Radiacode 102

[Dan Knapp]

I just came across this compact scintillation detector/spectrum analyzer that sells for only 259 Euro (made in Cyprus). It has a smartphone app that displays the radiation spectrum and will identify radioactive isotopes. This seems like quite a deal for the size and price, and I was wondering if anyone on the forum had any experience with it. It lists a resolution spec of "9-10% FWHM for Cs-137." It wasn't clear to me what they mean by stating the resolution as a percent of FWHM. (www.radiacode.com)

[Richard Hull]

Seems amazing and uses your smart phone as a readout/display. Thanks for the heads up on this.

Richard Hull

[William Turner]

FWHM represented as a percentage is done with scintillation detectors because of their relatively low resolution. 10% for Cs-137 means 66.2keV FWHM (HPGe could be 1keV FWHM for a Fano Factor of 0.1). A high resolution capable detector in this device would be wasted, since many of these pocket / sound card gamma spectrometers are almost toys. To achieve good resolution and accurate results a significant amount of signal processing is required. There is a lot going on inside a NIM spectroscopy amplifier (Ortec 572/672, Canberra 2025/2026).

[Richard Hull]

In addition to significant A-D high speed conversion and other electronics, the scintillation detection system is always the limiting factor regardless of electronics.

This item above might serve to identify warm to hot gamma sources with ease but might not separate close peaks for study purposes.
I wonder how it would do on Ra with little red riding hood and the 3 bears.

Richard Hull

[Dan Knapp]

Attached is a radium spectrum. It looks pretty good to my untrained eye.

[Frank Sanns]

Dan,

The spectrum looks too good. I know much progress has been made with miniature detectors but this seems almost too good.

Replacing the PMT with a solid state one wins quickly. 10 mm detectors are all of a sudden possible.

Do you happen to know the innards? Nothing giving a clue on the website.

What is curious is the resolution at both high and low energies. Maybe it is a Ce doped LaBr3 crystal but those are not cheap.

Then there is the diameter of a scintilation crystal needed. There are edge effects where a gamma is not fully stopped by a crystal that is too small. This manifests as a lower energy random pulse that adds to the noise. This of course is a ratio of stopping volume vs gamma rays to be studied. It is not such a factor for low energy but becomes more pronounced as the energies goes up.

Detectors like CZT are great at the low energy spectrum but fail miserable at the higher.

BGO is great for stopping high energy gammas but has lousy resolution.

No amount of filtering or even a 3 hour count time like in your photo can accurately clean up a a low resolution detector. Which brings us to the question of what the detector might be or what the software is actually doing.

If there is a built in reference spectrum in the software, it may be filling in the spectrum from some fairly poor actual measurements. In the case of radium or any other nuclide, there are only a very limited number of isotopes that would give any more than noise of a particular isotope characteristic energy value. (Wow, there is a noun string!). It is conceivable that that is what is going on with the detector. If it sees a peak or two unique to an isotope, then it must be so and it fills in its high resolution spectrum.

I do not know this for sure but am totally speculation base on other systems I have seen in other technologies.

It would be really interesting to see how the spectrum fills in vs time. What does the spectrum look like in 5 minutes? 15? What does it do if you have a second gamma source their at the same time. Like Cs and Ra?

[Richard Hull]

Yes that is a hot source 31,000 cpm. Yes, that is a 3 hour count. the low count at 600kev tells me the thing is not great at the high end like a 2x2 NaI would be. As the 600kev is always higher than the well represented goldie/3 bears.

Frank is right about the collection time. I have a 3x3 NaI:Tl in a Canberra system and with the same source I get this spectrum in 60 seconds of collection time on medium point counts and in 2 minutes on fine data point resolution.

Still, I am amazed this little toy does this well even over a tedious 3 hour collection time. I have no smart phone so I don't have a suitable display should I wish to get the radicode 102.

