EDIT: there is a dedicated blog on this project, here. Make sure to follow it for latest news.
A few days ago I announced my intention of creating my first crowd-funded project, but I didn't tell you what was it about. Well, I'm planning quite a challenging project - to bring my uRADMonitor to a global scale, and create a network of distributed detectors, with data centralized on one server, so everyone, regardless of location can access radiation related data. With crowd funding, we'll be able to produce enough of the uRADMonitor units to cover the most important cities and locations, so let's make this upcoming Crowd funded project a success!
Summary and updates
July 8, 2013: Intro and General specs
The final device should be able to run both as a mobile unit (on batteries) or as a fixed monitoring station (with a waterproof quality aluminum case and connected directly to your home internet router). Besides reading the Gamma radiation levels, it would come with additional sensors, for measuring temperature, humidity, barometric pressure, sky luminosity. To map the data to the real geographic location, the device will have GPS capabilities combined with IP localization (when connected to the internet). For offline use, as a mobile detector, the data would be logged on an internal SDCard, and the content uploaded to the server automatically as soon as a network connection becomes available.
July 9, 2013: The Crowd-funded project
My plan is to finish the final prototype and start my crowd funded campaign (Kickstarter, Indiegogo, Technofunding, etc) where I'm counting on all the support I can get, including yours, readers of my blog! Up to this point, using my private resources, I've been able to fund my own research and prototype developed, and I will continue to do so, but additional funding can help speed up the process.
The idea behind this project is to help people gain access to radiation measurement devices but on a global scale, by providing the uRADMonitor devices that anyone can install simply by plugging in the internet router network cable and a DC power supply. The data is then centralized, and the server located at www.uradmonitor.com allows free access to the data collected. With a few clicks you'll be able to see the levels in your city, or any other place. So easy!
The device also works as a data logger, especially when used as a mobile dosimeter, storing on an internal SDCard, GPS location and the environmental readings for Gamma Radiation background, temperature, humidity, pressure and luminosity.
All you need to do is get your own uRADMonitor device, place it outside or use it as a mobile detector, and the data is logged and sent to the server as soon as you connect it to the Internet. Then in front of the computer see your own readings or data collected by others. Not only a Geiger counter, but a powerful environment surveillance tool that builds a network of readings mapped to the entire Earth!
I decided to make the construction log publicly available, so others can see the story of what it takes to make this project happen:
July 10, 2013: The first prototype
I wrote the software to take the maximum out of the 84x48 display: special characters have been added (defined with 0 and 1 like in the the old BASIC days) for micro symbol, CPM, radiation sign or battery icons. Printing text, there's a total space of 14x6 characters: first I show the internal time (hh:mm:ss), the battery level (4 indicators for charging, full battery, discharged and empty), the accumulated dose registered since the device was powered on (measured in micro,mili or Sieverts, depending on the value size), then there's the battery value in volts, under the battery charge symbol. The third line shows the equivalent dose (measured in micro,mili or Sieverts per hour, again, depending on value size, the software does the switch automatically), a text description of radition level with 4 states (LOW, NORMAL, HIGH, DANGER) set to <0.10uSv/h, <0.20uSv/h, <0.30uSv/h and more than 0.30uSv/h for danger indicator. Finally the last line shows Geiger tube voltage (regulated to a constant of 400V +-5V), the inverter duty cycle (aprox. 34%) and the CPM, or counts per minute (measured as CPM, Kilo CPM or MegaCPM, depending on value size).
July 12, 2013: Ra-226 source test video
Here is a video showing a high count detection test, using the vintage 1B24 TR Switch radar tube, containing a small amount of Ra-226:
July 15, 2013: Geiger tubes comparison
Having two devices in one - portable dosimeter and fixed monitoring station, certainly complicates things. The former needs a battery, the station will go for DC source. The portable device must be small, the latter can be bigger in size. Is this important? Well, let's take the size into account: to be convenient for the user to carry it, the portable dosimeter should be no longer than 110mm and fit the black case, as in the picture above.
