Gamma Spectrometry Forum

I just acquired an interesting device - the FLIR ICX Raider. It features multiple CZT detectors with resolution <3.5% + 3He Neutron counter.
Gamma dose rate (DR):
Number of Detectors: 8
Material: Cadmium-Zinc-Telluride (CZT)
Volume: 1cm3
Sensitivity (137Cs): ≥6 cps/µR/h
Energy Range: 30keV-3MeV
Dose Rate Range: 5 µR/h to 150 mR/h

Gamma identification detector (ID):
Number of Detectors: 1
Material: Cadmium-Zinc-Telluride (CZT)
Volume: 1cm3
Energy Range: 50keV-3MeV
Sensitivity (137Cs): ≥0.75 cps/µR/h
Energy Resolution (137Cs): <3.5 % @662 keV
Spectrum Length: 1024 channels
Dose Rate Limit for ID: 150 mR/h (137Cs)

Thermal Neutron:
Number of Detectors: 2
Type/ Material: Proportional Counter, primary
gas 3He, fill- pressure 15 atm.
Overall Dimensions: Ø0.5" by 1.62"
Energy Range: Thermal to 10 MeV
Sensitivity: 6 cps/nv ±20%
Flux Range: 1.3 ∗ 10−3 to 3.74 ∗ 10−4% nv

The device was used but in a very good condition. Apparently the price of a new one was ~25,000 EUR and it was originally manufactured by a German company (ICX?) which was later acquired by FLIR and the technology was incorporated into their newer devices.
It has a very sensitive detection/search mode, can do quick and more extensive gamma spectroscopy and ID a set of pre-defined isotopes.
When I first booted it, I noticed a very familiar user interface and sound. In the past I worked as firmware/kernel/OS engineer on almost all versions of Windows CE and this looked very familiar to me :) Indeed the device is running on Windows CE 5.0. Further inspection has shown that it's also based on the Intel/Marvell PXA270 ARM-based CPU, exactly the one I worked with the most..

The device has its own web-server (even though using outdated and no longer supported TLS 1.0). Connecting it to a PC via USB RNDIS will allow extensive configuration via its web-based interface. Many parameters can be changed, detailed reports and spectra files in various formats can be downloaded. However, it doesn't allow adjusting calibration data, which is a pity as asking the FLIR service to calibrate it would be hell expensive (a calibration+minor service repair by FLIR did cost $6000 in the past!). I guess I'll have to use my WinCE knowledge to hack into it :)

I was really pleased how well the device is equipped, it's several levels above other devices like the Mirion HDS-101 for example.

However there's one issue. The device is equipped with a Li-Pol internal 3.7V 610 mAh battery and user-replaceable 3xAA battery pack. By default the device is powered from the main AA batteries and if they are discharged it switches to the backup internal battery. The internal battery is also charged from the replaceable ones. The problem here was that the device drains the batteries very fast, much faster than the manufacturer-declared 10-24 h run time. So I suspected the problem might be the old Li-Pol battery (which should be replaced every 2-3 years). According to the date code on the battery (year=0) I assume it's from 2010.. Sending this to FLIR was no option due to the astronomical prices, so I decided to disassemble the device and look inside. Was quite nervous as I expected it to be densely packed with electronics and didn't want to break anything, so I also asked a friend skilled in electronics for help..

So here we go, the inside. The device has multiple PCBs packed closely, here the bottom 2 PCBs: Here the CZT front detector with own PCB: Unfortunately I forgot to take a photo of the 3He counter, next time.

The VARTA EZPack S has proven faulty indeed, designed internal impedance should be 115-150 mOhm, measured was >600. It's no longer manufactured, so I ordered a Li-Ion replacement and now waiting for it to arrive and assemble the device back.
Will post more information once the device is (hopefully) working again.
Any questions, let me know...

Very interesting device and post Martin, thanks for posting it! It will be also interesting to show measurement results when you are done with the repair.

Neutron counter: Internal battery has been replaced but the problem seems to be in battery monitoring circuitry. It doesn't properly estimate replaceable batteries capacity and switches to internal backup battery. So more work is required...

Device is finally repaired, it was just a small problem with contact on battery holder.
So first let's test the resolution using Cs-137: The claimed <3.5% seems to be real, calibration data in firmware claim 3%.
Apparently the CZT detector doesn't require calibration and results are quite precise. 661 keV missed by a small margin, Am241 59 keV almost precise.

Martin,

Well done, I know how satisfying it is to find the fault with equipment that someone else discarded as useless or not worth repairing. Often it is less complicated than you would at first imagine.

Not related to radiation detection, but my 8 year old son has taken up playing the saxophone in the school band, so I looked on ebay and found a good brand saxophone in need of repair (auction), ended up with a winning bid AUD $65.00 (USD $45). I changed all the pads and now it plays like a new instrument. Putting it back together was a real head-scratching puzzle, but very satisfying in the end 😉

Sesselmann wrote: ↑

30 Jun 2022, 12:33

Well done, I know how satisfying it is to find the fault with equipment that someone else discarded as useless or not worth repairing. Often it is less complicated than you would at first imagine.

Especially with a product in price range that high. Great job and deal, Martin!

Indeed, this was a big risk but fortunately paid well! I bought it for 3000 EUR (incl. shipping), so almost 1/10 of the original price.
Even that was quite expensive and I hesitated, but given the capabilities of this device it was probably the best deal I ever made (on detectors) ;-)
Apparently this device belongs into the dual-use items group, so there are also certain (export) restrictions involved...
If I only could hack into the firmware and adjust some things there.. The developers were smart and encrypted the firmware image package and configuration data in a ZIP with a strong password which I couldn't break yet. Maybe I will attempt to read the internal flash content, but this would again require complete disassembly and JTAGging it, which is not a trivial job.