User Tools

Site Tools


Navigation Menu

PiGI - Raspberry Pi Geiger-Müller Interface

Geiger counters are basically just devices which enable us to measure ionizing radiation. In the context of human activity we have to deal with “natural” radiation sources and “artificial” ones. Some of the materials emitting ionizing radiation are used for medicinal purposes (Radiation-Therapy), as an additive in paint and even in smoke-detectors but the majority of it is used to create electricity in nuclear power plants and to stockpile thermonuclear weapons of mass destruction (well, except in iraq, as we all know), so many could get scared and a few could get rich.

EU Nuclear Reactor Sites and recent problems

Source and more detail with per site incident information on this mashup:

https://www.google.com/maps/d/viewer?mid=1JvNHsKIGS2e2SY0qmlhMUfzY0mQ

But how do we deal with something we cannot actually see, hear, feel, smell or taste? Exactly! We just don't think about it. We never perceive it as potentially dangerous to our health until an accident happens and we have to find someone with a geiger-counter, we have to trust, who can tell us how sunny the “weather” is? Better we do it ourselves, so that we can independently get all relevant realtime and historic metrics in a decentralized p2p system, with a far lower risk of getting tainted/biased/wrong/no data, especially when we as people really need it.

That should be in common rational awareness instead of ignorance or fear. Just like rain. We all know rain is neither good nor bad. It's just rain. And when we stay out too long in the rain without the appropriate protective clothing, we might get soaked so bad, that we might get sick. Our brain determines our maximum allowed in-rain time and the level of protective gear we need to repel the rain so we just visually perceive how many rain drops we see and how big they are. That gives us the freedom to decide whether to stay indoors or to wear some form of raincoat or just to risk it.

Unfortunately our body won't help us with detecting ionizing radiation (although it can suffer badly from it). Lost sources and depleted uranium shells, that irradiate uncontrolled, are as much a real problem as reactor incidents or catastrophic accidents like Chernobyl or Fukushima. Not to mention future generations, which will have to deal with the long term nuclear waste deposition. But that's where geiger counters come back in: They let us “see” radiation levels and therefore give us the freedom to decide if and how we want to go on as safely as possible.

After the recent nuclear disaster at the Fukushima Daiichi nuclear power plant in Japan a lot of people started to realize this and began to develop, build and buy geiger counters and geiger counter kits. Many share their local radiation levels through social networks and simple maps, just like a rain radar (e.g http://blog.safecast.org).

This would also be a great metric to collect for Argus and to display the data as an overlay in DSpace. Now imagine hundreds of these cheap solutions spread all over the world and sharing the metrics with one another to get a global Open Radiation Monitoring (ORM) network.

Open Radiation Monitoring

PiGI Module

PiGI V1.0 Prototype Board

The Raspberry Pi is a perfect platform to be a cheap but very versatile geiger counter. It can be connected to a TV/Monitor to display a nice graphical interface, it can play the nostalgic tak, tak, taktaktak sound via audio output and it can also serve as an autonomous geiger counter sensor node (even solar powered) collecting and sharing real-time and historic data. All you need is:

  • A Raspberry Pi
  • PiGI Module
  • Geiger-Müller Tube

The PiGI is designed & built as a simple open-source plug and play module for the Raspberry Pi to transform it into a cheap and hackable multi purpose geiger counter. It generates the necessary high voltage (up to 1000V) the GM tubes require to operate (typically 400-600V for beta/gamma GMTs).

Some pancake sensors (mostly used for alpha radiation) require an even higher voltage, theoretically the circuit should cover it. However, that's untested, only a LND712 (new, see picture) and two old Phillips tubes (Frieseke&Hoepfner FHZ74/76) could be tested, due to limited prototyping resources.

For every impulse the geiger tube registers the PiGI pulls a GPIO Pin of the Pi to ground (falling edge detection). On the Pi runs just a little daemon called counterd which waits for an interrupt to register the count and notify the display/audio/data-storage/network handler. That's why it can also easily be connected to any other embedded/micro controller system like:

  • Arduinos
  • ATMegas
  • PICs
  • Spark-Cores
  • Other embedded Linux ARM/MIPS systems with GPIO Inputs (Netus G20 etc.)

HOWTOs can/will be provided, drop a comment if you need one for connections/pull-up configuration.

Use Cases

  • Local/Handheld Alpha/Beta/Gamma Radiation Measuring
  • Autonomous/Remote Federated Radiation Monitoring Network (Argus/DSpace)
  • Radioactive Decay based Hardware Random Number Generator

Specifications

Board Dimensions 40x43mm
Output Voltage Up to 1000V
Current <2mA @ ~0.09uSv/h local dose rate
  • Efficient design - very low energy consumption - supplied by the Pi
  • Simple circuit / Low BOM count
  • Simple layout for easy hand-soldering (even for SMT beginners)
  • Dual layer (top/bottom) only 3 vias - for easy DIY replication
  • Very cheap design (Prototype costs per Module: <EUR15)
  • Dual stackable for low/high dosis counting with 2 tubes
  • Micro controller input protection (impulse inverter)
  • Free & Open licencing: CERN OHL 1.1
  • Plug and Play for Raspberry Pi

Hardware

If you want to know more about the design or want to build one yourself, check out the PiGI Hardware Section for a detailed circuit description, schematics and printed circuit board layouts.

Software

We've invested a couple of days into the new geiger counter software. Even though it is only in an early rough state it is already fun to work with. It's implemented as a full mobile-first HTML5/CSS3 Websocket based web application, than can be used in latest chromium & firefox browser. You can learn more about it in the software section

Preliminary Screenshots

Main Instrument View
History View
Ion Trace Visualizer


Implemented features

  • Live Status
  • Live (15min/60min/24h) Graphs
  • Analog gauge
  • Ion Trace Visualizer
  • History
  • Tick Simulator (For show and development)
  • Hardware RNG entropy generator

