Ion Chambers and Geiger

Comments and ideas for the measurement of Radon and radiation

Two SBM-20 in parallel and a GeigerAdapter form a simple probe, very sensitive and inexpensive.

This image shows a measure of environmental fund with Theremino_geiger

More information
– Software download:
– Electrical schematics and Assembly plans:
– Hardware, DIY and kits:
– Images and videos:
– Article about Open Source Electronics:

In the documentation folder (accessible from the Help menu) you will find all the information you need to use the program and probes.

The HAL program, required to connect to the Master, you download from here:

Measure with Geiger

This image shows a measure of environmental fund 15 minutes, followed by the provision of a sample of dietary salt. At the beginning of each measure was pressed “Start”

Precision measurements you make with long integration time. While precision measurements you do not have to change the conditions of measurement. Move the samples during the measurement would be like taking a fever without taking the thermometer in the same spot for the three minutes required.

All ’ start of the measurement You should press the button “Start new measure” that is how “Reset the thermometer” Before measuring a fever. Without this initial zero setting it is likely that the fever has not indicated its, but someone else.


To make precise measurements You must respect the agreements with the program, the instructions are explained, in detail, the procedures to be followed.

To make quick measurements, just set the “usual” 2 or 3 minutes of integration, using all portable geiger..


Some users would like to avoid having to follow the measurement protocol. In essence they would like a button “Start” you press alone, any significant variation of the radioactivity.

Early versions of Geiger had this option, but we found that this made the measurements unreliable and bore, in some cases, completely wrong measures, without being able to inform the user ’ and without any possibility of control by software. These errors were not due to an inaccurate implementation of ’ algorithm, but are inherent in the measurement method and any appliance that is automatically reset can avoid them.

As far as l ’ algorithm is sophisticated and set the trigger thresholds, There are always a few sequences of changes that lead to a zero delayed, resulting in pollution of the first part of the buffer of integration, or to a non-zero or, even worse, to reset drivers, they do lose hours of data.

To miss any appliance that resets by itself, proceed as per the following example:

1) Let it settle to the bottom

2) Bring a sample very radioactive 20 cm so that it resets and empty the buffer

3) Continue moving the sample up to contact, but very slowly, so as not to over take the reset threshold.

The first part of the buffer will be filled with samples too low, the threshold it clicks more and the measure will continue to be systematically flawed, How long you have the patience to wait.

This example sequence is extreme, but there are many real opportunities to attract more or less large errors, If you allow ’ user to move samples at any time, or add and remove fragments in a sample already to an extent.

The devices designed for measuring “on the field”, where it is important to react quickly, are not suited to precision measurements and can, in many situations, provide incorrect measurements.

Measure with the ion Chamber

Ion Chamber documentation files do not show enough that, in some cases, the pulse rate can be very low.

In environments with little radon (d ’ summer with wind and opened Windows) Radon concentrations may fall under the 0.1 PCI/l (4 Bq/m3). In these cases, you get even less than one pulse every 5 minutes. It can also happen to not measure no pulse for ten or fifteen minutes. The randomness then back to place the media, concentrating many impulses in 10 minutes after. The next image shows this behavior.

In this picture have gone well 9 minutes, before receiving the first impulse. In the central zone is also a period of 15 consecutive minutes without pulses.

That is why we designed a very large room (from 1 liter). With many commercial appliances impulses would also 100 times less frequent. For example a “Safety Syren Pro3” (a few cubic centimetres), would produce on average, Pulse every 7 hours and could spend long periods, even more than a day, no pulse.

Test of Thoron

If you live in an environment with little radon, the low frequency of impulses might suggest a malfunction of the room. Fortunately there is a Thoron, a similar Radon gas, but much more convenient for testing. More information about the test of Thoron, on pages 21 and 22 of the document “Radon_IonChamberElectronics”, download from here: /hardware/inputs/radioactivity-sensors

In the first part of this picture you see the Environment Fund. In the second part the mantle containing Thorium was positioned in the saucer, and been pressed Start. In the third part, the Start button was pressed again and in a few minutes the measurement has stabilised at about 5 PCI/l. In just over half an hour ’ l ’ error fell to 10%.

