Geostationary communications satellites are vulnerable to low earth orbit EMP attack. Presumably the m288 orbit would be more vulnerable. \James Bowery\ ----- An EMP (Electro-Magnetic Pulse ) burst results from the deposition of fast electrons from a nuclear burst on the upper atmosphere, before the slower positively-charged ions arrive. The asymmetrical nature of the charge distribution results in a huge electric field. However, this voltage impulse is spread out over many microseconds, so the high frequency components roll off rapidly above 100MHz or so. EMP damage can occur because of damage to the electrical power systems, and damage to the radios. The radios will be tuned, so only energy within their band is a problem. A sensitive, high bandwidth radio will presumably have smaller ESD (electrostatic discharge) protection diodes, which may blow out more easily, allowing the LNA (low noise amplifier) front end to be damaged. While EHF radio systems in the 30-60GHz range may be more sensitive to smaller energies, there will be much less of it to do damage, because of the EMP frequency rolloff. Designing LNA front ends that are rugged against EMP will be a tradeoff between performance and military survivability. In time, designs that are "RF nonlinear", with parasitic elements that subtract power from the front end under high overload, may be possible. This will have less effect on the noise figure for the radios. Other sensitivities occur in large satellites because the whole satellite acts as antenna, and large electrical transients can blow out power converters. Thinsats are small, run on much lower voltage, and the power absorbing loops must be designed out to avoid inadvertent mag-torquing. The power supply itself is a solar cell, which acts as a huge bypass diode, so it is extremely robust to power transients. Damaging power transients can be avoided. Thinsats are 7x closer to the upper atmosphere than GEO, where EMP attacks originate, so they will pick up 49x the energy. However, they also have a smaller view angle than a GEO satellite, so most will be completely shielded by the earth from single explosions. The constellations may be severely damaged by one EMP burst, but most will survive. LEO EMP attacks can only be mounted by nation-states with significant capabilities: nuclear weapons, and (at least) suborbital launch capability. This is not something that can be done as a stealthy act. The nation-state that launches such an attack is not likely to survive the consequences, and the leaders will probably suffer the Saddam option. On the other hand, if server-sky is developed as an asset to all nations, then the attacker nation suffers from the attack as well, with the leaders removed by their own internet users. The best defense against attack is to be costly to attack, and beneficial to survive, for all the potential attackers. Nations become their own hostages. Overall, EMP is a useful problem to consider. Many other problems, such as static discharges between too-close thinsats, will behave similarly to EMP pulses. In manufacturing, ESD impulses can damage circuits, and ESD protection is important to maximize product yield. So, "design for EMP" may help with "design for ESD". Since thinsats are cheap, testing to destruction is also cheap, and we can learn a lot about armoring thinsats through test and observation. \Keith Lofstrom\ -----