This implies using an appropriate Geiger tube, since these guys are coming in given sizes, and must have everything else match them. A good choice would be the SBM-20, but this is less sensitive than the bigger SBM-19, used in my first uRADMonitor variant . Background counts for SBM-19 is at 80cpm, while the SBM-20 barely touches 20. For short term readings this might result in unwanted variations from the true environment radiation dose, but would not be an issue when integrating the data over a larger time interval (a few hours or so). SBM-19 is 19cm in length, while the SBM-20 is only 11cm:
July 28, 2013: Designing the tiny 400V inverter
I've been busy designing the next prototype. The target is to make the circuit small enough to fit the aluminum case I opted for. I know that this is just another constraint that complicates things, and that the case should be amongst the last things to opt for, according to the final board, however as stated previously, this device should also be portable - this is why the case should have a reasonable size as well:
The circuit above (prototype 1) fits nicely, but it doesn't leave much space for batteries, GPS, Ethernet module and the SDCard storage module. To minimize, this new board will be created with SMD technology.
Going for SMD, everything is nice and easy, except the 400V inverter (for the Geiger tube). Instead of using the ferrite transformer, I have to design a boost converter (step up converter) that uses a choke! Just compare the pics below to get the idea:
The choke, visible in the second photo is the tiny coil between the two black transistors placed on the breadboard. With this setup I am trying to get something close to 400V from only 3V of input voltage.
The simple Boost converter setup doesn't provide enough output voltage (and it's not a big surprise as according to the literature, a certain maximum transformation factor is indicated). The simple boost converter (VAR-1) will only produce something close to 100V only, so far from the 400V required for the Geiger tube. But with the second variant (VAR-2), that uses a second transistor (T2) to discharge the base of the main switching transistor (T1), I got better results:
But the discharge transistor (T2) also changed the shape of the PWM signal considerably. I can no longer rely on the original PWM oscillation parameters . Normally a frequency of 10-20kHz is perfect for the ferrite cores I use (both chokes and transformers). To exit the audible spectrum completely (and make sure the device doesn't produce annoying ultrasonic sounds), I usually go for a minimum of 17KHz. The signal looks clean, a nice rectangular signal as expected. But with the VAR-2 modification, things change. Here are the oscillograms, measured in point 1 , 2 and 3 (points indicated on the schematic):
So instead of setting the microcontroller code to generate a 17KHz signal, I will probably go for strange values like 100KHz, actually resulting in a nice output voltage that is close to 400V. The last image already shows the CH2 (in cyan) reaching the Geiger tube voltage interval (388V, more than enough to run the SBM20).
This was a minor change in the booster converter topology resulting in higher voltage output. I also had to use a different T1 switching transistor (this is highly critical), as not any would do. Finally it now works, time to finish this second prototype, but this time with SMD components.
July 30, 2013: Wireless connectivity
Several people suggested that adding WLAN/Wifi capabilities to the uRADMonitor would be a great feature to have. And indeed, a data radio link is preferred over having to install Ethernet/UTP cables all around the place, or to have to take them outside or where the monitor is to be mounted. There are many obvious advantages, but as expected, there are also a few issues:
- a WLAN 802.11b/g Module consumes additional power, where we are already at the limit having so many modules and features
- Configuring the WLAN link would require data input at least for the security key (WEP/WPA): a minimal alphanumeric keyboard, or a bluetooth link to a smartphone - but this over-complicates the entire project.
A better alternative seems using a RF transceiver or even a simple RF module pair, like the FA1000A, presented here.
The uRADMonitor would send the data wirelessly over the RF link, to a receiver connected by cable to the internet router. So instead of having to install two cables, there could be a single power cable, and a secondary module to act as radio receiver/gateway and redirect the data to the internet router.