The gas mantles containing thorium are not all the same and the room may not fill completely due to leakage between the saucer and the room. Then the measured value may be different, Come on 5 PCI/l of this picture.

If you measure from 2 PCI/l to 10 PCI/l the test passed, Otherwise please email us.

In this picture the mantle with thorium and saucer were removed. L ’ indoor air was grossly cleaned up, waving a cardboard for a few seconds. Then the Chamber was lying horizontally, so leave the holes open to both terminals and pressed the Start button. In just over half an ’ now the room is completely cleaned up and returned to 0.1 PCI/l.

The ion Chamber should always be handled with delicacy. The bursts of pulses that are produced when the room is shaken, mild vibrations are produced by central wire and do not indicate a malfunction. Even dust bunnies can cause bursts of pulses, You may need to add dust filters. Read pages 25 and 26 of the document “Radon_IonChamberElectronics”, download from here: /hardware/inputs/radioactivity-sensors

  1. Paul says:

    Hello,putting the form geiger in an inner box and externally smb20 probe what cable should I use ? the maximum length between geiger and probe is 3 meters.

    Thank you

    • Livio says:

      Geiger Geiger tube Adapter and module must be together and together form the probe. The maximum distance between the GeigerAdapter and the Geiger tube is a few centimeters.

      Many projects that are on the network connecting the Geiger tubes with cables and connectors. Try, the pulses arrive, and they're happy as well. Even our GeigerAdapter work. But in doing so the Geiger tube must suffer and his life would be reduced.

      Parallel tubes manufacturers specify that there must be the minimum possible capacity. A cable, for example, three meters, It has a capacity from 300 AI 1000 pF. This ability is charged with 400 Volts, and you download onto the pipe at each pulse, generating very strong currents that “FRY” the pipe in a short time.

      Conclusions: Maximum 2 or 3 PF in parallel to tube. Up some centimetres of wire. Better to use unshielded wire. Better than the positive wire is as short, then turn the tube with the positive toward the Geiger Adapter.


      Attention: Geiger pipes outside wants a shed for the Sun, otherwise the container may exceed 40 – 50 degrees, In addition to which the Geiger go crazy. Of winter, the cold condenses moisture and high tension starts to get empty. At a minimum, the container must be airtight and you have to put the bags of Silica Gel. But even with all the care may not necessarily be able to make it work long.

      Our probes are made to make laboratory measurements, or with portable measuring set. When not in use should be stored in a dry place and reasonable temperatures. Environmental radioactivity measured outside is pretty much always the same, useless place a fixed probe outside. In the case of a nuclear accident there are already thousands of active and always available on the net.

      • Paul says:

        can I put it all out… though the probe want put outside the box to limit the length of unshielded cable will cm20, technically goes fine?

        • Livio says:

          You can do it, But why have two separate pieces with 20 cm of wire in the Middle, When you close it all together in a single probe.
          – A unique probe is easier to perfectly seal.
          – A probe with Geiger and Geiger Adapter is a complete unit, interchangeable with other sensors.
          – High voltage must stay all in probe. No wires with high voltage around. Much less if you think placing it outdoors.
          – Multiple threads are long and more electromagnetic fields induced by lightning increases. With twenty centimetres of wire, instead of two centimeters, the probability of counting lightning increases tenfold.

        • Livio says:

          Anyway “technically” You can do it. Twenty centimeters of unshielded wire make about 20 PF and not pain too the hose. If you use due care, mechanical seal on all, then okay.

  2. Luciano says:

    Hello Livio,using it as a smb20 probe on 60 CPM measured how many uS/h would be?
    CPM 60 = uS/h 0.3420 is correct?

    Thank you

    • Livio says:

      And’ more or less fair,
      If you have a tube of CPS/mR/h = 29 and don't count the BKG tube, 60 CPM make 0.348 us/h
      otherwise a BKG of 0.2 (normal for SBM20), make about 0.28 us/h.

      I to do these tests use the script “Geiger Simulator” you find examples of Theremino Script (to fill out and keep handy). Then I reckon by Geiger and not risk to make mistakes.

