RadiaCode-101 (радиакод-101, RadiaCode-102, RadiaCode-103)

[Radiacode-101]

Ja

Please feel free to ask me if you have any questions about the device and its specifications. There's no need to make assumptions or reason logically when I can provide you with all the specifications.

It's very simple!

RADIACODE, what is the energy range for compensation?

Answer:

The compensation is calculated up to 3 MeV, and anything above 3 MeV is considered as 3 MeV in the last channel.

itte zögern Sie nicht, mich zu fragen, wenn Sie Fragen zum Gerät und seinen Spezifikationen haben. Es ist nicht notwendig, Annahmen zu treffen oder logisch zu argumentieren, wenn ich Ihnen alle Spezifikationen liefern kann.

Es ist sehr einfach!

RADIACODE, bis zu welcher Energie reicht die Kompensation?

Antwort:

Die Kompensation wird bis zu 3 MeV berechnet, und alles über 3 MeV wird als 3 MeV im letzten Kanal betrachtet.

[Radiacode-101]

Das bedeutet aber, dass derzeit drei Dinge sind falsch sind, einmal suggerieren die Geräte-Spezifikation,

It is evident that you want to present your desired observations as actual data. First and foremost, I would like to point out that the random measurement error is calculated for each channel separately. The uncertainty in the count rate is determined as 1.96 divided by the square root of the number of pulses, not as you mentioned, 1 divided by... Since we calculate the uncertainty for a 2-sigma confidence interval, which corresponds to 95%, the uncertainty calculation follows this approach. For the dose rate, the random error is calculated based on the sum of the pulse numbers in the channel, taking into account their contributions to the dose rate. This means that until the minimum number for the entire range is reached, with their respective weights, there will be a high uncertainty in the dose rate channel. I did not understand what you meant by "uncertainty" in your message and why you divide activity by the square root of the number of pulses. It seems like a translation issue.

Energy compensation works for energies calculated for the last channel, typically around 3 MeV on average. A small variation of +-200 keV will not significantly affect the dose rate readings. Regarding the low number of registered pulses, that is a separate issue. The absorption efficiency of the body will be significantly lower than that of cesium iodide due to the different densities. Following your logic, detectors based on Geiger-Muller counters should be highly inefficient. The detection of the thallium-208 peak does not affect compensation because the field density does not depend on the detector's size. The energy distribution extends from the maximum value of the photopeak to the Compton photons generated in the scintillator and shielding materials.

I understand that everyone is concerned about deviations in readings due to background radiation. However, it should be noted that these values are generally no more than 30%, which is acceptable for a consumer device. Of course, I understand that we will continue working on refining the compensation mechanism (we already have some proposed solutions). However, after comparing the energy dependence with professional instruments, we concluded that the differences in readings due to energy do not exceed the values typically specified in the specifications.

Es ist offensichtlich, dass Sie Ihre gewünschten Beobachtungen als tatsächliche Daten darstellen möchten. Zunächst möchte ich darauf hinweisen, dass die zufällige Messabweichung für jeden Kanal separat berechnet wird. Die Unsicherheit der Zählrate wird als 1,96 geteilt durch die Wurzel der Impulszahl berechnet, nicht wie von Ihnen erwähnt, als 1 geteilt durch... Denn die Unsicherheit wird für ein 2-Sigma-Vertrauensintervall berechnet, was 95% entspricht. Für die Dosisleistung wird die zufällige Abweichung anhand der Summe der Impulszahlen im Kanal berechnet, unter Berücksichtigung ihres Beitrags zur Dosisleistung. Solange die Mindestanzahl für den gesamten Bereich mit ihren jeweiligen Gewichten nicht erreicht wird, wird es eine hohe Unsicherheit im Dosisleistungskanal geben. Ich habe nicht verstanden, was Sie mit "Unsicherheit" in Ihrer Nachricht gemeint haben und warum Sie die Aktivität durch die Wurzel der Impulszahl teilen. Es scheint sich um ein Übersetzungsproblem zu handeln.

Die Energiekompensation funktioniert bis zu den Energien, die für den letzten Kanal berechnet werden, im Durchschnitt beträgt dieser Wert 3 MeV. Eine geringe Variation von +-200 keV wird die Dosisleistungsmessungen nicht wesentlich beeinflussen. Was die geringe Anzahl registrierter Impulse betrifft, ist das eine separate Frage. Die Absorptionseffizienz des Körpers wird aufgrund der unterschiedlichen Dichten erheblich geringer sein als die von Cäsiumjodid. Wenn man Ihrer Logik folgt, sollten Detektoren, die auf Geiger-Müller-Zählrohren basieren, äußerst ineffizient sein. Die Detektion des Thallium-208-Peaks beeinflusst die Kompensation nicht, da die Feldstärke nicht von der Größe des Detektors abhängt. Die Energieverteilung erstreckt sich vom maximalen Wert des Fotopeaks bis zu den Compton-Photonen, die sowohl im Szintillator als auch in den Abschirmmaterialien erzeugt werden.