August 01, 2013: Offline SDCard logger
For data logging I've considered using either an EEPROM module or a SDCard. The latter has many advantages like huge memory capacity and compatibility with PCs, so you would be able to take it out of the uRADMonitor and see the data on your PC.
I did some progress on this, and the SDCard module is ready to accept sensor data (radiation/gps data logger for portable use). More on SDCards and microcontrollers, here.
September 3, 2013: SBM-20 Geiger tubes
I had some very nice progress developing the next prototype. After waiting for some modules to arrive from China, they finally did, and also the first batch of Geiger tubes, of type SBM-20:
October 3, 2013: Three variants: A, B and C
I have decided to create three different uRADMonitor devices, that will be part of a single, global radiation surveillance network. From simpler to more complex these are:
- Model A: a small compact detector, this device contains a Geiger counter (SBM-20), a temperature sensor and a network interface. Encased in a small aluminum box, exposing only the DC power connector and the Ethernet port, intended as a fixed monitor. It's main advantage is the relative low production cost, while offering continuous radiation surveillance. The radiation detector is built upon a SBM-20 tube.
- Model B: the same compact size as the above, while also offering a LCD and a built in battery for remote operation. Model B can be used both as a monitor (when connected to the network via Ethernet cable) and as a mobile dosimeter showing the radiation levels on the LCD. Data is centralized only if the device is connected to the network (via the Ethernet cable).
- Model C: the most complex uRADMonitor design, features the same LCD as the previous model, but comes with more features and sensors: a GPS unit for mapping radiation data to location and also for standalone use as a GPS Tracker, a SDCard slot for logging the data when a network connection is not available, a wireless network connection to avoid installing complicated Ethernet cables - only DC power is required. Model C logs data to SDCard when Internet is not available and then uploads the data automatically when a connection becomes available. The wireless connection is via 2.4GHz radio waves to a local receiver connected to the Internet router by cable: so you can mount the uRADMonitor station outside, and have it upload data wirelessly to a receiver placed inside and connected via a short Ethernet cable to your Internet router. Very much like in this diagram.
October 4, 2013: Model A Ready!
The Model A first prototype is ready. The board fits the case perfectly and it slides on the side rails when taking it in/out. The software is also close to the finish point and the device is already sending data to the uRADMonitor servers. I am happy with the size of the device, and the rugged feeling given by the robust aluminum case.
For this device I developed a tiny 400V inverter module, all with SMD components including the ferrite choke. No need to wind any other coils with hundreds of turns for ferrite core transformers like I did before. I'm sure this will be a huge time saver.
Simple and elegant, the board has plenty of space for adding all the extra features in Model B. Yes, the same case size will be used. Model B will also feature an LCD display.
The second picture shows the inverter module, just at the right of the Ethernet connector, on top of the red light background.
October 5, 2013: Completing the software
Now it's the time to complete the software for uRADMonitor Model A. First go the timers, the PWM (for the inverter), the temperature sensor code and the Geiger tube pulse counter, taking into account the microcontroller frequency set by an external quartz crystal of 8MHz.
For the inverter, I am using one main switching transistor driven by PWM, and controlled with a discharge transistor for sharper pulses and higher voltage. This was explained above already - I will not repeat it. This part of the circuit is entirely done with SMDs, so I get a lot of functionality in such little space. Given the SBM-20's working interval of 350-475V, my plan is to drive the tube at 376V.
In order to find the inverter's best combination of parameters (frequency and duty cycle) , I did some comprehensive measurements and put them into a chart:
The complete list of values is available in this PDF document: inverter_performance
Given the results, the best inverter settings seem close to duty cycle 50%, but don't forget I am using a discharge transistor - as a result the main switching transistor is ON for only a fraction of this.
The code makes the duty cycle auto adjustable for the target of 376V so we have a regulated voltage applied to the tube. My frequency of choice is 19KHz, way above the ultrasonic audible threshold. If you look at the chart, my choice of parameters corresponds to the cyan peak (that hits little over 380V).