Status / History

2013-04-15 Project Idea
2013-04-17 Basic circuit definitions
2013-04-23 Identified and ordered all parts
2013-04-28 Finished V1.0 Prototype Layout
2013-04-30 Ordered 7 prototype boards from MultiPCB
2013-05-05 Soldered first prototype
2013-05-13 Added dual-stack option (V1.1)
2013-05-19 Changed R1 from 3k9 to 1k to make T1 work more reliable
2013-05-22 Bumped prototype layout to V1.1 for production release
2013-05-23 Started long-term stability test
2013-06-17 Stopped long-term test for live-demo (stable/reliable 100% until now)
2013-06-17 2nd prototype board finished
2013-06-23 Updated + Added V1.0 board images
2013-07-25 Pre-production planning
2013-08-04 Indiegogo.com campaign drafting has begun
2013-08-09 Selected CERN OHL 1.1 as licence for now
2014-02-08 Personal Log: PiGI Testing
2014-03-04 PiGI Software Hacking & Hackathon
2014-06-13 Generating Entropy From Radioactive Decay With PiGI
2014-06-28 More PiGI prototype boards finished
2014-10-15 Added PiGI Longterm Test Data to VFCC Dashboard (ORM Precursor)
2016-03-21 More PiGI PCBs available through small-scale crowdsourcing
2016-12-13 PiGI adopted & deployed @ https://tdrm.fiff.de/messtechnik.php
2016-12-19 Re-added missing C5 (330pf - 0805) to mouser BOM

Next steps:

  • Run a crowd funding campaign
  • Get adafruit or seeedstudio to produce, sell & ship PiGI's for a fair price (You could help by sending them a note with your interest and link to this page, so that they know there is demand)

Repos

I want to have it

Since the PiGI is ultimately cheap to produce once you reach the 1000 pieces boundary it would be great to put this out to the Pi-Community as a kit or even ready-to-go modules, the only thing that would need soldering is the tube itself (no SMT solderig). If you're interested in having one, please vote for the option you would like best so that we can determine how to structure a kickstarter/indygogo campaign that suits the community:

What would you like to have?
  Just the PCB ~EUR5* A kit with PCB+Components ~EUR13* 1 Ready-to-go Module ~EUR20* 2 stacked ready-to-go Modules EUR40*
3ernhard    
123    
A>  
AaronM   
AC  
AdrianFZ    
agx    
AIB  
al    
alan    
Alessandro    
Alexndr    
Andi    
Andreas    
AndreaZzZ   
andy    
Anony    
Anonymous    
AnonymousZebra    
armac    
Armas    
Armin (alarmschaben)    
Arthur    
Arthuro    
Arto K    
Artyom    
Auto    
Banana0042    
Barthm    
Becks    
Ben B    
Bennie    
Bernd    
Bernhard   
Bert   
Birne    
Bitru  
blaugrau
boelek    
bof    
brain    
Bruno   
bunfer    
CabbyHat   
cagan    
caldimerda   
carsten   
Chad L.   
Charley    
chris    
Chris L   
ChrisM    
Chriss SQ3FYK    
Christian   
ChristianC    
clemens    
Clemens    
Cody    
Connorl   
ConSi    
count-me-in    
crazy lemmon    
Crusty SOCK   
CrysisLTU   
CS    
CyberPunk   
dafosy    
Dalail   
Dan   
Daniel   
Daniel Fernández     
Darryl    
Dave     
Daver    
David    
DavidR    
db0B    
Dennis    
DerPatrick1987    
derpsson   
dimm    
DL8EBL    
DrPhilXray    
dude    
Edriee    
elvis    
energyideas    
erich    
Erik    
Ethan  
FabienD    
Fabio    
Fabs   
fader   
Fafnir   
Feek    
fehlfarbe   
Felix    
Fenrir    
Ferdi    
Ferdinand    
Fiesiker    
fjoe    
foo    
fr    
Frances    
Francis    
Frank    
Franky   
friedri    
Frillip   
Fritz    
Fumfoo   
Fuwo    
fuzzyhypothesis    
G7TKK    
Gaudeamus    
geologica    
Georgeos    
Glubschi    
goibhniu    
grantorax    
great project  
Grenthal    
guenter    
gzx    
Günther   
hajola   
Hal   
Hamish    
HansD   
harpss1ngh    
HarryBalzonya    
Harry_Bonanza    
hbeevers    
Hegz    
Henry  
Herb    
Herbert  
hiange
Hi Mom    
Hobson  
Hogan    
Hogelmann   
HohesC    
ian    
ibot   
iggy    
Ingo   
ismael   
Itxaka    
J   
jacopo    
jambam   
JamesT   
JanK  
Jannis    
Jari   
JasenFree    
Jens    
Jeremy    
jFMd    
jg    
Jindrich   
jjeffries    
JK From Finland    
Jo    
joe   
Joe    
John    
john    
Johnny Bravo   
Jorge    
Julian    
K   
Kev    
kjm   
KL   
KomKon    
konrad   
kyr0    
Lars
Laura    
lbaonline  
LD    
LHackett   
link   
Lionel    
Liow  
Loic    
Lord Loh  
LorenzoLamas   
Louis    
Luca    
lule    
luqas   
M   
mactron    
Mad    
Magyar Legény    
Makoland   
Manfred    
Mani   
manu    
maqd     
Marcus    
Mark    
Markus    
Martin Smith  
Martin want it    
Mathias    
mattia    
Mattias     
MaxiFrett    
MD    
Me    
me   
Merv    
Michael   
MichalSz    
MidnightCommando    
mike    
Mike A    
MikeBurton_2000    
Mike G    
miksaup    
milanco    
milutzu_k    
misisanta
mj   
mjk    
mkb600    
Mo    
MOe   
Momo    
mrmrmr    
mulles   
Mynasru   
n8foo    
Nemmich  
NemoN    
Nick    
Nico    
Nikolay_Ch    
Nils Hitze   
njo   
norbert    
Nordrunner    
Nuaberhopp    
Nuc    
NukeTom    
okgi    
osi  
P    
PA7T    
packishman   
Pankratus    
PapyFeel   
Paradox   
pascal   
Pat    
Patrick    
Paul    
paul    
PaulM   
Pavo  
peddi from germany    
perry    
Peter M    
peter s    
Phil    
Pip   
pk (no joke: all)
Plamen    
Prothon   
ptarjan    
ptr    
PumpkinEater    
R    
RadioInActif    
Rainer    
Ralf    
Ralf2    
ralphi   
ratzifummel   
RayJ   
recantha   
Reiger    
ric    
RiccardoJ    
Richard    
Richard Liow     
Rico    
RoBBBoY    
Robert   
Robert from Poland   
RobG    
RobS    
RockyBoulder    
RockyMountainSlim
Roland   
Salipotens    
Sam    
sbiella    
scavenger    
Schtubb    
scm    
Scott    
Sebastian    
Sebbe    
sepp    
Serge   
sexy beast    
Shade    
Simo  
sm    
SmokeMaster    
SolderingIsFun  
Sparky    
starfinger   
Stephan    
stephn  
Steve    
Strainu   
taqd  
teg    
Terry    
tgage    
ThatGuy    
Theo   
theo    
Thiange Doel: 20 pieces    
Thomas    
Thomas from Germany  
Thomas L.     
Thomas W.    
tianxiaobo    
Tigoud    
Todd Smith    
Tokyo Loose    
Tom    
tomgidden   
TomMaverick  
TonyD    
Tosebach    
tsf    
tungl    
TWF    
Uli    
Unixversal  
UnNamedGER    
VC    
Vdliv    
Vendus    
Vicus   
Vitas   
Vitti    
Wal    
what    
What is PCB? Hi mom!    
whatnick   
Wickex    
wiesel    
winni   
wm    
Wofei    
xiaobotian    
Xman   
Yani   
Yanik    
Zaptac    
Zarumba    
Zarumba_2    
zaza    
zboson    
zendick    
zgub   
Zoltan   
ZwaZo    
üps :O)   
ʞɹɐɯuǝp ɯoɹɟ uɐɯɹoN    
Count: 17 164 286 62