Ich verstehe, dass alle besorgt sind über Abweichungen der Messungen aufgrund der Hintergrundstrahlung. Es sollte jedoch beachtet werden, dass diese Werte im Durchschnitt nicht mehr als 30% betragen, was für ein Gerät im Haushaltsgebrauch durchaus akzeptabel ist. Natürlich verstehe ich, dass wir weiterhin an der Verbesserung des Kompensationsmechanismus arbeiten werden (wir haben bereits Vorschläge, was getan werden kann). Nachdem wir die Energieabhängigkeit mit professionellen Geräten verglichen haben, sind wir zu dem Schluss gekommen, dass die Unterschiede in den Messwerten aufgrund der Energie nicht die üblichen Werte in den Datenblättern überschreiten.

[DL8BCN]

Ich habe eben eine Vergleichsmessung mit 4 Strahlungsmessern gemacht mit einer Packung Thoriumelektroden in der Plastikschatulle mit 10mm Plexiglas als Betaschirm darüber.
Meßzeit war über nur ca. 10 Minuten mit folgenden Ergebnissen:
RC101:                  0,23 μSv/h
RC102:                  0,29 μSv/h
RD1008:                0,40 μSv/h
Graetz X5C Plus:   0,30 μSv/h.
Insgesamt scheint mein RC102 etwas höhere Messwerte anzuzeigen, als der RC101.
Wobei er sich bei längeren Messzeiten mehr dem RC101 annähert.
Insgesamt scheint mir der RC102 sehr gute Zahlen zu liefern.
Diesen Test müsste man noch mal mit einer etwas stärkeren Quelle wiederholen.Wobei solche Messungen immer mit Vorsicht zu genießen sind, wegen der Meßgeometrie.

[DG0MG]

Wie machst Du denn mit dem RadiaCode eine Messung über 10 Minuten? Bzw. wo liest Du den Messwert ab? Im Geräte-Display? Welche zugehörige Standardabweichung wird denn angezeigt?

[DL8BCN]

Ja, Anzeige im Gerätedisplay.
Abweichung war ca. 15%.
Und nur 10 Minuten, weil die meisten User, wie auch ich zu ungeduldig sind länger zu warten.
Ich habe nur darauf geachtet, daß sich die Anzeige stabilisiert hat.

[opengeiger.de]

Please feel free to ask me if you have any questions about the device and its specifications. There's no need to make assumptions or reason logically when I can provide you with all the specifications.

It's very simple!

@Radiacode-101 I will follow your proposal! Thank you so much for your comments on the concerns that popped up here in this forum regarding the inaccuracy we see with our RC-101/RC-102 devices when measuring low environmental dose rates since the introduction of energy compensation. What you state is that energy compensation works up to 3MeV. You explained that you are determining the statistical uncertainty in each channel separately based on a 2-sigma interval. This Is understood now, thanks for clarification. You also stated that after you compared energy dependence with professional instruments you concluded that the difference in readings do not exceed the values typically specified in the specifications. What exactly means "values typically specified in the specifications "? This explanation is not really convincing given the physics of a scintillation counter with a small scintillation crystal. It can hardly explain the behavior of the RC-101/RC-102 during our practical experiments. I will prepare a more detailed description of my concerns and I will publish it here soon.

But let me ask you a first question: Given three radiation sources that show the same dose rate in presence of an accurately calibrated professional equipment (traceable to national standards), a first source containing a radionuclide such as Lu176 that shows two major peaks at E1a=202keV and E1b=307keV, a second source containing Cs137 with the major peak at E2=662keV and a third source containing a radionuclide such as K40 with a major peak at E3=1460keV. If you do not have data from Lu176 and/or K40 at hand, you can alternatively assume other radionuclides, however the following should be valid: E1<<E2<<E3. Due to the small size of the CsI crystal in the RC-101 and the RC-102 the count rates below the peaks of the 3 sources should be very different. For this case, can you please specify the following for a measurement with the RC-101/RC-102: When T1, T2 and T3 are the measurement times required to reach an uncertainty(accuracy) of 20%, by what factor a should T2 (Cs137) be larger than T1 (Lu176) and by what factor b should T3 (K40) be larger than T2 (Cs137) when a user wants to achieve measurements with an accuracy of 20%? Can you please specify at least coarse numbers for a and b?