  3. Lucignolo says:

    Hello Livio
    I made two geiger with 4 SBM20 tubes each, one and’ mounted in a plastic container (and’ posted at Theremino the page after this) and the other in a die-cast aluminium housing with a thickness of 3/3, 5 mm, This for “block” beta rays and only detect gamma rays. I had placed two whiskers geiger nearby and set the bottom in CPS to have the usual value in nS/h of 85 in two graphs, This regulation and’ lasting several hours. The values in nS/h of two geiger (clearly closed in their respective containers) are different especially if it rains, and’ major in geiger in plastic container, While the CPM value and’ minor. I opened two Pulse counter and actually counts reflect the CPM.
    How come the CPM value and’ greater where nS/h and’ minor and vice versa, There’ an explanation?
    Thank you

    • Livio says:

      You set two different values in the BKG their tubes, the BKG greater subtracts more and is less uS/h although the CPS are more. I suspect you're asking too much accuracy in pipes that are allegedly made to work by many uS/h up. Virtually all Geiger tubes are made to find out if it's time to get away and it is already so be able to measure the environmental fund with a fair approximation.

      We've written repeatedly that the measures you can do are related. I.e. compare and say “This is a little’ more radioactive than this one…”. You can't measure the Bequerel and the accuracy that can be obtained are always very coarse.

      To have a minimum of precision is necessary to know well their probes and understand how they behave under various conditions. I'm sorry but I don't know tell you why these things happen, You must understand you doing tests. Probably just variations of geiger in time or parameter setting errors. What I can assure you that this is not software defects or counters, or USB or serial transmission.

      I once did an experiment, I measured the bottom BKG a SBM20 and then I tapped and rocked long. After this treatment the BKG had fallen very, If I remember correctly, almost 50%.

      This is not to say that pre-condition the tubes with a vibrator (also because I do not know what would happen the next day) but I mean that you should not expect much accuracy by Geiger (not because they are of SBM20, but precisely because they are Geiger tubes)

  4. The University of Regensburg organizes a conference on radioactivity twice a year since the Chernobyl accident. The conference in Autumn 2015 shall be dedicated for the first time to amateurs measuring radioactivity. In the past we had professionals only. The development done by amateurs is impressing however. Theremino MCA for example makes this technique, which was rather expensive in the past, available for hobbyists. We invite speakers to report about such projects. There is also an exhibition of measuring equipment.

    The location of the conference is a museum near Regensburg, Germany. Conference language is German, a talk in English would be appreciated also. Participation is free of charge.

    You can view the topics of the past conferences on this website:

    This is rather far to travel for people from Italy or France, We have participants from this countries however each time. Maybe there are also serious amateurs in Germany who read this. If you are interested as a speaker or participant please contact:

    Juergen Putzger
    Universität Regensburg
    Fakultät Physik
    93053 Regensburg


  5. Paul Zoso says:

    It would be a complicated thing, add a function to the software geiger to send mail when it exceeds a certain value or radiation level?
    I'd leave it on continuously to monitor environmental radiation and can configure an alarm.
    I'd also like to be able to load all the data in a database on my Synology nas and create a website (always on synology perhaps with joomla) with the real-time graph of radiation (, but unfortunately not an expert programmer and I wonder how much might come to cost developing a thing.

    Thank you

    • Livio says:

      Since the system Theremino is designed to separate modules, It is not fair to change the Geiger, for each new requirement. Would be fairer to set it, to write the data of radiation, in one Slot (the Slot to use you set in the options panel). Then it would take, two lines in Theremino script, to send an email or do other operations. You should find a friend who knows how to program. Unfortunately few of us and we cannot. You wouldn't know, but schedule requires a great deal of time, even for tasks that may seem small. For example, to act on something called synology and joomla, You should buy them and know them… whereby weeks of studies and experiments, at a minimum.

      With regard to send data on a site, the Geiger does it normally (Panel Logs), and there are dozens of sites, that they are doing.

      Here's some links:

    • Livio says:

      I missed the “how much would cost”.
      For this question you should contact a professional programmer, We do only free and Open Source projects.