Richard Hull

[Alexey Khrushchev]

I have a Radio Code -101. This is a pretty good dosimeter-radiometer with a gamma spectrometry function. It easily detects a small content of Cs-137 in the soil that was thrown out as a result of an accident in the city of Elektrostal (Russia). With the help of this device, I even recorded photons resulting from the annihilation of positrons at an altitude of 10 km, during an airplane flight.

[Dan Knapp]

The detector in this device is a CsI (Tl) crystal 10x10x10 mm
+ solid-state photomultiplier.

[Richard Hull]

The 10mm xtal speaks to the lack of a high energy fall off. Still with long collection times and good code to compensate in the count regime one can make the spectrum look rather normal, I might imagine. If they have such a long term algorithm to adjust the spectrum, that is sweet. I assume you need not keep the smart phone on during the collection cycle, but can check in over the long term collection period.

If all of my suppositions above are true, then this is quite an interesting piece of gear. Is it standalone as a data collector minus the phone?

Richard Hull

[Dan Knapp]

Yes, you can use it without the phone.

I invited a person from the company making the device to join this discussion, but he encountered a problem with the registration validation question being less than clear. It is not clear to me also. Is the number being requested the number of neutrons plus two or is it three divided by the number of neutrons plus two?

I made the plunge and ordered one of the devices.

[Richard Hull]

I hope you have a few varied sources to give it a good test. Let us know what you think regarding collection time on weak versus stronger sources.

Richard Hull

[Frank Sanns]

The question gives and answer that is not in the public domain so AI cannot find the answer online. I will leave the answer up for a couple hours then delete it. xxxxxxxxx3

[Dan Knapp]

That’s what I thought the answer should be but the “the3\” was confusing.

[Dan Knapp]

Richard, I'd welcome your suggestions for tests to put the Radiacode 102 through its paces. I have radium dial instruments, Americium smoke detector sources, a Cs test source, uranium, and potassium chloride.

[Richard Hull]

Radium is number one. It has a number of peaks. Run a series of accumulations like 1,5,10 minutes and see what is there.
Americium should be easy and quick as it is effectively a hot 59kev gamma emitter
I would be amazed if it showed the KCl signature at all over any sort of short time span.
I hope there is a mark on the device where the crystal/scintillator is located so you can jam the source next to it to warrant maximum count and detection speed.

The key aspect is how does it respond over reasonable time frames related to more "emergency" determinations and identities.
For simple tests where it can be accumulating for hours is nice, but short term tests with usable results would be nice.

Richard Hull

[Emma Black]

I looked up this little device a while back after seeing it used in this fairly interesting video about the EBR-1 site. I think it was a sponsored video by Radiacode so take the demo with a pinch of salt.

Be interested to see the results of the testing. Planning on some more "nuclear tourism" soon and having something pocket sized like this could be really handy.

https://www.youtube.com/watch?v=K0kvQJ-j2nk

[Ryan Ginter]

I happen to own a Radiacode-102, so I'll see if I can answer some of these questions.

Regarding the spectrum image posted previously, long exposures on high activity sources do produce a rather clean spectrum. Low activity sources require many hours however. A spectrum I made of trinitite required multiple days to produce clean peaks, though its gamma activity was not much higher than the background. The software does allow for background readings to be stored in a library and subtracted from subsequent readings to improve the clarity.

Speaking of software, it isn't necessary to have a smartphone to control the device. There is also official software to allow the spectrum data to be viewed on a computer by connecting the device to USB. The computer doesn't need to be connected at the time of measurement, the Radiacode could be used out in the field and then the file could be uploaded when you got home.

Looking at the data collected, the device keeps a log of every pulse detected and notes the time of the event and the energy/channel that it was detected in. To the best of my knowledge, the device does not alter the spectrum data by filling in with an external library. There are a few algorithms present, but as I understand it only one is used on the device itself. From what I remember, the device originally had a lower number of channels. An update was pushed that added an algorithm to help increase the resolution of the lower energy pulses.