* These are the current sale price target estimations (excluding shipping), based on a worst cast scenario. As soon as there is a more solid count it is very easy to get the figures what it's really going to cost. The PCBs and part prices drop very much at 1000, 5000 and 10000 pieces breaking points.

Thank you for your time to state your interest. You might want to consider subscribing the Apollo-NG RSS Feed to get instant project update notifications. As this project is currently very dynamic in nature you are always welcome drop in on #apollo on freenode IRC for questions and discussions.

Discussion

Danny Iland
2013/06/12 08:05

I am a PhD student at the University of California, Santa Barbara. My research group focuses on Wireless Networking and sensor networks. I would love to add your sensor to my solar powered Wireless Pi! I would be happy to push some python code to your repo if you send me a prototype to develop with.

Danny Iland

iland@cs.ucsb.edu

chrono
2013/06/12 09:45

Cool stuff Danny, have a look at the github repo, there is a basic counter daemon implementation. It's only prove-of-concept code for now. IMHO it would be best to build the counter daemon as simple as possible (unix style) since the PI's interrupts are a bit flimsy. It might also be that my hackish code is just bananas and there is a much better way to do it but I'll leave that to the guys who are real software developers and know what they're doing :)

Then we could just use some event mechanic to notify other pieces of software to do the actual counting and calculation, logging, etc. using SHM, a unix socket or whatever is cheap and can be used easily with python, nodejs or whatever else we might want to use :)

I'll try to solder some more prototypes within the next 7 days and will come back to you once they're ready.

Michael Horne
2013/06/13 05:23

I created the Picorder, which is a Pi-based version of a Star Trek tricorder. It was featured on the Foundation homepage. I'm very excited about the PiGi as it'll make an excellent addition to the device. Just wanted to wish you well with developing the board :-)

chrono
2013/06/13 05:29

That's awesome Michael, another hacker and I were discussing building an open-source tricorder with interchangeable measurement/monitoring modules so it's even better, when someone else already got it started :)

Michael Horne
2013/06/13 05:34

:-) I'm currently on version 2. The first version used an Arduino nano-sized clone to read the analog inputs. Now I'm using a Key Lime Pi from nwazet.com so it's Pi-only now. The bulk of it is made out of Lego!

Aaron M
2013/06/23 00:56

This is just a really cool idea to increase awareness about something few people even realize exist. I don't understand it much myself, but something like this would definitely bring the subject into the common awareness. I was thinking about cases for the project too, and what better than a surplus ammo can? What better way to store something that is used to measure radiation?

chrono
2013/06/23 21:21

Yeah, these are pretty sturdy & nerdy and I know a lot of people who make good use of these surplus boxes. Just keep in mind, that you will not be able to detect alpha through these boxes. For beta/gamma tubes it sure is a nice hackish case when you've got a surplus around :)

john
2013/06/29 10:08

i have worked on something similar for arduino last year, i was considering trying to port it over to the raspberry pi when i discovered your project. I would really like one of your modules and would be willing to share ideas

chrono
2013/06/29 21:05

That's the beauty of open-source: I was looking for a geiger-counter device based on the Raspberry Pi but there wasn't anything. Then I found a couple of Tom Napier's geiger-counter implementations and the circuit seemed easy enough and was obviously working out in the wild. From there on it was just a matter of basically understanding the kickback principle, learning more about geiger tubes and finding modern SMT components that could match or surpass the old specs.

So, they aren't really “my” modules, they are just the current result of the ever increasing speed of technological evolution through internet access and people willing to share their ideas, failures and results globally. It's a team effort, even when the players involved don't know each other and that is also why this project cannot be anything but open-source again.

It was just a little step to put it all together into a neat layout that is simple (DIY) and cheap (mass production) with direct RPI drop-in capability while still maintaining modularity and hackability to use it with what ever else is counting impulses.

At this point the V1.1 layout seems pretty ready for production and it would be nice to have a few reviews of the schematics, the layout and the DSO signals. All tests have indicated a solid design but fresh eye never hurts. Other than that we need to focus on the software side, I'm already preparing a new version of the counterd and I would love to have some real developers by my side.

The poll above is already going pretty well but we still need more people to at least loosely commit so that putting time and work into a full crowdfunding/production run can be justified. So anyone can help by spreading the word.

Apart from that, please feel free to join our mailing list or #apollo on freenode. If there is interest, we can also schedule a mumble conference for voice discussions.

chrono
2013/06/29 21:11

I hope all you visitors from space-, military- and defense companies like

  • DLR
  • Pentagon
  • Armed Forces Pacific NNIC (Navy Network Information Center)
  • AFB (Andrews Airforce Base)
  • Thales UK (Optronics?)
  • US DoD Network (Alexandria)
  • US Air Force Space Command (AFSPC)

are only coming here as hobbyists/scientists. It would really piss me off to know that open-source designs are used in military applications to build machines/infrastructures which help killing people with a totally insane price tag, paid for by the tax money of the very people who actually developed and freely shared that technology…

jamest
2013/07/02 08:03

Hi!

