DEV Community

Benjamin Klein
Benjamin Klein

Posted on

Cosmic Interference

The Earth is constantly bombarded with radiation. Some of this radiation comes from the Earth itself and some from the Sun. There are a variety of different types of radiation that react with various materials. Some forms of radiation interact with hardly anything at all. A particularly interesting form of radiation comes from throughout the universe and is referred to as cosmic rays. This term, however, is a misnomer. Radiation coming from the cosmos is made of high-energy particles, rather than rays. The majority of these particles (roughly 90%) are singular protons, and the remainder are various types of nuclei. Cosmic rays are formed in the hearts of dying stars as they go supernova or radiate from black holes. These rays can affect our computers, and in fact, you have experienced this interference before. Before we can understand what causes this phenomenon, we first have to understand what bits are and how semiconductors create them.

Computers use a special material called semiconductors to conduct computations. A semiconductor is a material that's conductivity is dependent on its current state. Conductivity in a material is dependent on valence electrons, so let's brush up on our chemistry.

Protons, neutrons, and electrons comprise all atoms. Protons and neutrons clump together in the center and electrons orbit them in groups called electron shells. The outermost shell is called the valence shell, and the electrons in the valence shell are called valence electrons. All atoms are 'happy' when they have eight valence electrons. Atoms can share their valence electrons, so they can both have eight. This is called a covalent bond. For a material to be conductive, there needs to either be extra electrons or extra protons. Extra electrons can flow freely through a material. Extra protons provide a place for electrons to travel to. This is all the chemistry you need to understand semiconductors.

The base element semiconductors use is silicon. Silicon has four valence electrons so it forms a covalent bond with four other silicon atoms. Since there are no free electrons and nowhere for the electrons to go, silicon is an insulator. To make the silicon conductive, it has to be 'doped.' Silicon is most often doped with either boron (which has three valence electrons) or phosphorous (which has five valence electrons).

Image description

Silicon doped with boron is called a P-type conductor because a positive charge can flow through the wire. Since boron only has three valence electrons, when it forms a covalent bond with four other silicon atoms, the boron atom only has seven valence electrons and it wants eight. When a new electron is introduced, it can hop into a boron's valence shell. Electrons can now flow freely through the material.

Image description

Silicon doped with phosphorus is referred to as an N-type conductor as a negative charge can flow. Because phosphorus has five valence electrons, when it forms a covalent bond with four silicon atoms, there are nine valence electrons in its valence shell, which is one too many. If a positive charge is introduced to this material, the valence shell will shed an electron and the electrons can once again flow through the material.

Image description

What allows a semiconductor to change its state is a PN junction. A PN junction is where a P-type conductor meets an N-type conductor. The extra electrons from the N-type conductor can jump to the empty spaces in the P-type conductor and a depletion region forms. Since there are no longer extra electrons nor places for the electrons to go in the depletion region, the material loses conductivity.

If we take a battery and apply the positive terminal to the N-type side and the negative terminal to the P-type side of the semiconductor, the depletion region will expand making the material progressively less conductive. This is because the extra electrons from the N-type side of the semiconductor can leave the material while the empty spaces in the P-type side of the semiconductor are filled with electrons from the battery.

To make our semiconductor conductive again, we can reverse the flow of current by applying the positive terminal to the P-type side, and the negative terminal to the N-type side. If more than 0.6 volts are applied, the electrons will be pulled out of the P-type side of the semiconductor, leaving empty spaces behind. On the N-type side, electrons will be forced into the material so there are once again extra electrons that can flow freely through the material. The depletion region collapses and our semiconductor is now conductive again.

To test our semiconductor's current state, we can apply no more than 0.6 volts with the positive to the P-type and the negative to the N-type. As mentioned before, if more than 0.6 volts are applied, the depletion region will collapse if the semiconductor is in a nonconductive state, but less than 0.6 volts will not be able to pass through. If we apply less than 0.6 volts and the voltage can pass through, then we know our semiconductor is in its conductive state.

This is how computers can generate bits. If the semiconductor is in its conductive state, its circuit is open which is represented by a 1. If it is in a nonconductive state, the semiconductor's circuit is closed and is represented by a 0. These bits are how computers store data. By setting some semiconductors to open and some to closed, current can be passed through to re-evaluate that data.

Unfortunately, there is another way a semiconductors state can be changed. When a cosmic ray strikes a semiconductor, it can leave electrons behind or knock electrons clean off, opening or closing the circuit. This occurrence is called a single event upset or a single event error. Sometimes single event upsets will just crash your computer, but sometimes they can cause substantial problems.

On October 7, 2008, Qantas Flight 72 set off on its way from Singapore to Australia with 315 people aboard. While cruising over the Indian Ocean, the aircraft suddenly entered a steep dive. The Airbus A-330 descended 200 meters in only 20 seconds. The aircraft experienced 0.8 negative G's and anyone who wasn't wearing their seatbelt at the time slammed head-first into the ceiling. 118 people were injured, but luckily all survived.

A subsequent investigation showed no errors in the flight computers and no issue with any control surfaces. Everything was functioning normally when the event occurred. The culprit was most likely a cosmic ray. After further analysis, investigators found that the event could be replicated if a singular bit was flipped in the flight computer, that was holding onto the aircraft's current pitch orientation. Though it is impossible to prove it's highly likely that a single event upset caused the flight computer to believe that the aircraft was pitched dangerously high, so it violently pitched down to compensate. There are however cases where we can measure and register single event upsets.

Nasa, knowing of the dangers of radiation, designed the space shuttle with five flight computers. If any of the computers encounter an error, the other computers will overrule it. When this occurs, if the error is not traceable, it is registered as a single event upset. The space shuttle once registered 161 bit flips on an only five-day mission.

Luckily this doesn't occur as often on the planet's surface, as it's better protected from radiation, but it still does. Every day, particles that have been traveling for millions of years over millions of miles strike our computers, corrupting data and causing errors. It is a constant reminder of how small we truly are and how hostile the universe actually is.

Work Cited
https://en.wikipedia.org/wiki/Qantas_Flight_72
https://www.youtube.com/watch?v=33vbFFFn04k
https://www.youtube.com/watch?v=AaZ_RSt0KP8

Top comments (0)