I also noticed the interesting rectangular shapes formed by the different combinations of parameters, that are shifted both in amplitude and frequency. Maybe someone can explain this phenomenon - just use the comments section if you have any idea on it.
October 8, 2013: Rugged aluminum cases
I now have completed the second unit, and have the first two units of the uRADMonitor unit Model A ready. The last hardware job was to complete the case lids to make space for the two connectors, the Ethernet and the DC plug. For the next units, a chinese manufacturer will take care of all the details so these two first units are entirely hand made (pcb+smd soldering+case power tools work). I am quite happy with the result and the robustness of this product so far.
The next few days I'll run tests for checking the quality, the stability and the behavior when exposed to various sources of radiation. Adjustments in software will be added where required. Here are a few photos I took today:
The photos show the compact size, and the few external objects have been added for comparison purposes. Of course, the peanuts are not needed when monitoring the radiation levels.
October 10, 2013: Data sample charts
While keeping an eye on the stability of the first two units, I have decided to publish the data, because I'm sure many of you want to see these first tests. But remember, all data show at this point is experimental and for testing purposes only.
The units are currently indoors, and the temperature sensor is inside the case, so the temperature level can be higher.
|Device "#2" (round case):||Device "#3" (rectangular case):|
The dose in uSv/h is approximate. Some people seem to react exaggeratedly to this conversion, resulting in this note (like what other measurement is anything but yet another approximation).
As you can see the Geiger tube voltage has been set to 390V , the inverter's frequency is set to 17KHz and the duty cycle goes around 41% (but changes automatically to maintain the voltage constant).
EDIT: The test closed on October 29, 2013. The data charts above are showing the last data recorded. You can download the complete database data of this test, here
October 13, 2013: Comparing the first two units
Today for my 13th update to this project, the plan is to add a comparison graph between the data measured with the two prototype devices, #2 and #3.
Looking over the numbers recorded so far, I quickly observed that while there is a negligible difference in CPM levels (probably due to small differences between the two SBM-20 tubes used), both radiation graphs for #2 and #3 follow the same trends.
It seems that even small changes in the radiation field at my location are being picked up in the same way by the two detectors, which is great news given the small values involved (only the background levels).
To make viewing data easier, here is a comparison chart for #2 and #3, showing the last 24hours data, integrated at 30 minutes (better resolution than the previous charts):
I plan to run an experiment soon, in which I a radiation source will be positioned equidistant above the two detectors to see the results. Then, another experiment will only have the source on top of one of the two detectors.
EDIT: The test closed on October 29, 2013. The data charts above are showing the last data recorded. You can download the complete database data of this test, here
November 22, 2013: SI-29BG Geiger tubes
November 26, 2013: Model A with SI-29BG
After receiving the batch of SI-29BG tubes, I had to design a compatible PCB board - a variation of the Model A board for the SBM-20 (presented in the previous posts):
Despite using different tubes, the Model A software compensates the readings and we get consistent indications. A SI-29BG based device is already running in Carei, Romania, and real time data can be seen here.
Having completed a few more devices, I had some time for a few pics. See them below, devices with both the SBM-20U and the SI-29BG tubes:
January 09, 2014: Google maps integration
As the first few prototype units were already up and running, pushing streams of data to my server, I was able to make the next step and implement a very basic radiation map plotted upon a Google maps layer. This is just a very basic first integration, as I plan to change the radiation web portal in the months to come:
From now on, anyone can access the real time radiation data on www.uradmonitor.com .
January 11, 2014: First unit overseas
First uRADMonitor in the USA has become operational today in Chicago, IL. You can see the measurements on the radiation data portal.
January 12, 2014: Model A Case design
At this point I have to outsource some of the effort needed for building these units. First is the aluminum case, I found a promising Chinese manufacturer that should build the cases according to these specs:
The front lid offers minimal hints for the two connectors, and the high voltage warning (as you know, the Geiger counter uses a 400V DC supply, not dangerous due to its low energy, but can give quite an unpleasant shock). The body shows the logo and the uradmonitor portal address.
