I agree, particularly given the redundancy we're discussing.
One thing we'll have to look hard at as design progresses -- common-mode failure.
73, Jim wb4gcs@amsat.org
Phil Karn wrote:
Jim Sanford wrote:
Lou: Is it time to write a strawman concept proposal & get people to review some specific proposals? Or do we need to wait on the results of your experiments?
I'd like to see measurements of the accuracy and variability of the voltage references used in the various DC-DC converters. This knowledge is necessary to set the voltage dead-bands in the battery module controllers. We must be able to precisely control, over all possible temperatures and operating conditions, the allocation of load and charging currents to the individual battery modules.
This would provide maximum versatility and reliability in managing a variety of energy storage devices (Li-ion batteries, NiMH batteries, NiCd batteries, supercaps, etc). Here are some of the things we could do with this system:
- Chemical batteries live longer (i.e., produce more amp-hours over
their lifetimes) with shallow discharge. So if we have more than one battery of each type, we want to match their states of charge as much as possible.
- Li-ion batteries live longest at about 50% charge, so we should try
to maintain them at that level.
- Because supercaps have (or are claimed to have) a much longer cycle
life than chemical batteries, during eclipse we'd probably want to discharge them completely before we start to discharge any of our chemical batteries.
- For test purposes, it would be useful to selectively cycle
(discharge, recharge and equalize) a given battery module independently of the others and of the spacecraft's power balance. This might require a dummy load in some situations, e.g., when we want to test a battery under heavy discharge and there's no other place to dump the energy.
All these things require the ability to program arbitrary, accurate and stable I/V curves into each battery controller. All this may seem rather complex, but given that battery failure has ended most of our past missions I think the added complexity would be more than worth it.
--Phil