On 12/7/2020 13:27, John Brier wrote:
I hope for this to be a purely engineering based discussion that could shed some light on this question.
Please only respond from an engineering perspective in this thread.
Data:
According to the "AMSAT Fox-1 Systems Engineering Documentation" the satellites were expected to last at least five years. [1]
It's nice to see this discussion going on, a thank you to those who join in.
I want to comment on the five year expectation. That certainly is/was the goal for the Fox-1 series however there are some things that I think soften that requirement statement, which as Burns mentioned was also made "before my time". Not that I disagree with it. As the overall project was a first, given the reduction down to a 1U CubeSat albeit packed with stuff, there were assumptions made based on data and many times minimal or lack of data. The five years is overall and not specifically based on battery lifetime, in fact operation after the battery died was part of the original napkin and so it was a consideration in the five years as well.
The cells originally scoped were KR1700 and were out of production at the time I joined in 2011. As the years passed and the cells sat awaiting someone who knew matching, by 2014 when we did find someone who could do that it turns out that those cells were by now less capacity than the KR1400. Which were also out of production. Because of the A size form factor designed and being built, we were rather stuck to using old NiCd.
Only five sets were made because of the lack of quantity of those cells. Further newer cells were rejected as opinions of our battery providers was that they were of a much lower quality as intended for "toys" and none would recommend that we put them on our satellites.
The battery circuitry was designed for simple charging, no cycling. I think that given one Fox-1 and an expectation that it would launch right quickly after initial designs, no one understood that it would be years before the batteries flew. I considered the possibility of trying to properly tend the battery cells with Fox-1B and -E but there was really no way to just plug a tender in the umbilical as blocking and voltage reduction diodes would prevent discharge, although you could discharge it using the voltage (DVM) pin, but that would require a specifically wired tender and we determined that the risk was too high. Of course, even AO-85 had not failed yet at that time.
And the cells sit awaiting launch not only in Fox Labs where I could at least charge them, but in the case of AO-85 for 7 months it was untouchable being integrated and out of our hands, prior to launch. AO-91 -B did have a much faster turnaround from the Labs to launch, about 2 months IIRC without looking. AO-92 -D was years from completion to launch although somewhere around 3 months from integration to launch, again by memory on the number.
The basis of the battery cell lifetime was essentially depth of discharge, to keep that at 20% or less per orbit. I do not know the how and why for that decision, sorry, again that was a before my time thing. In any case none of the Fox-1 except AO-95 -Cliff experience that little DOD on a regular basis anyway. Interestingly as this comes about, Fox-1Cliff AO-95 is the one that has become a benchmark in a sense, as it certainly has less than 20% DOD. How its cells behave and how long may be useful for conclusions on this battery behavior "study". Unfortunately being one of the early, actual the second, Fox-1 builds it does not have thee specific cell telemetry either but only total voltage of the battery.
As you see there are a lot of unseen things that you may not all be aware of, although I have discussed them before either in Symposium or Journal columns here and there, and most recently because of what is happening. The five year lifetime as a whole was based on best understanding of what was being done, and the design, but had deviations introduced due to lack of knowledge and time to delivery. And here we are in an era now, where getting a launch to an orbit that even lasts five years might put the EOL on re-entry instead of battery life for Fox-1E.
I appreciate the community discussion of what's going on, I am curious as heck about it but never have time to actually do anything to satisfy that curiosity... While the outcome may not be of use for future satellites given the changes in stored power it is definitely part of AMSAT's purpose to learn and educate from everything we do. And it could be fun! (Maybe Paul will send a commemorative QSO to whoever arrives at the definitive answer first.)
Jerry Buxton, NØJY