      But there is good news!
      We are preparing the instructions to send mails with Theremino Automation. With the new version you can simply write the following type three lines:
      If Slots(1) > 1000
      SendMailTo foo dati_smtp testo_da_inviare_nella_mail password it at
      End If

      So if the Slot 1 (prepared by ThereminoGeiger) will exceed 1000 us/h, the mail will be sent.

      • Paul Zoso says:

        OK thank you very much , I'll wait for info to use Theremino Automation, in the meantime I will see if I can get something, If I will share it all here.


  6. Paul says:

    Salve, I was interested in understanding how estimates the error range of Geiger. I thought it was up by the square root of the number of counts detected but from source I saw that and’ so’ only between 100 and 10000 pulses (even though from the comments to the code it appears that initially were not present this distinction). There’ any particular reason why under 100 and above 10000 and’ was decided to use a different Predictor?
    Thank you and congratulations for your program.

    • Livio says:

      The statistics says that the error is in proportion to the root of the number of pulses and this is definitely right, But what is meant as an error can?

      We talk about errors that will not be exceeded in 90% of cases, or in 99% of cases? If you intend to errors that will not be exceeded at (even with the most unlikely that could happen every 1,000 years) then the values would be much higher, of percent we are used.

      What we were interested in showing Geiger, It is not a pure as mathematician, but a percentage that, “in most cases”, was consistent with reality. That is a fact that it was a good estimate, What to expect in the next minute, or in the half hour following, or in the three hours following (Depending on the sample activity).

      This is basically to predict the future…

      And we have verified in practice that, for very high counts (> 10000) and very low (<100), the reality didn't match the theory. We approximated roughly these two cases, with root fourth and a linear relationship.

      There are two alternatives commented in green, You can try them, the third is exactly equal to the statistical theory.

      I don't know tell you why our formula works better than pure statistical theory. Probably the "risk" to err, perceived by humans for short periods, is not the same as the "risk" for a long time. In other words you must modulate "confidence" at 90% or to 99% or to 99.9% Depending on whether the seconds, minutes or hours.

      If you find better explanations write them here in the blog, for the good of all geigeromani.

  7. Marco says:

    some time ago I came across this site accidentally and I was pleasantly surprised by the technical level of proposed projects. I recently developed a sensor and an app for Android (RadMeter) functioning much like Geiger. The app is free and can run on virtually any sensor (using interfaces such as Geiger Adapter or Checker :-)).
    I am convinced that this app has since now features over Geiger Pocket, used here in the past, but I think it can be further improved by suggestions or recommendations collected within this community.
    I would be very happy if RadMeter could be mentioned within this site or if it could be proven by users thereminici.
    In any case, I congratulate you for the great job you're doing outreach.


    Here is a link to the app:

    • Livio says:

      Very nice application, the'll link each other soon with some pictures, Maybe the page of GeigerAdapter. Send an email to Alexis that he often uses smartphones for radioactivity and write surely his tests on our Facebook and Twitter pages.

      Before proceeding, however, you should explain two particular:
      – The sensor SS05 that sensitivity has? (CPS/mR/hr)
      – How much is it? (If we're going to advertise would be nice at least to someone writing prices – products without prices are lousy – only in Italy we have this mania to hide everything)

      And then… no hurry, but in some future release… would it be possible to add the BKG to each tube and take them into account in calculations? Otherwise the bottom of scale (that is the most important) suffers from an exaggerated non-linearity and provides values systematically mistaken (even the 200 or 300%).

      • Marco says:

        Thanks Livio for availability and for the speed of response.
        Here's my replica:

        >- The sensor SS05 that sensitivity has? (CPS/mR/hr)

        >- How much is it?
        I agree with your account, in fact the cost in question is written on the site :-) and is established when 59 Euro (See also
        I would like however to point out that the app is not bound to using this sensor, It is present in the internal database on a par with other sensors and Geiger tubes without any preference of treatment. On the site I tried to give directions, and if necessary I will look into the matter, regards the interfacing with different types of sensors application hardware.

