voyage and use the bluetooth link (through the acrylic cover) to verify and record the ERGO data until the moment of release.
4. Mount the weights (30 pounds seems fine) on the bottom with a line much weaker than the lift line (50-60 pound test). Make the connection between the ERGOnaut and the weights out of something that will dissolve or otherwise separate in a few hours to assure the ERGOnaut returns to the surface.
5. Never allow the diver to get close to the bottom
6. Use an entirely different kind of winch, one that’s able to hold a cable that will lift 500 pounds. The electric reel that we used on this dive was just big enough to hold 4500 feet of very thin 200-pound-test spectra line; the new winch will be sized to hold 5km of 500-pound-test stainless steel cable.
7. Try a much shallower dive to a depth that we can dive to with scuba before going to greater depths, then work our way up to 500ft, 2500ft, and finally 15,000 feet (east of the Bahamas, in the “tongue of the ocean”).
So, we’ll start building ERGOnaut II soon. It will have a longer cylindrical section to allow the boards to mount on end rather than flat, and so that it will have some positive buoyancy. We’re looking for ideas on making the connection between ERGOnaut and the weights so that the weights will automatically drop in three hours or so.
The first outing for the deep-sea ERGOnaut cosmic ray detector will be Saturday, Nov 24. We’ll be taking the research vessel out into the Atlantic Ocean to a depth of over 8,000 feet, and we’ll drop the high-pressure-sealed ERGOnaut capsule down to 2,500 meters depth, if everything works as planned.
At that depth only a few cosmic-ray daughter particles (mostly muons) will penetrate. During the mission we will be continuously recording detected events, and we’ll publish the results as soon as we have them.
This is the working version of the Arduino shield which includes the GPS receiver, timestamping, and signal conditioning for a Geiger-Muller detector. Plug this onto an Arduino, add a detector and software, and you have an ERGO Beta5. On the way to 1,000 units…