Really like it, but what about LEDs? One green Power LED and one red LED that shows the activity. That way you clearly see whether it works or not.

chrono
2013/07/02 08:24

I thought about that but went against it. It would only drive up the costs. When the Pi has power the PiGI will have power. Even with a low current LED it would be a useless waste of power for the rest of its operational lifetime.

The signal LED makes even less sense to me since the impulse duration is somewhere between 90-200 microseconds (depending on the tube), so it's way too fast to catch the short LED flash with our eyes.

If someone really wants/needs a LED I would like to propose to do that in software: The counterd will flip another IO pin of the PI where a LED is connected (with a larger holding time, so that we can actually see it flash).

jamest
2013/07/02 08:06

Ah and some mounting holes if you don't want to mount it on the raspi directly?

chrono
2013/07/02 08:13

Have a look at the V1.1 layout PiGI Hardware: The board size was increased from 40mm to 43mm to leave room for a mounting hole precisely at the place where the Raspberry Pi board mounting hole is. The V1.0 prototypes (photos) don't have that since dual stacking wasn't considered at the time :)

Parkview
2014/02/10 13:14

Yes, it must have mounting holes! Mine will run from the RPi via a cable, so I will need mounting holes.

chrono
2014/02/10 13:41

Already covered in the 1.1 spec :)

jamest
2013/07/02 08:36

Yes, good idea. You could calculate a blinking frequency based on the counts/s. Could there be space on the board for a LED and a resistor, optional of course if someone wants to do that?

chrono
2013/07/02 09:16

Technically there is still a lot of free space on the PCB but not all of it is easily routable to the header. I'll have a look into it, anyone else requesting this feature (GPIO connected LED support on the PiGI board)?

Gene
2013/08/13 14:49

Solar Power so you could just leave it outside somewhere (Near Wi-Fi) and it will record the amount of radiation in the environment.

jamest
2013/08/13 18:32

For that, a simple microcontroler with some wireless interface would be much cheaper and easier to power. For example an Arduino nano and a RF24 or Xbee interface. Because if you place it outside, you simply don't need the processing power and interfaces of the RPi.

chrono
2013/08/13 19:39

True. That will be a bit cheaper and more power efficient (smaller solar panel).

On the other hand it would be much more inflexible and wouldn't allow for direct internet access (Data API/Webserver) without some other form of always-on/connected computer which would receive, manage and store the Xbee/RF datagrams from the uC/Arduino to make them available to anyone on the internet.

So I think flexibility wins, those who already have running boxes in RF range can do computing, storage and delivery there. Then it makes sense to connect the PiGI to a uC/Arduino. Others may like the Pi as a complete module without further dependency (except internet access) where everything is done autonomously with just a Raspberry Pi.

Peter
2013/12/19 05:32

I am also interested. Is there any recent progress?

chrono
2013/12/19 10:20

The prototypes are done and working properly. James-T wanted to write about his experiences with his PiGI but he obviously hasn't had the time to do so yet.

Also, John from DIY geiger counter wrote and asked about the transistor and the specs and updated his project accordingly. Unfortunately, he hasn't really understood the spirit of the net and collaborative work and consequently didn't throw a link back to PiGI to make it clear and give people more choices and a better understanding. But that's the problem when people do something to generate money, they need to keep it to themselves and have to appear as the “specialists” to their customers and are always afraid to get less money if they share something.

Max from pi3g seems interested in production/distribution of the module (and/or a kit) and we're going to have a talk about that to make sure that the non-profit aspects of Apollo-NG won't collide with the interests of Max's business. “Outsourcing” would be nice, because we wouldn't have to allocate time for all the logistics tasks but it all depends on the specifics to decide whether it makes sense or not.

Alternatively, we've already prepared an indiegogo campaign we can launch when we want to go for a crowdfunding approach and just do it ourselves.

I know it's bad to have to wait for something (you can always build it yourself when you really want to - everything you need is available) but IMHO it's better to wait for a campaign than to wait for actual delivery because the people doing the campaign (we) don't have enough time available to fulfill it in a reasonable time frame.

At this point it's the day-jobs that keep us from doing more but since there is still more money needed for the Aquarius, I still have to have it in order to keep the whole project progressing, even at a slower pace now.

Peter
2014/02/04 08:11

Oh man. Please take our monies :-)

chrono
2014/02/04 08:41

Well, when you put it like that :) We'd really love to. And thank you for verbalizing your appreciation and desire. Not because of the money but we want people to have funky geiger counters as cheap as possible. More self reliance. More independence. The poll has also still gathered more people wanting to have one so we feel rather confident that enough people out there would support PiGI production in a crowdfunding campaign.

The decision to wait was my responsibility because until now there was no money involved. It is easier to wait for something to happen when you haven't already transferred money to the person(s) you are waiting for. So please bear a little longer with us until we can figure out how to proceed without putting ourselves and our families at higher risk just because we want to believe that this kind of system can work much better than the old one.

Peter
2014/04/03 11:09

Hello again,

How are those thinking processes going? :-) Sorry to bother you again - it's been 2 months and I'm anxious to hear about any progress ;-)

chrono
2014/04/03 11:20

No problem - thanks for coming back and poking us constantly - it keeps us motivated :)

We've already pushed the software further in the meantime (have a look at the github repo) and would like to use the next two weeks to consolidate everything into a campaign which hopefully will be launched by the end of it.

We've also talked to Max from pi3g last week - he also offered help and resources for the fulfillment/after campaign distribution so with a little luck things will start to move soon.

Peter
2014/04/03 12:16

Yay! That is good news.

Looking forward to things moving along!

Nick
2014/04/05 10:04

Can't see which tube it's using … often a problem to get hold of the right one - presumably designed to use one that is available reasonably cheaply and easily ?

Nick

chrono
2014/04/05 10:50

We've designed it to be used with almost any available tube so that anyone can pick a tube that is available/affordable. Although we haven't tried it, due to lack of funds to test more tubes, even 1400V pancake type sensors should work. The only thing that needs to be changed with different tubes is the voltage level, which can be easily adjusted with R10.

xychix
2014/04/28 06:50

Hey Chrono, how is the kickstarter idea progressing?

chrono
2014/05/29 09:50

Ahoy there, iggy and I are still hacking on the software, in order to be able to have something we can show people, you know. It's much easier to promote a video of the whole system in use with real world examples than to explain the technical details with lots of text :) We've locked down the issues for the first “official” release and will then generate enough visual material to launch a successful campaign.

peter
2014/05/29 09:23

Thanks for providing this procect. Can't wait for buying a package somewhere soon. I am wondering hat it doesn't need an external power supply.