        About the value of BKG it is already accounted for in the calculations and is present in the database of the sensors. Because I felt very valid program Geiger, I tried to make the app so that it was very similar to it and was of immediate use for users already accustomed to this program. Printing errors, the calculation algorithm of RadMeter should be equivalent to that of Geiger.

        If Alessio deems interesting this app and want to delve into any technical detail about it can contact me freely in pv via email.


        • Livio says:

          I hadn't seen the price (a little’ Altino if it is a single BPW34…) nor had noticed the FAQ page. There are now more info, glad, Thank you!

          Very good your program, you added the BKG! And you notice that responds well, makes a nice chart very significant (practically the same visibility of Geiger), has the logarithmic scale and also does error bands! Very nice!

          And also very nice round logarithmic scale!

          You will do a good review but to be clear before on some particular. We do not like recommending items to 50 Euro, If we are not sure that go well. How can they be 5 CPM/mR/hr?
          From our trials it would take dozens of BPW34 in parallel and the total parallel capacitance stoop so much the pulse height, making it unstable and precarious the limit between pulses and noise.

  8. Joseph says:

    Hello Livio, I don't know how to say this….but really congratulations for the work you do. I read the manual of the CNC latest version and you have done a tremendous job. I can't wait to get to work on CNC. Meanwhile I'm still ravanando on theremin, the VST etc and I'm having a great time.
    Two things, however, I wanted to ask in reference to Ionic room for radon.
    I saw a while ago that you are working with one Chamber in stainless steel (and I saw that you have taken the copper rods 2 mm…I while ago I suggested to use them but I recommended the square ones that weren't going well):
    1) What do you use as an insulator of the wand?
    2) There is a tutorial for the construction of the room in stainless steel or can I follow mutatis mutandis the old one?

    Thanks again…Joseph

    • Livio says:

      We use chopsticks to brass tube 2 mm in diameter. It is found in the shops of modelling, for tubes of models with petrol engine. To isolate them okay any plastic or nylon with a 2 mm hole. Sure the finger FET, must be completely inside the inner Chamber, fully shielded from electronics and from external electric fields. Brass is an excellent material for electrode and the advantage of a pipe, compared to an electrode full, is that the relationship between rigidity and weight is much greater, Why shaking the room pulses do not occur.

      Ion Chamber treatment of mechanical construction is even more important than what you thought at the beginning. A room at ions with slight defects (residual dust of construction, blunt tips, rough parts, Rusty or not rounded) It can work well for a long time and then stop working, just the shakes or humidity exceeds a certain level.

      For the construction of the room there is still no stable project, mechanical solutions have been tried, all difficult and also horribly uncomfortable to fit the electronics.

      However the research continues, a test Chamber is at the harp to verify the calibration and maybe we can even eliminate double room outside and keep the exterior in massa. For now let us not create too much confusion, with intermediate versions that are outdated before they even finish it. As soon as we publish them valid and definitive solutions. It will take months, as it took for the CNC, but worth the wait.

  9. luciano says:

    A salutone to all,After quite a long time I rewrite on this blog for chiederen to Livio information this time on the raspberry.
    I saw that you created a img (ENV) for this mini pc,the question is also works with Geiger?.
    I still in the drawer of his master and geiger hired time and would like to experience what.
    the raspberry is the new one 1 Jig with 4 CPU-



    • Livio says:

      Until last year, between Mono, Libusb, coprocessor not implemented and slow running, use the Raspberry was so difficult and uncomfortable that none of us has continued the tests. Then last year came out Tablet, just more expensive than a Raspberry, that give so many accessories and performance that they prefer in any situation. Even if only for the fact that having it all do not bind the USB ports for Wifi and Mouse. But especially because, having a true Windows, the installation is easy and fast and precise execution. Besides, it also becomes possible to use Visual Studio and debug the software. On Raspberry you should use “Mono Develop” that is extremely uncomfortable, If not impossible to use, for debugging.

  10. Livio says:

    The new version 6.3 by Geiger is here:

    The improvements to the hardware of the ions are here:

    The discussion on measuring units “Concentrazione” and “Activities” is here:

Leave a Reply

Your email address will not be published.