[X] My Assistant

[dvandorn]

Just saw an interesting (and disturbing) report on CNN. As y'all have probably heard, as soon as the new set of solar arrays was deployed and activated, the computers in the Russian segment of the ISS went offline. Attempts to reboot them and bring them back online have failed. The best they have done is establish "limited communication" with a couple of the "lanes" of communication between the computers (the main and terminal systems) on the Russian segment and the equipment they control, but that the communication is not trustworthy. They've now been shut down totally while ground experts study the problem.

According to the Russians, the best guess right now is that the new American solar arrays are generating an induced EM field that is blowing the Russian computers out of the water. There may be no choice but to leave the Russian segment without its main control systems until the computers can be replaced -- which is not a simple process.

This isn't a minor glitch, folks -- for one thing, there is apparently no way to use the Russian attitude control system without the computers. If for any reason the primary attitude control system -- the CMGs -- has to be taken offline (due to gyro saturation or any other reason), this would leave the entire complex uncontrolled in attitude. The Vozhduk (sp?) air regerator system is also controlled by the now-disabled computer system.

For me, this kind of thing is why you have to gain experience in deploying large structures in LEO before trying to send them elsewhere in the Solar System. Certainly the best American and Russian engineers never considered the possibility of induced EM fields from the solar arrays that could have such a devastating effect on the internal electronics. In my humble opinion, this is *exactly* the kind of learning experience that ISS must provide before humans can move out to the asteroids and planets. Let's just hope it's not a show-stopping lesson...

-the other Doug

[mcaplinger]

According to the Russians, the best guess right now is that the new American solar arrays are generating an induced EM field that is blowing the Russian computers out of the water.

That sounds like BS to me. Avionics, in the US at least, are extensively tested for electromagnetic interference problems. For digital electronics, shielding is not very difficult. And if there are sensitive sensors like magnetometers involved, they should have thought of these sorts of problems long before now.

[nprev]

For digital electronics, shielding is not very difficult. And if there are sensitive sensors like magnetometers involved, they should have thought of these sorts of problems long before now.

I beg to differ in a way on a different tangent, MC. True, the boxes themselves are undoubtedly shielded against EMI, but it seems that what's going on is a massive problem with input power. "Clean" line power is enormously important for digital electronics; it really doesn't take much fuzz in this area to flip a bit, with unfortunate consequences.

On another level, though, I agree with you; a rigorous risk analysis may well have revealed this possibility, and good system engineering should have identified this risk and provided mitigation procedures as well as design changes to avoid the risk entirely.

I've heard reports that a defective power switch seems to be the root cause (I presume that it's providing a less-than-ideal conductive path, with consequent loading), and the workaround has been, God help them, a set of jumper cables.

All that being said, I have to wonder about power filtration onboard the ISS in general. Surely the designers at least anticipated transients from solar phenomena and unequal surface charging across the full area of the station; don't these mission-critical computers have something like a FERRUPS to smooth input power and provide backup capability? (Disclaimer: That link is not intended as an endorsement of any product or service.)

[mcaplinger]

I beg to differ in a way on a different tangent, MC. True, the boxes themselves are undoubtedly shielded against EMI, but it seems that what's going on is a massive problem with input power. "Clean" line power is enormously important for digital electronics; it really doesn't take much fuzz in this area to flip a bit, with unfortunate consequences.

A large fraction of MIL-STD-461 testing is associated with conducted emissions and susceptibility (what you're calling line power) and typical avionics will operate through all kinds of horrible noise on input power lines, by design. Typically one uses box-level DC-DC converters with extensive EMI filtering. ISS uses 120Vdc main power distribution (at least the American side does.)