BTW graphs can be very easily generated with the help of Munin. Just needs a script to output the actual value to the cmdline.

chrono
2014/05/29 09:39

Thanks for your feedback. The only power supply you'll need is the 5V/USB power for the Raspberry Pi. Since the PiGI Module draws only very little power, it can just use the available 5V pin on the Pi's header.

We've currently ordered our issues on github and are trying to handle them for the final release to have the first software version finished. We have come up with our own storage/rendering system since the munins, cactis, collectds, rrds, graphites, […] were all either too cumbersome and heavy set for the job to perform well or just not flexible enough to produce what we wanted.

Right now, the Pi stores all metrics & annotations in a local LevelDB instance and only pushes data to the client via websocket. The browser then handles the rendering of the graphs with dygraph. This approach scales very well with the number of clients, since the heavy work is done on each client, keeping the Pi free to do its data gathering/storing/distributing jobs undisturbed. It's pretty fancy, check out the code on github and run it locally (when you have a GNU/Linux system). It has a demo mode and a tick simulator, so that you can play with & help develop the software even without a Pi or PiGI module.

For the future and global monitoring in the context of what we call ORM (Open Radiation Monitoring), where PiGI users put up their nodes to collect current environmental data, we currently see influxdb as a basis for a similar concept, where all nodes push their data not just into the local LevelDB but also to the network of influxdb instances. These metrics can then be mapped and analyzed from anywhere with just a browser.

ptarjan
2014/07/08 06:40

Hey guys, what's up? Some progress update would be nice…

Roland
2014/07/09 11:31

Congrats to this PiGI - it's really cool what you made so far! I like the compact board and the great looking software! … i'am thinking of building a prototype by myself - but not sure if I get all the parts ;/ Any idea of a timeframe to provide this as a kit?

Roland
2014/07/09 11:54

btw. is it possible to plug-in a common display to the GPIO-port of the Rasberry Pi on top of PiGI?

ptarjan
2014/08/23 05:56

@Chrono: Your estimate on April 3 was that things would start moving along two weeks from then. That was 4.5 months ago. Care to revise that estimate?

tristram
2014/09/16 09:13

I like that it consumes very little power and that it would work with Arduino, RaspPi (the B+ consumes 0.5 to 1 watt less power :)) so that a portable geiger counter which consumes less watt and runs longer is nice.

Still there? How's Kickstarter/Indygogo/… campaign doing? I just want to be able to log the CPMs to a file on rasp's sd card or show the CPMs directly in the command line and plot it later with Gnuplot, so don't waste too much time on writing your own plotting software :)

ptarjan
2014/09/16 10:56

Tristram: I second that. Just roll out the hardware - it would be expensive to do it individually. The software side will take care of itself in the process.

Come on guys - a lot of enthusiasts are eagerly waiting for you!

tristram
2014/09/16 17:00

Yes, it's working hardware already and Gnuplot can also do an average (15min, 60min, whatever min) plot and much more. I can also change colors (not that I would) so it looks h4xx0risch lol. Stop wasting time on your own plotting software ;-)

wasnhierlos
2014/10/18 09:05

Immer noch nix? Für live plots gibts Flotcharts (und andere), spart euch die eigene Software! Brauche kein Kickstarter & Co. 10€ mehr is mir egal. Wäre schön wenn man es jetzt also kaufen könnte.

chrono
2014/10/18 10:07

Yeah, it's not about the software, as you can see in the commit logs. It's about time and since I've been forced to move out of my current location, it's more a question of personal/biological survival and getting a new place where I can store the lab stuff and maybe even sleep. I'm sorry for that but life is just the shit that happens, when you're planning to do other things. Actually, I'm rather grateful that I trusted my guts and didn't start the campaign yet, this would have been the worst timing ever to have to handle both challenges and the logistics involved.

Roland
2014/10/23 10:11

Dear Chrono! Thank's for your honestly post! It is really sad to hear from your bad siuation and hopefully it will change soon and you find a good place to stay with all your stuff! Wish you the very best!

Asdf
2014/11/23 17:21

chrono, Monat später und wie siehts aus?

chrono
2014/11/23 18:04

Well, technically I shouldn't even waste time to reply to such a blatantly obvious attitude of utter carelessness (see name and language selection). It's easy to demand things quickly when they are for free, right? Maybe I haven't made the situation clear enough, so here we go again:

Please don't forget that the recommended model of our current society is to “protect” all the “intellectual property” with patents and closed-source bullshit in order to sell it for a shitload of money, so that a few people can get rich (successful).

Instead I choose to invest almost all the money I earn (by trading a substantial part of my lifetime in a day-job) into this project and all its sub-projects, only to invest the rest of the remaining lifetime in order to actually do something and then document/release everything that I could learn. Open. For free.

This is the key spirit of Apollo-NG and all other projects/groups around the planet like it. Because we (the people behind them) don't crave money or personal fame, we simply believe that increasing the speed of knowledge transfer by freely sharing our solutions (and keeping them open) will increase the speed of open technological development and enable people to do so many incredible things in such a way, that a new form of global mindset/society will eventually emerge from it naturally. Without all the scrambling and hassle we humans often seem to fall into when we've waited long enough without doing anything, to actually let the shit hit the fan.

If you want it that bad, everything you need to produce it yourself (slightly more expensive than a big batch though) is made available. Just create the pcbs with any DIY technique available to you (search the net for diy pcb making) or just send the eagle files to a pcb manufacturer and order the components from mouser/digikey. It's easy. If we can do it, so can everybody else.

Now I am soon to be completely homeless, because I chose to trust the wrong people and a new compatible place to live and survive in Munich is nowhere in sight. I am working every wake minute now to find a new place and to get the rest of this site into a more comprehensible/presentable state to start a crowd-funding campaign for the survival of myself and Apollo-NG.

Having a mobile habitat like the Aquarius (Landing-Module) would definitely be enough to sustain my basic needs as a human being and would also rid me of the necessity to keep a day-job because I wouldn't have to pay monthly rent anymore. And I would never have to rely on people keeping their word again.

