Detecting cosmic rays using CMOS sensors in consumer

Detecting cosmic rays using CMOS sensors in consumer devices Matthew Plewa

Introduction • What is a cosmic ray • Stable • Typically striped nucli • Charged • What is a secondary? • Gammas, muons, protons, pions and electrons

Common Detections Methods • Citizen Scientist • Cloud Chambers • Vapor condenses along the particle path (can visually see) • Scintillator Panels • When excited by ionizing particles they emit photons

Common Detection Methods • CMOS Sensors • When light is blocked from the sensor cosmic rays deposit a charge on the sensor. This charge is then interpreted as light in the image. • The sensor must be capturing a frame at the time the cosmic ray enters it. • Small active area

DECO (CMOS Sensor) • Funded by American Physical Society Knight Foundation and Simon-Strauss Foundation • Works on android phones (IPhones soonish) • Great learning tool and is becoming a great scientific tool with upcoming upgrades

CMOS Sensor Benefits • Easily attainable • Most of us have them in our pocket • Easily used • CMOS sensor just needs to have light blocked • Tools are being developed for detection in the frames

CMOS Sensors • Due to the small area (~. 15 cm^2) livetime is extremely important. • For this reason we had to look at what method would increase livetime • Video mode • Still capture

Video vs Still Capture • Still capture (8 MP) • Livetime ~5% • 1 out of every 20 events detected • High resolution (Low data rate ~1 a sec) • Video mode (2 MP) • ~95% livetime • Low resolution (High Data rate ~30 a sec)

Video vs Still • What would work best for high altitude ballooning? • The more events that can be captured the more analysis that can be done. • Lower resolution means track analysis is less accurate • More events seemed to be the best choice for initial testing

Using a Go. Pro • Readily available device • Knew the sensor size (. 24 cm^2) so determining how many events that should be seen at ground level was easy. • Fairly reliable hardware

Post Flight Processing • Video needed to be decompiled into individual frames • Current method of detecting events is a simple threshold cut of the R, G and B values for each pixel • Altitude data was not recorded so events couldn’t be correlated accordingly

Results • 66 “events” were detected in 2 hour span • Electron worms and Alpha particles (decay on the surface of the sensor) • Demonstrated the viability of detection using CMOS sensors • Need better filtering parameters

Future work • Raspberry Pi & Pi cameras • Potentially larger surface area. • Higher resolution • Data logging • Temperature • Altitude • GPS

Long Duration Exposure • DECO V 2 (full resolution) • Android 5. 0 + Camera 2 api • Able to set ISO • Full exposure control up to 2 seconds • Automatic datalogging • Database collection • Highly recommended for future ballooning

Track Analysis • Need to determine how to reduce electron and alpha particle noise • Track traceback and pointing of particle origin (path before entering sensor) • Will be easier with higher resolution

Questions?