I'm on paper designing a nanosatellite power system. The satellite would be orbiting in LEO. And I am wondering how and what type of radiation would affect its electronics. The system has a microcontroller, a few inductors, a few transistors, capacitors and some op-amps. How does one actually shield a nanosatellite from radiation? Does the metal box of the power system help? Or should the nanosatellite be made of some particular material to protect the subsystems from radiation?
The electronics of a satellite can be changed by radiation because radiation can be just photon and photons can excite electrons or holes creating unwanted error.
Radiation can be electrons or positrons and if it is positrons (β+) they can annihilate with electrons or if it is electrons they can fill holes which is not good in semiconductors.
There are different type of shieldings for different type of radiation.
Just Google how to protect from different kind of radiations.
And those mentioned are just the simplest case there could be α radiation but the changes are more for the CSE.
Radiation can flip bits in registers and memory. First minimise that, with shielding, and using 'rad-hard' components. Second, minimise the effect of that, with system design.
If the flipped bit in in the program counter, then you can expect the program to run off into the weeds. Look up 'crash bars', instructions to reset peppered amongst the program proper, so that a bad PC will hopefully eventually reset the device.
Look up 'watchdog', hardware to reset the microcontroller should it fail to do something expected, like flip a bit every 100 mS or whatever you choose.
If the flipped bit is in the program flash, then it's game over, unless you're routinely checksumming flash and have a way of correcting it with FEC (forward error correction) and/or multiple copies of the program.
Look up 'redundancy', using multiple processor running the same program, and suitable hardware voting on the outputs, and resetting or otherwise repairing the 'odd processor out'.
Silicon breakdown voltage shrinks with lithography and contamination.
Radiation shield effectiveness depends on wavelength and atomic structure of metal for Xray to alpha,beta to gamma radiation.
But consider the E field and junction size and normal breakdown to conducted voltages can be small, but large for nano wave particles as the impedance rises in the dielectric.
This contamination is R in a dielectric C reduces the gap length and causes partial discharge,PD and might sound like a Geiger counter. When mV/um is energized by photons of alpha and gamma waves it goes through metal and is harder to shield thus larger litho older technology was generally better than newer small litho. But as contaminant levels improved by processing and process, newer smaller litho became more tolerant.
Think of it like a unijunction partial discharge , PD relaxation oscillator effect but more random excitation from cosmo solar flares. So E field is important and wavelength for radiation hardened IC’s .
They do this by Xray bombardment precap to “oblate” the contaminant particles that might be in the PPM concentration levels making it more robust, or it could make it worse if the particles fuse together and remain behind thus extremely low current HIPOT to devices might seem practical but expensive to substrate might be done now (?) which makes more expensive $$ yet verified partly. But it must be done by a reliable process of visual inspection and XRAY bombardment.
without this then you have a Monte Carlo machine. do you feel lucky? Then add redundancy with detect and correction. The takes EMI design to a much higher frequency level of physics.