I'm sorry that I don't have better news. So siehts halt aus.

asdf
2014/12/25 21:46

Danke für Deine Antwort. Ich muss zugeben, dass meine Frage ein bisschen, aber nur ein bisschen, provozierend gestellt (so als ob ich gar keine Ahnung hätte) weil sich eben nichts tut. Das ändert aber nichts daran, dass ich (und hoffentlich andere) Deine Arbeit respektiere, u. a. weil Du dich für open source entschieden hast bzw. der “spirit of Apollo-NG”. Ich nutze Linux wo nur möglich (privat überall Linux/FOSS). Schade, dass deine private Situation gerade so kacke ist weil du den falschen vertraut hast. Leute sind immer mehr getrennt durch all die negative Propaganda von allen Seiten. Rom ist auch nicht wegen der Bleileitungen (Kalk im Wasser bildete sehr schnell eine Schutzschicht), sondern wegen der Moral untergegangen. Hier so lange mein Beitrag für eine tragbare low power Lösung: http://forum.43oh.com/topic/6198-low-power-geiger-counter-with-msp430g2553/

chrono
2014/12/27 09:41

Lieber asdf,

in dem Sinne war meine Antwort sicherlich auch davon gepraegt wie viel Frust ich ueber die momentane Situation empfinde, weil ich es hasse wieder weiter vertroesten zu muessen und es so sinnlos verschwendete Zeit ist, in der wir z.B. endlich PiGI produzieren koennten, damit auch Leute, die sich noch nicht ganz trauen das Ding komplett aufzubauen, die Moeglichkeit haben fuer ganz wenig Geld einen netzwerkfaehigen Geigerzaehler zu bekommen. Um zu experimentieren, zu lernen, um zu messen und ein “Gefuehl” fuer Strahlung zu bekommen, z.b. “Wie sieht eigentlich meine Ortsdosisleistung im Normalfall aus?” Bei zukuenftigen nuklearen Zwischenfaellen waeren wir nicht mehr darauf angewiesen die Angaben der Medien zu schlucken. Wir brauchen uns dann nicht mehr fragen ob wir dem Glauben schenken wollen, weil wir unsere eigenen unabhaengigen Messwerte haetten, von denen wir wissen, dass sie nicht durch Manipulation oder Interpretations oder Uebertragungsfehler verfaelscht sind: Open Radiation Monitoring

Ich sehe wie viele Menschen da draussen sind, die langsam aufwachen und sich fragen ob hier alles richtig laeuft, sich von der staendigen Propaganda abnabeln, die ihnen repetiert vorgefuettert wird um sie fleissig als Realitaet zu kopieren und assimilieren. Und all das haben wir letztendlich dem Internet zu verdanken. Wenn man mal durch den ganzen Wust durch gewatet ist, erkennt man am Ende das das Internet uns Menschen in dieser Zeit ploetzlich eine Moeglichkeit bietet die keinem anderen Menschen jemals zuvor zu Verfuegung stand: Unser Knowledge, unsere Skills und unsere Daten miteinander zu vernetzen und zu teilen. In Echtzeit.

Frueher konnten Wahrheit und Wissen von einer kleineren eliteren Gruppe (Kirche etc.) vergraben und “geschuetzt” werden, um die generelle Kontrolle ueber die Definition der Realitaet der Bevoelkerung zu behalten und ueber soziale Strukturen (Kultur) deren Fortbestand und Selbstreplikation (Erziehung) zu gewaehrleisten.

Die meisten damaligen Nerds/Hacker wurden gekreuzigt und als Ketzer verbrannt und nur wenige haben es geschafft Ihr Wissen in hand- geschriebenen Notizbuechern zu konservieren und dann in muehevoller Handkopie im Untergrund zu verteilen und Ihre Erfahrungen mit anderen zu teilen und es brauchte teilweise Jahrhunderte um sich in der Realitaet der Gesamtbevoelkerung zu manifestieren. In der neueren Zeit beschraenkt man sich nur noch darauf solche Leute entweder einzusperren oder irgendwelche lustigen Syndrome zu erfinden, damit man die Leute oeffentlich “krank” erklaeren und deren Gehirnkapazitaeten dann chemisch vernichten kann. Ganz zivilisiert halt. Und macht ja auch guten Profit fuer Pharma. Win-Win fuers System.

Dafuer koennen wir heute in Echtzeit mit jedem Menschen auf der Welt der Zugang zum Internet hat unsere Erfahrungen, unser Wissen und sogar einen guten Teil unserer Werkzeuge teilen. Man muss ja kein Genie sein um zu erkennen, dass die gegenwaertige Entwicklung nur der Anfang der Beschleunigungskurve ist, da wir jetzt erst langsam feststellen wie grossartig das eigentlich ist und Tools entwickeln, die uns noch einfachere globale Zusammenarbeit und Teilen ermoeglichen.

Deswegen laeuft hier ja auch eigentlich alles in Englisch :) Nationalsprachen sind eine zu grosse Huerde, man muesste jedes Stichwort, welches man sucht, im Kontext uebersetzen und ueber alle verfuegbaren Nationalsprachen (und deren Kontext) suchen um alle Ergebnisse weltweit zu bekommen.

Zum Thema Open: Ich wollte nen Geigerzaehler haben, damit ich mir mein eigenes Bild machen kann. Meine einzige Quelle fuer alles was ich gebraucht habe um dieses Projekt zu realisieren war das Internet, d.h. ich war nur in der Lage das Wissen zu erlangen, welches ich gebraucht habe, um PiGI zu entwickeln, weil andere bereit gewesen sind Ihr Wissen und Ihre Erfahrungen zu teilen und mir zum kopieren angeboten haben. Ist doch am Ende dann mehr als logisch, daß die gesamten Infos zu dem Projekt und damit der aktuelle Wissensstand dazu auch wieder ins Internet muessen, damit jemand anders es 1:1 kopieren und bauen oder mit dem Wissen das eigene Projekt vorwaerts treiben und auch wieder publishen kann. Ist doch nur wichtig, dass beim Teilen fuer ALLE der Grad an Freiheit erhoeht wird und wir jedesmal ein Stueck schneller vorwaerts kommen mehr Alternativen zu vorhandenen Systemen und Strukturen zu finden und entwickeln, ohne auch nur ein Joule an Energie oder eine Sekunde wertvoller Lebenszeit damit zu verschwenden, sinnlos ueber ein kaputtes System zu jammern oder gar zu versuchen dagegen anzukaempfen :)

Wenn Du erkennst, dass das Pferd, welches Du reitest, tot ist, steig ab.