In our community, some colleagues did a comparative measurement with respect to the above-mentioned issues on official reference areas with well-known contamination resulting in elevated dose rates as given by our national authority for radiation control named BfS. Whereas our colleagues measured with RC-101 / RC-102 and other consumer instruments, BfS measured with professional instruments. The goal was to clarify if there are substantial differences in the measurement results for the different instruments or if the instruments match with negligible differences. According to one member of our community, you as the manufacturer of RC-101/ RC-102 had been somehow involved in this activity. We therefore would have expected that you either proudly present matching results or you tell us that you saw significant differences for some reason and you therefore will do changes to fix the issues. Interestingly, the activity took place but the results were kept secret so far. We as RC-101 / RC-102 users who didn't take part in this activity are now rather puzzled. What is the reason that you keep this secret? This definitely causes the suspicion that there is something that needs to be disguised. Could you please explain from your perspective?   

[anon_226]

Good day to you all! I am one of the leading engineers in the development of our product and therefore decided to personally register on the forum. Part of the responses were formed by my texts.

First of all I want to answer the question what are "the values typically specified in the specifications" - this is the energy dependence of detectors from such companies as SNIIP-AUNIS, Polimaster, ecotest and several more. For these devices the energy dependence is on average +-30%, and the permissible relative error is 15%. Both of these errors are defined in the test procedures.

The energy uncertainty is determined by measuring the corresponding energy of the isotopes, and the basic relative uncertainty is defined as the measurement uncertainty in the measurement range. Based on my experiments with the reference sources previously listed, the energy error was -40+20% for the different isotopes, which is well within the range of professional instruments.

The basic relative error is determined by the range of measurements, but because we do not have so many powerful sources, this experiment has not yet been conducted. Taking into account the repeatability of the characteristics of our detectors and dead time correction I can confidently say that the deviations in the measurement range due to the maximum count rate should be quite insignificant and fall within the standard 15%.

About energies and different number of pulses - yes indeed for different energies the sensitivity will be different, and therefore different statistical error. Therefore, the greatest contribution will be made by the maximum energy of radiation, but the number of low-energy quanta is much greater than the high-energy quanta, and under normal conditions, the contribution to the dose will be made mostly by low-energy gamma quanta. Also the dependence of the dose coefficients on the energy falls very quickly and at a certain point the difference in energies becomes not great, which does not require to measure the energy of the quantum and its weight in the corresponding channel. I cannot fully disclose the data on how we calculate the statistical error, because this information is for internal access, but I have described to you the basic principle of calculation. Also I can skip, but I want to say that for long-term measurements it is necessary to use the spectrum mode - in the android app the time of spectrum measurement is not limited and in this way you can achieve the minimum required statistical error.

About experiments - we have no secrets from our users and we will be only glad if correct comparisons will be made on known sources. Personally I carried out researches on several standards, and also comparison with devices of the above mentioned series on small sources.

Regards Evgeny Solovyev  

[opengeiger.de]

Hello Evgeny,
Welcome to this forum! It's not that usual that developers of instrument companies join such a forum! Therefore, we highly appreciate your contribution and we hope that our discussion will make your product better and helps other users as well. Thank you for your detailed explanations. It gives us more insight in what the instrument actually does and thus increases our confidence in it. There are still some open questions in particular around the energy compensation of the RC-101 / and RC-102. It still appears somewhat as silent "black magic". To get more confident with this, we wish to get some more information. I think confidence in measured does rate values is of utmost importance to most of the consumers particularly consumers who use your device and potential customers who try to get informed reading comments in this forum. Consumers typically are not dealing with strong sources but are more concerned about radiation issues in the environment. In this respect the black magic inside the device should be minimized. Perhaps a more abstract explanation may be the best compromise. Of course, we understand that staying competitive also is also of high importance for a small company as yours with a narrow portfolio of products. Therefore, you need to find the right balance in what you reveal to the public and what you keep internal. We'll respect this.  At the end however, a good understanding and the confidence in your product will strongly influence the acceptance of your products in the market and will determine your success, don't underestimate it!

Regarding the issues we have with the silently working energy compensation algorithm in your device I'm currently compiling a document that describes my findings and concerns in more detail. I will publish it here in the next few days as a document instead of typing long posts. You may discuss this then in more detail in your company and come back with some feedback if possible. But one suggestion in advance: Would it be possible to show the compensated spectrum optionally in the spectrum view? This would give the user the option to better judge the correct functionality. Another option would be to allow the turn on and off of the energy compensation feature with a button in the device settings when a user doubts about the correct results. What is your opinion about this idea? 

