Jerry -thanks for that detailed explanation.
One of the challenges I think with the Fox design is the way the manifesting and integrating cubesats may affect battery lifetime in unexpected ways compared to historical experience with spacecraft. As best I recall, in 1992-93, the NiCd batteries used in what became AO-27 and IO-26 were sourced from someone working with SSTL who had NiCd matching down to a science and was able to work with relatively newer cells. I think the SSTL satellites were using the same batteries, and I don't recall supply chain / discontinued product issues Fox had to deal with. Also, AO-27 was accessible to the development team until a few weeks at most before launch when integration of all 6 secondary payloads was completed. I believe there were provisions to keep batteries charged before launch but not 100% certain. In contrast, the realities of cubesat launches almost guarantee a lengthy gap between final spacecraft integration and launch, with no opportunities for battery maintenance. So the requirements for the battery in a future mission might include "shall survive without maintenance for xx months until launch and deployment without adverse effect on battery lifetime" or some such (easy to spec, probably way harder than I know to deliver!).
Managing depth of discharge has been vital for long battery life. AO-27 was managed very aggressively - 15-20 minutes on early in it's life, 4 min at a time today. Something is better than nothing? I'm not suggesting we should have done that with the Fox satellites, I am curious if there is data from SO-50 as it continues to be used heavily and seems fine after 18 years in space. Keep in mind these satellites are larger compared to Fox, have larger capacity cells, less dramatic thermal changes, operate at a higher bus voltage (AO-27 is nominal 10v) and there may be changes between the early-mid 1990s NiCd battery chemistry and manufacture vs. 2010s that individually and in sum affect longevity in space. Tradeoffs are tough design decisions that have to be made to get from something perfect but unbuildable to feasible and flying.
There are a bunch of lessons here affecting all sorts of requirements and resulting engineering tradeoffs and operations decisions - battery capacity and chemistry, supply availability and shelf life before integration, self-discharge, ability to monitor individual cells, ability to manage depth of discharge after assembly and from integration until launch, and managing the battery once in space. More capability also means more circuitry, more software code, and more places something has to be tested and could fail. And have sufficient time in development and test to validate and incorporate good ideas that improve reliability and longevity (we hams may be the only people who would get excited about using satellites 46 years old (AO-7), 27 (AO-27), turning 8 (SO-50), or failed to deploy from the launcher (RS-44)!)
Steve KS1G