Altes Indianersprichwort

asdf
2014/12/28 23:39

Danke für deine ausführliche Antwort. Ohne auf alles einzugehen (weil wird sonst auch viel), aber zur Gesundheit weil so wichtig, denn wenn man krank wird, gibt man bekanntlich all sein Geld aus um wieder gesund zu werden: Das mit der Big Pharma stimmt auch. Und zu “Syndrome zu erfinden”: Das geht noch weiter: Symptom nennt man etwas von dem man nicht weiß was dafür der Grund ist. Z.B. gibt es Leute, die angeblich HIV haben, die raus gefunden haben, das es das HI-Virus (HIV) gar nicht gibt. Auch einige Nobelpreisträger sagen, dass es das nicht gibt. Man stirbt an den Nebenwirkungen der Medikamente die genau die gleichen Symptome auf dem Beipackzettel beschreiben, die das angebliche HIV haben soll, man stirbt also an diesen. Ein Virus der erst nach >20 Jahren ausbricht, genau, für wie dumm halten die da “oben” uns eigentlich? – bloß keinen Test darauf machen lassen – Ähnlich mit Hepatitis und einigen anderen angeblich existierenden Viren. Die da “oben” erfinden irgendwelche Krankheiten (es wird z.B. auch kritisiert, dass man mit dem Buch DSM-5 mittlerweile jeden als “psychisch krank” einstufen kann, wobei das zwar jetzt die psychischen Sachen sind, aber das schneidet sich) und profitieren dann nicht nur von Geld sondern von Kontrolle. Andere Themen-fakes sind hier exposed: https://www.youtube.com/user/freeradiorevolution/videos , richtig ist geschlossen worden, aber wende diesen Benutzernamen in die YT-Suche eingibst, findeste die Videos/den User wieder. Anderes: Impfungen und Autismus, oder auch solche Sachen wie: “Kaiserschnittkinder werden meist vor der 40sten Schwangerschaftswoche geholt. Zu früh sagen Experten: eine Studie aus Schottland zeigte, dass sich dadurch das Risiko für spätere Schulschwierigkeiten verdoppelt. … Auch Hebammen wird das Leben schwieriger gemacht.” – absichtliches dumm-machen der Menschen.

Zum Thema Wissen teilen: ich hätte den tragbaren Geiger auch nicht so schnell bauen können, hätte nicht jemand schon die libs fürs LCD und basic Code für den MSP430 gestellt. Ich habe ihn auch gebaut, um, wie von dir erwähnt, nicht auf die Angaben der Regierung warten zu müssen bzw. diesen dann vertrauen zu müssen, denn die Vergangenheit hat gezeigt, dass diese entweder nicht stimmen, oder diese nix raus gibt. Das was mich an PiGi auch angesprochen hat sind die relativ niedrigen „<2mA” at background. – Und: schon Jodtab. zu Hause für Jodblockade? Ich glaube und hoffe, dass diese nie nötig sein werden, aber man weiß ja nie. Typische Jodtabletten sind P/L mäßig viel zu teuer und was wenn schon ausverkauft wenn gebraucht bzw. es ist dann möglicherweise eh zu spät und man muss diese schon vorher zu Hause haben. Den Wirkstoff, Kaliumiodid, gibt zig fach günstiger wenn als Chemikalie gekauft. Einfach “Kaliumiodid reinst” in die Suchmaschine eingeben, dann ist man noch weniger auf die R. angewiesen und muss nicht warten bis diese die J-Tab. verteilt. Übrigens: In Japan nehmen die Menschen täglich über 12mg (12000 μg) (Küstenbewohner der nördlichen Insel Hokkaido sogar, 200 mg Jod täglich) zu sich und wir hier sollen mit nur 100-200µg gerade mal den Kropf verhindern (in Amerika ist der tägliche Jodbedarf sogar noch weiter runter gesenkt worden, die Menschen da sollen wohl echt krank gemacht werden – Amerika am abbauen?)…(Jodmangel Symptome googeln sage ich da nur). Den Menschen fehlt aber noch lange nicht nur Jod – Krankheiten nehmen zu. Also keine Angst vor Überdosierung – kaum möglich – laut Studien haben die meisten außerdem einen Jodmangel, so dass wahrscheinlich auch Du dazu gehörst.

Anderes Thema: laut dem aktuellen Vortrag von Jacob und Laura (Reconstructing narratives) auf dem 31C3 sind GnuPG und OTR die einzigen mit relativer Sicherheit von der NSA ungeknackten Standards/open source Tools. Einfache Mathematik mit sehr großen Zahlen und keinen möglichen Schwachstellen – diese Tools wären deswegen auch mein Tipp gewesen.

ptarjan
2015/02/02 16:57

People, as this does not seem to be going forward anytime soon, I have an offer for the first person to reply here. I had the PCB made from the schematics, and I have a spare in a lovely purple color. I would like to exchange that for the parts on the Mouser BOM. (The problem with ordering them directly is that the shipping is a ridiculous $40 to Europe.)

So: I offer to send the PCB and 10 USD (PayPal) to someone who will send me all the parts from the BOM at mouser. No fancy shipping, just chuck them in a padded envelope - I'll do the same with the PCB. Any takers?

Peter

chrono
2015/02/02 19:35

Yes! Great! It is unfortunate that suddenly my ability to survive has put a break on this project because my goal here in the end is to help other people to get easy and cheap access to DIY geiger counters. That's why everything here is actually open-source, so everybody can take the already tested and prototyped board file and give it to a manufacturer.

But I fear that if I am gone, all of this will be gone too because no one else is stepping up to drive it forward or even just keep it alive. So I have no other choice than re-prioritizing and focusing on getting the means and necessary conditions to keep Apollo-NG and all of its subprojects alive to finally be able to push projects with full focus and not being deprived of my time by the bullshit of world.

In the meantime, I'd love to see you help each other. From the statistics of this page, there are a lot of you out there looking at this project and I'm sure if you guys would connect it would be possible to find a way to share load, risk and time. If we need a mailing-list for that purpose I can set one up very easily to coordinate things a bit more convenient (if there are more than two people actually willing to chip in with time). We already have a full stack of working distributed collaboration means, from etherpads to live video conferencing to quickly coordinate efforts.

