Page 1 of 1

CMOS Neutron Detection

Posted: Sun Jan 05, 2020 3:53 pm
by Jim Kovalchick
Has anyone attempted to use a camera CMOS to correlate neutron numbers?

Based on different instances of different neutron performance over the last few days for similar voltage and current, and clear differences in the amount of speckles I saw in images taken by my cell phone camera, I think the speckles are mostly neutrons and not xrays.

I think that snap shots may be used in the same way bubble detectors are but much more cheaply.

Key to this would be doing some research to confirm neutron correlation with repeatability, eliminating xrays, and determining a calibration relationship for a given set up.

I may save this for when I have a consistently running fusor.


Jim K

Re: CMOS Neutron Detection

Posted: Sun Jan 05, 2020 4:17 pm
by Richard Hull
Jim, I would think that x-rays would be the more likely candidate for the flashes, "speckling", as they are the predominant radiation from the fusor as it ramps up with more voltage and current, just as the neutrons.
Admittedly the neutrons are at 2.45 mev compared to the 20-50kev x-rays. It is the flux of x-radiation that far exceeds the neutron flux. My fixed camera has been in place since 2004. It now has a bout 20-30 dead pixels or permanently turn on , (white). I can't speak of how many may be black (blown out), but I think they are there and might, on close inspection, be visible and almost as visible as the white dots.

Your idea would have merit if we could be sure it is neutrons and not the torrent of x-rays blasting into the camera that causes speckling.

Richard Hull

Re: CMOS Neutron Detection

Posted: Sun Jan 05, 2020 5:57 pm
by Jim Kovalchick
A quick read of a couple academic paper abstracts backs up my idea. Video CCD's behave differently than CMOS by the way. Proof of course will be in practice.

Re: CMOS Neutron Detection

Posted: Sun Jan 05, 2020 6:32 pm
by John Futter
Jim some time back I destroyed a new colour CCD camera at work taking pics Cherenkov radiation of a 12MeV 22nA proton beam in air. viewtopic.php?f=19&t=7995
If you play the vid frame by frame you can see the CCD damage some of which is temporary while some is permanent.
The neutron detector in the vid is on its least sensitive fullscale range of 1.4 millirads per hr x 1000 ie 1.4 rads per hour detector about 2 meters away from proton beam.
the video is edited so that the lights are off in the lab to see the Cherenkov radiation from the beam and with the lab lights on to see the neutron count --which was well overrange so I have no idea of how high it actually was.
The accelerator pipe end with the thin Be window which is made from aluminium has to be locked up in the nuclear store for a couple of weeks as the aluminium gets activated by the external beam
Needless to say it was worth sacrificing a new camera for the pics

Re: CMOS Neutron Detection

Posted: Sun Jan 05, 2020 11:32 pm
by Jim Kovalchick
There is no doubt that CCDs and CMOS show interaction events for gamma and higher energy xrays. Some of the interactions cause pixel sparkle and sometimes permanent damage. In my years of work at a nuke plant I've seen plenty of toasted cameras, mostly ccd, toasted by gammas.
My readings are far from complete, but it seems that neutrons are also 'seen' by camera detectors and a couple academic papers are out there discussing CMOS detectors being used for neutron detection and dosimetry.
I became intrigued the other day by this possibility because I took a series of camera images with my cell phone during identical voltage and current conditions but very different neutron production rates. The images with the most sparkles correlated to more detected neutrons by my BF3 tube. In both cases xrays were raging out the view port. I did not quantify the xrays between the cases, but further studies would do this. Application of lead shielding would also be a good to eliminate any question of xrays.
I have attached two images. Each image is a crop from two full images taken of separate runs of 37 to 38 kV at approximately 10 mA. I cropped the same position reference to the grid in each. I then maximized contrast and removed color. There is a clear difference in image spots associated with radiation. I would expect any spots from xrays to be affected by xray energy and current but not fusion reaction rate.

I may be off my rocker, but I want to run this one to completion if I can. Hopefully the forum will hear more from me by spring.

Jim K