Regarding the spectrum viewing software, there are two algorithms that I know of. One is a logarithmic curve that can be superimposed over the spectrum to allow easier visualization of the high energy pulse counts. The other is a filter that attempts to smooth and average the peaks out. Both of these algorithms are optional, and are controlled by moving a slider at the bottom of the screen. When both are set to zero, the spectrum appears to be raw data.

The device itself is very compact, and the location of the scintillation crystal is marked with a radiation symbol on the plastic housing. The screen is a simple LCD panel with glass. Count rates, dose rates, total accumulated dose, and spectra can all be viewed on the screen. I would not recommend anyone use the LCD panel for viewing spectra however, as the resolution of the screen is much too low for it to be useful. It does give a general sense of how much the data has filled in however.

Next we will move on to viewing the spectra taken on a short time scale. I have to leave for work soon, so for the moment I will only be uploading the Radium Bromide spectrum. The spectrum was collected from a clock face roughly 3" in diameter. The activity of the clock face was measured by taking a 10-minute count with a 2" Geiger probe and subtracting out the background. Measurements were taken with the protective mesh of the mica window in contact with the glass of the clock. Activity was measured at 3,780 CPM.

Likewise, the spectrum was taken with the Radiacode-102 in direct contact with the clock face glass, with the scintillation crystal in the center. All spectra were taken without using the smoothing or amplifying filters.

1 Minute Radium Bromide Spectrum

5 Minute Radium Bromide Spectrum

10 Minute Radium Bromide Spectrum

15 Minute Radium Bromide Spectrum

30 Minute Radium Bromide Spectrum

The software also has an isotope library overlay. Clicking on a spectrum channel will display a line marking the closest isotope photo peak from a list of common isotopes, as well as additional lines for its decay daughters. Here is an image of the overlay with the 30-Minute Radium Bromide spectrum taken in logarithmic mode.

I have other spectra I can upload if anyone is interested. From what I've seen overall, anything over 20-30 counts per second(as measured by the Radiacode) can be identified within a few minutes. Lower activity samples will take multiple hours to properly identify.

[Richard Hull]

Thanks very much for the report on the 102 model. It is as I thought. Not a great substitute for a large 2X2 or 3X3 NaI:Tl crystal and PMT hooked to a more formal gamma spec instrument.

However, its "wander about" tracking of radiation is of great value and far exceeds its value as a gamma spec.

Richard Hull

[Frank Sanns]

Longer count times will give better spectra up to a point. The original spectrum still looks too good to me.

It does not seem that the 30 minute spectra that you posted has any chance of looking like the 2 hour 50 minute original photo that you put up. It looks too pretty. Still looks like a reference spectrum than what was collected. I am still skeptical that a little crystal could ever produce that first photo that started this thread. I have a 2 inch LaBr3 that has to work fairly hard to get something that pretty in an hour or two.

[Ryan Ginter]

There seems to be a bit of confusion, the original spectrum was posted by Dan Knapp. I believe that original photo comes directly from Radiacode's website. One thing to keep in mind, the spectrum from the official image was taken with a much higher activity source than mine.

The spectrum pictured below is one I made that was taken over the course of three hours. To collect a data set equivalent to Dan Knapps image, I would need to collect for almost 12 times as long.

I would also be willing to bet Radiacode had a lot less bremsstrahlung radiation in their data set. The brass of the watch hands and the clock face will likely distort the spectrum. Still, I'm not sure if Radiacode's spectrum is legitimate or not.

Pictured here is a thirteen minute spectrum I took of a high activity Cesium-137 check source. I'll collect a spectrum on it for a few hours after I get home from work. Real world testing is the only way we'll know for certain.

[Richard Hull]

Nice short run Cs137 spectrum. Source strength always wins.
The radcode is just not up to really weak source work.