It's just that at this moment, it's impossible for me alone to handle all involved tasks and the usual perils like part shortages, delivery delays, customs delays, RoHS, REACH, CE, European Product Safety Regulations and all the other system-related inhibitors which could lead to fuck-up and need to be managed when scaling up numbers. But I'm sure we could pull of a small series when we do it together.

Another way to make this happen of course if mailing and buggering adafruit, seeedstudio and the like. Seriously, they're just doing business and as such, if they see a clear demand (coming from more than just one person) and we can come up with a production scheme with them it would be win-win for everyone: We would get these modules cheap because we can produce numbers and also have an already implemented payment, production, QS and fulfillment infrastructure, shipping world-wide.

Kro
2016/02/04 09:36

Hello Chrono,

I will manufacture some PCBs and test them. Are You shire, the RPi caches all hits of the tube (while not in dead time) ant isochrononly? (Linux is not known to be real-time-enabled.) Might it be better a better solution to use a small MC to stricly - catch all hits - on an isochronous scale ???

Then the MC would transmit the data to the RPi, for example by SPI oder I2C.

Peter

chrono
2016/02/04 09:47

That is a valid concern I also raised early in development of the PoC: The primary bottle neck I see is python itself. I've always opted for a C based interrupt counter that just counts and gives the count stats via RPC or a socket to the python framework in order to mitigate these issues.

So in short, the higher the count, the more likely it will be to have lost counts that fall either victim to dead-time issues or more likely to the Pi maxing out its python/script based interrupt handling capability. A tinyAVR or some other small, SOIC-8 based uC could mitigate these problems.

However, most really expensive “professional” geiger counters have two GM tubes: One for low- and one for high-dosage counting to address that problem, hence the idea to make them stackable to have the same ability.

In any case, in the real world, the more this becomes a problem, the less likely survival is anyways. Maybe something for 2.0 or for your adaptation efforts. The design goals of the original implementation were simply to make it as cheap as possible, with the lowest BOM-Count and the least amount of complexity, so that anyone hacking around in a garage, ghetto or whereever could copy the concept and plans in order to have at least something to measure your normal LDR (Local Dose-Rate) or if the levels get to something far less comfortable.

It's not designed to be a scientific, high-event, high-accuracy radiation measurement device, but to give basically anyone the ability to identify, verify and get a ballpark number of ionizing radiation being subjected to.

ptarjan
2015/05/05 18:16

I have an interesting plot to share: https://www.dropbox.com/s/l8cqwqqnymf98g4/geiger_time.png?dl=0 This is data from the PiGI board and a Philips 18504 Geiger tube. As the title says, it shows in a histogram the distribution of time intervals between two consecutive hits. (I made it log-log so that I could span several orders of magnitude.)

The first thing you can see is that the tube's dead time is ~84 microseconds (where the histogram starts) or ~109 microseconds (where it has the huge peak), depending on what you call dead time. Both are realistic, in any case.

The second thing you should note is that this distribution should be strictly exponential beyond the dead time. Well, it is, though not quite from the beginning: the red line is an exponential fit.

Third: where the hell do the two big dips (one between 100 and 200 us, the other between 300 and 400) come from? Physically that can't be right…

Fourth: Geez, what is up with the periodic structure??? From the histogram those peaks are spaced around 125 us apart, which translates to a frequency of about 8kHz. To me that suggests a slight ripple on the output voltage which makes it periodically easier & harder to trigger the tube. Is 8 kHz maybe the coil charge/discharge frequency?

I would like to stress that the PiGI does very nicely what it promises to do, i.e. drive a Geiger tube and catch the pulses. I'm just geeking out and trying it to understand the inner workings of both the board and Geiger tubes.

Peter
2016/03/09 23:21

Hi Chrono,

You are right, I do not need no calibrated geigercounter. Now my first pigi with a RFT VA-Z-114 runs, tested by a scope. Next week I will go on with it and install the software. Did You find a new home?

Thank You, Peter

PeterKa
2016/11/17 10:00

I have running about 10 stations with FHZ76V and five more pigi-cards. Those five do not run, because their voltage drops and do not reach 500VDC. Any idea?

Peter

chrono
2016/11/19 09:27

Hmm, let's see:

1. Have you used some giga-ohm resistor to actually measure the voltage (as precise as we actually can) or are you assuming it's not enough because those five just don't register any counts?

2. Are the components of these 5 from a different source/batch which may have lead to mixed up part values or fakes/relabeled 555 to be sold as TLC555 (there is a difference).

3. How long are the connection wires between tube and pigi?

4. Have you checked if the 555 is actually switching on its output?

5. Have you tried adding a 10M resistor between the tube's anode and R3?

See picture:

chrono
2017/01/24 07:18

For those who want to have a PiGI, but don't have time or skill/confidence to roll your own PCBs and do SMT soldering:

We've been contacted by Seeed Studio and used the opportunity to pitch PiGI to them as a non-profit deal again, asking that they take the PCB files and BOM to produce and sell PiGIs in their webshop for a fair price so that more people could get PiGIs easily.

We've left it open to Seeed Studio to compensate us for time/resources spent for development and testing (basically become a sponsor for Apollo-NG and allow us reasonable access to PCB prototyping or products we require for further development of projects within a certain percentage of their profit margin of PiGI sales). That would be great so that you guys would not only get PiGIs but your purchase in Seeed Studio's shop would also help us. Win-Win for everyone.

However, we've made clear that this proposal is by no means a show stopper. If we don't get anything at all, that's okay too, as long as we can spread more PiGIs and give people the opportunity to have networked ionizing radiation measurement gear for a very low price compared to deprecated industry products.

HardNut
2017/01/24 08:58

Peter
2017/01/24 14:31

This is great news, when do you think we can actually buy them and how much are they going to cost?

Funky
2017/03/06 20:00

Nice job! I would definately buy one.

alarmschaben
2017/03/09 09:58

Nice design! I'd actually like to buy not one PCB kit with components, but five. I was successful in persuading some of my friends do build the same.

Comment text

Please type the letters you see on the left into the box on the right:

   __   ____    ____   __ __   ____
  / /  / __ \  / __/  / //_/  /  _/
 / /__/ /_/ / / _/   / ,<    _/ /  
/____/\___\_\/_/    /_/|_|  /___/