On 8/22/11 9:18 PM, Greg D. wrote:
Ah, interesting. So, values around +/- 0 mean that the battery is (or thinks it is) fully charged, and the satellite is running on just the solar panels. Since that happens pretty quickly after start-up (low MET values), that certainly supports the conclusion that the battery has nearly zero capacity left in it.
But if it's "fully charged", there must be something else going on
I think it supports the conclusion that the battery has become an open circuit. The voltage wouldn't get that high if multiple cells had shorted. This is lucky for us since the satellite can continue to operate on the solar panels when lit.
I'm really beginning to think about new ideas for satellite power systems, especially Lou McFaddin's idea (published some years ago at the Symposium) for a modular power system with a parallel bus. Each battery is in a module with its own microprocessor and DC-DC converter to match the standard bus voltage, and each is designed to remove itself from the bus when its battery fails.
What's nice about this is that it's very easy to combine different battery technologies and even supercaps and have the system degrade gracefully over time. The computer would keep track of the total available capacity, and at first it could operate all systems normally through eclipse. As modules fail, it would shut down non-essential systems one by one as necessary during during eclipse until the sun reappears. The highest priority load would be the IHU memory. With some supercaps in the mix it should always be able to keep the computer alive. That would avoid a lot of spurious computer resets and associated loss of information like recorded telemetry and store-and-forward messages.
-Phil, KA9Q