[opengeiger.de]

Nach dem obigen Post des Chefentwicklers glaube ich, es war wohl eine falsche Fährte die Raddet da gelegt hat indem er andeutete die Geheimniskrämerei um die Wismut-Referenzmessungen hätten was mit den Entwicklern von Radiacode zu tun. Wenn man dem Post oben Glauben schenken darf (in dubio pro reo), dann wissen die von nichts und es scheint eher ein Forum internes Problem zu sein. Mir ist allerdings immer noch völlig unklar warum. Ich hoffe nicht, dass das noch lange ungeklärt bleibt, denn sonst glaubt am Ende keiner mehr die Ergebnisse, wenn die so lange "reifen" müssen.       

[DL8BCN]

@Bernd: Um an noch mehr Infos zu kommen, könntest du dich im RadiaCode Support bei Telegram anmelden. Da schreiben auch die Entwickler. Ich bin da auch lesenderweise dabei.Ich war damals auch Beta-Tester als ich ein Problem mit meinem RC-101 hatte, habe ich einen speziellen BugFix von dort bekommen.

[anon_226]

The main problem in correctly estimating the natural background radiation (NRF) is the lack of known reference values of the NRF. When we modeled the conditions of interaction of gamma radiation with a crystal, we had only a small number of available reference sources, which we used to check the values obtained after processing the spectra. Well, at present in our company the number of sources is also limited, so we have to trust only those sources that we have. To make an accurate estimate of the ERF after modeling with OSGI sources was not as easy as we thought. At the moment, we will work on correcting the algorithm, but for now I can only present the energy dependence for those isotopes that our company has. If you're interested, of course)

[opengeiger.de]

At the moment, we will work on correcting the algorithm, but for now I can only present the energy dependence for those isotopes that our company has. If you're interested, of course)

Of course we are very interested in the energy dependence for those isotopes!   Since the use of sources is tightly regulated in Germany non-professionals have lots of difficulties to determine the ERF.

I meanwhile composed the promised document that explains the concerns I have and possibly others as well with the current implementation of the energy compensation. In this documentation you can see the modeled ERF I used for the RC-101 to at least explain the functionality and to highlight the issues I'm concerned about. It would be great to get some feedback on it! We all highly appreciate your work and the intention of the attached document with its comments are meant to help improving your product and to set expectations of customers right.

[hg1978]

@opengeiger.de Ich bin beindruckt. Danke für Deine Arbeit. Ich hoffe, die Entwickler nutzen diese Infos. Ich bin als Neuling im Strahlenthema etwas besorgt, ob ich mir mit dem Radiacode das richtige Gerät gekauft habe - im Sinne korrekter Werte...   

[Lennart]

Ich bin als Neuling im Strahlenthema etwas besorgt, ob ich mir mit dem Radiacode das richtige Gerät gekauft habe - im Sinne korrekter Werte... 

Es geht hier um leichte Abweichung bei der Hintergrundstrahlung. Niemand kann seriös irgendwelche Gefahren ableiten, nur weil die Anzeige um wenige Nanosievert danebenliegt. Im Test auf den Wismut-Flächen hat das AD6 von Automess (wie erwartet) eine gute Figur gemacht. Für schlappe 1.339,94 € mehr kannst Du Dir dieses (weniger empfindliche) Gerät ohne Spektroskopiefunktion kaufen, Preis siehe: https://shop.ziegler.de/product/A0088111/dosisleistungsmessger%C3%A4t-6150-ad6-e

Mit der entsprechenden Szintillationssonde bewegt man sich sicherlich im preislichen Bereich eines PKW. Nur um mal die Relationen aufzuzeigen.

[hg1978]

Ich bin als Neuling im Strahlenthema etwas besorgt, ob ich mir mit dem Radiacode das richtige Gerät gekauft habe - im Sinne korrekter Werte... 

Es geht hier um leichte Abweichung bei der Hintergrundstrahlung. Niemand kann seriös irgendwelche Gefahren ableiten, nur weil die Anzeige um wenige Nanosievert danebenliegt. Im Test auf den Wismut-Flächen hat das AD6 von Automess (wie erwartet) eine gute Figur gemacht. Für schlappe 1.339,94 € mehr kannst Du Dir dieses (weniger empfindliche) Gerät ohne Spektroskopiefunktion kaufen, Preis siehe: https://shop.ziegler.de/product/A0088111/dosisleistungsmessger%C3%A4t-6150-ad6-e

Mit der entsprechenden Szintillationssonde bewegt man sich sicherlich im preislichen Bereich eines PKW. Nur um mal die Relationen aufzuzeigen.

Verstanden