A decent trinitite specimen can take a number of hours to show well on a 3X3 bircon NaI:Tl scintillator and a Canberra 40 gamma spec.

Richard Hull

[Ryan Ginter]

Just finished the three hour Cesium-137 spectrum using the same source as the one from my last post.

The additional time appears to have hardly made any difference. The data appears to be siding with Frank here, I find it difficult to believe the spectra shown on the Radiacode website were taken with a production model 102.

Perhaps I'll try taking a two day measurement of my Radium sample in order to gather an equivalent data set to the Radiacode website image and finally put this to rest.

As a bonus, here is the three hour Cesium-137 spectrum displayed on a logarithmic scale with the background subtracted out. This is about the cleanest I've seen from the Radiacode-102.

[Ryan Ginter]

The original image posted by Dan Knapp displayed a spectrum taken from a 518 cps source for a duration of 2h 50m 42s. 518 cps for 10,242 seconds comes out to 5.31x10^6 counts.

The new spectrum I recorded used a slightly more active Radium dial than the last one I used, the Radiacode-102 measured it at 63.9 cps. 5.31x10^6 counts divided by 63.9 cps gives us 83,100 seconds, or 23.1 hours. This is the duration for which data should be collected to make a fair comparison between the spectra.

Pictured here is the 23 hour spectrum I just finished. The photopeaks did clean up a little compared to the shorter duration spectra, but it still doesn't look like Radiacode's official image.

I think its safe to say those sorts of results are unrealistic.

[Frank Sanns]

This one I believe 100%. It looks right for the size and type of the crystal. Thanks for posting the longer run time.

For the record, I think the unit is fabulous for what it does. It was never my intent to dis the unit. That very first post of the spectrum just seemed too good to be true.

Those of us that have done a decent amount of spectroscopy know how difficult it can be to get nice clean peaks.

[Ryan Ginter]

I never doubted your intentions Frank. Logic is often mistaken for hostility, but skepticism is vital in the field of science.

[Richard Hull]

Ray, I really appreciate your effort on our behalf regarding the radiacode 102. The real value of any gamma spec is identification of isotopes in activation and which of the two major elements are contained in a mineral sample. Finally, it is the only way we can get any sort of accurate, relative gamma strength and emission from anything radioactive.

Sadly, any GM counter that is not calibrated against a known gamma source related to its GM tube's specifications is rather worthless for any sense related to gamma sources.

I must admit that I am rather spoiled by my Canberra 40 and its 3X3 Bicron head. I see these smartphone crushed X axis images with all the peaks jammed up together. I am use to a more huge x-axis display with the peaks spread out a bit. I assume that is the image is moved to a computer, one might expand it a good bit. I am so use to a simple line peak and the absence of all that fill under the line that I look at it as needless clutter, but that is just me, I suppose. I am used to no more than 1000 second counts to get a fabulous spectrum. Of course I rarely work with "whisper" sources. I guess it is what one gets use to.

For its price, in the hands of one familiar with its long collection times and application, the radiacode is quite amazing.

Richard Hull

[Ryan Ginter]

The below image is another screenshot from my phone.

The previous spectra were displayed using settings choosen to mimic the original image as closely as possible.

I agree, you can't beat the desktop style scintillation detectors when it comes to lab work, but it's really cool you can carry something like this around in your pocket.

[Ryan Ginter]

Just a heads up to anyone who may be somewhat interested in the Radiacode-102, it would appear they just released the 103 model. Looks like they improved the resolution slightly from 9.5% to 8.2% FWHM Cesium-137. This will likely also mean a price drop for the 102 model.

[Richard Hull]

Like the ever increasing iPhone/smartphone numbered models. So it is with Radiacode. I'll hold out for the Radiacode 110 blister packed in Target and Walmart for $39.95.

Ryan, nice single line log spectrum of radium above. Goldie locks and the three bears are easily seen.

Richard Hull