January 13, 2014: Parallel measurements
The plan for global coverage imposes strict demands on the calibration and tolerance of each of the uRADMonitor devices. First thing is to make sure that any two devices, exposed to the same radiation field, will return similar readings.
As I was preparing to ship out two additional units #11000003 and #11000004, one to Denmark and one to USA, I had to run a quick test to make sure the two units are properly calibrated. So my plan was to verify the readings for two scenarios: (1) exposure to a radioactive check source and (2) natural background exposure. For both, commercial dosimeters were set in place as a reference. One was the Terra-P MKS-05 and the second the Radex 1706.
(1) For the first scenario, the two uRADMonitor units were placed side by side. A small natural mineral sample containing thorium (thorite) was placed symmetrical to the position of the detecting element (SBM-20U) inside the two units. The Terra-P and the Radex-1706 were also used as a reference in exactly the same setup, placing them at the same distance (measured from the cvasi-point-like thorite mineral to their Geiger tube detector element, inside the case):
The two uRADMonitor units were left like described for an interval of aprox. 9 hours. The collected data is available below. Both the Terra-P and the Radex agree on a measurement value of aprox. 40uSv/h.
(2) For the second scenario, the two uRADMonitor units were placed alongside for yet another interval of aprox. 5 hours, but this time they were exposed to only the background radiation as the Thorite sample has been removed from the setup.
The Terra-P and the Radex show a result close to 0.12uSV/h for the normal background radiation level.
The data collected by the two uRADMonitor units can be seen in the charts below, both in CPM and the equivalent dose (uSV/h) estimation, click for larger images:
Note: The blue line is just the average value of all the values in the graph. It is not a coordinate on the vertical axis. The exact values plotted can be seen in the individual charts, where the horizontal axis is the time, and the measured values are printed directly on the graph line in black.
The higher values correspond to the exposure to the radiation emitted from the Torrite sample, exactly as shown in the previous pictures. Both uRADMonitor units have registered consistent values, 0.40uSV/h for the Thorite sample, and 0.12uSV/h for the background radiation (in the right side of the graphs, where the check source was removed). Here is a comparison chart, to offer a better view on these numbers:
And the linear aproximation to equivalent dose in uSv/h:
As pictured we got a very good consistency in the way the two devices registered the radiation they were exposed to. Also the equivalent dose estimates are matching those of the two commercial units.
Here are the other parameters monitored by the two uRADMonitor stations, also in good concordance:
As this test was successful, yet another uRADMonitor device was shipped to Denmark, hoping to see it up and running the next few days, as it gets to its destination.
January 22, 2014: New units in Denmark and Germany
Two more units got online today, one in Denmark (#11000003) and another one in Germany (#12000008). See the real time data on the uRADMonitor portal.
January 24, 2014: New model A PCB Design
It's been a busy day today, but I got a new PCB design for the model A. This will be used to derive the model B, as this latest revision shrinks everything even further, making space for additional components.
To name a few changes, the microcontroller was changed from DIP to SMD, the high voltage inverter has been reorganized (in terms of space), and the preamplifier also got a new location. The board is designed to fit any of the two tubes, SBM-20 or SI29BG.
The last picture shows a comparison with the PCB featuring the DIP microcontroller. Despite the ultra small components and traces, this latest device worked from the first test.
January 25, 2014: uRADMonitor portal updates
The previous version of the uRADMonitor portal barely presented the radiation data. It was very basic, and not quite user friendly. I now tried to change that a little, but there is still a lot of work to do. Here's what was changed:
- the station detailed info screen was improved, to show a few more details, all better structured:
The most important change was the Online / Offline status. If is computed by comparing the last data package received to the current time. Considering the data reports are sent one per every minute, if the last entry is 5 minutes old, we consider the station offline. This process is automated. When the station restarts its broadcast, it will change status to Online.
- The last entries as shown at the top, the last readings for radiation (both the equivalent dose and the counts per minute for the particular Geiger tube) and for the internal temperature (these units have an internal temperature sensor).
- Finally the history chart is now bigger, and shows data for the last 24 hours as recorded.
As the model A units are located using the IP , errors in localization are normal. For most cases we got enough accuracy to correctly pinpoint the city where the station is broadcasting from. But there are situations where manual corrections are required. Sometimes we also want to protect the privacy of the owner of the uRADMonitor station. This is why I added a manual location override option. Using it, I was able to correct the location for #12000007 to Carei, Romania (Instead of Bucharest) and #12000008 to Schecen.
Now there's also a Donate button. Those who feel that this project has potential to be useful to the community, are welcome to support it. I have started it with very little resources, but I am committed to make it grow to the global scale.
February 08, 2014: A new radiation monitor, in New York, USA
The news for today include two things. First, I finally found the perfect case for the Model A. It is a high quality aluminum case, that comes with a nice wall mount. It's a little smaller in width than what I've previously used, but I was able to make a new PCB to adapt the circuit to the new size. Before I can get enough of these, the previous black case will still be used for the model A.
For the model B, I will use the larger black (round) case, as I need to add a LCD, a battery and some additional internal modules, so the extra space is important.
I said there were two updates for today. Well the news is that the second uRADMonitor running on American soil is up and running at a location close to New York:
The radiation monitors can be seen on the map at: http://www.uradmonitor.com
February 16, 2014: Model A wall mount case
February 26, 2014: New unit in Arad, Romania
Unit 120000A has become operational today, in Arad, Romania. The unit appeared on the map as soon as it was plugged in, with a very good geographic accuracy thanks to the built in IP locator. Nevertheless the coordinates have been manually adjusted, for even better accuracy, closer to the real position.
Before sending this unit out, it has been calibrated and verified against another unit. The two units ran in parallel for several hours, covering two tests: exposure to a radioactive source (a natural Thorite crystal placed at given/symmetric distance to the two detectors), and another test consisting of exposure to the natural background radiation. Here are the registered measurements:
Temperature and voltage on the two tubes:
The two devices have registered the radiation levels in a consistent manner so the tests were successful. The background radiation level at my location in Timisoara, Romania where the tests were performed, was measured by the two devices to an approximative (attribute used because of the CPM->uSV/h conversion) value of 0.11uSv/h:
As the unit arrived at its destination, in Arad, Romania, I was surprised to see the live data shows a perceptible increase in radiation levels:
I am not sure what caused the spike in the chart, but the rest of the picture clearly shows an average value of 0.14uSv/h, and that's something, considering Arad is not more than 50km from Timisoara. Here is the station up and running on the uRadmonitor.com portal:
And here are a few pics showing the 1200000A running in parallel with the expensive Gamma Scout dosimeter. The results measured by the 1200000A , currently at 0.14uSv/h, are in good concordance to the values displayed by the Gamma Scout:
This uRADMonitor unit has been calibrated at a different location, where it showed only 0.12uSV/h, yet at the place of destination it was able to measure the radiation correctly , and in good concordance with the Gamma Scout dosimeter.
March 10, 2014: New unit in Augsburg, Germany
The second unit running on German soil has become operational today, in Augsburg, Germany.
It measures an average of 0.15uSv/h, relatively higher then the other units currently in the network. Before sending it out, the calibration tests performed on the 11000008 in Timisoara gave results closer to the 0.12uSv/h figure, considerably smaller:
Also the Augsburg natural radiation background level seems to be higher than the one measured at only 100km to the South-East in Schechen, Germany, by the uRADMonitor station 12000008:
The 11000008 in Augsburg averages 0.15uSv/h for the last 24 hours, while the 1200008 in Schechen only reaches 0.13uSv/h . The real time data is available here.
March 20, 2014: New unit in Washington, USA
EDIT: there is a dedicated blog on this project, here. Make sure to follow it for latest news.