I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ *AMSAT(R) Flight Software*
On AO-51, Jim White WD0E wrote a scheduler that ran on the satellite. We created schedules that executed mode changes at desired times (at 0000Z 5/19 execute Mode 3) where mode 3 was V/s repeater, or something similar. It worked very well. I could load schedules at my leisure and mode changes would be exact times, not “during the first pass over EL88 on Tuesday night”.
73, Drew
From: LTM ltm-bounces@amsat.org On Behalf Of Burns Fisher (AMSAT) via LTM Sent: Monday, May 18, 2020 9:38 AM To: Linear Transponder Module ltm@amsat.org Subject: [LTM] Scheduled automatic changes on LTM-1 and Golf?
I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ
AMSAT(R) Flight Software
Thanks, Drew. Did AO-51 have a fairly exact UTC on board? Do you know how it got it? GPS? 73,
Burns Fisher, WB1FJ *AMSAT(R) Flight Software*
On Mon, May 18, 2020 at 10:20 AM Andrew Glasbrenner < glasbrenner@mindspring.com> wrote:
On AO-51, Jim White WD0E wrote a scheduler that ran on the satellite. We created schedules that executed mode changes at desired times (at 0000Z 5/19 execute Mode 3) where mode 3 was V/s repeater, or something similar. It worked very well. I could load schedules at my leisure and mode changes would be exact times, not “during the first pass over EL88 on Tuesday night”.
73, Drew
*From:* LTM ltm-bounces@amsat.org *On Behalf Of *Burns Fisher (AMSAT) via LTM *Sent:* Monday, May 18, 2020 9:38 AM *To:* Linear Transponder Module ltm@amsat.org *Subject:* [LTM] Scheduled automatic changes on LTM-1 and Golf?
I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ
*AMSAT**(R) Flight Software*
Internal clock. It drifted a little bit, but we commanded it back correct when it got too far off.
On May 18, 2020, at 10:22 AM, Burns Fisher (AMSAT) wb1fj@fisher.cc wrote:
Thanks, Drew. Did AO-51 have a fairly exact UTC on board? Do you know how it got it? GPS? 73,
Burns Fisher, WB1FJ AMSAT(R) Flight Software
On Mon, May 18, 2020 at 10:20 AM Andrew Glasbrenner glasbrenner@mindspring.com wrote: On AO-51, Jim White WD0E wrote a scheduler that ran on the satellite. We created schedules that executed mode changes at desired times (at 0000Z 5/19 execute Mode 3) where mode 3 was V/s repeater, or something similar. It worked very well. I could load schedules at my leisure and mode changes would be exact times, not “during the first pass over EL88 on Tuesday night”.
73, Drew
From: LTM ltm-bounces@amsat.org On Behalf Of Burns Fisher (AMSAT) via LTM Sent: Monday, May 18, 2020 9:38 AM To: Linear Transponder Module ltm@amsat.org Subject: [LTM] Scheduled automatic changes on LTM-1 and Golf?
I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ
AMSAT(R) Flight Software
I concur with Drew's recollections. We uploaded the schedule file to the satellite.
Now I diverge---Remember, AO-51 (and Falconsat-3, a close cousin...) ran/run very elaborate OS and file systems, to support the Pacsat broad cast protocol BBS. That has advantages and disadvantages, as you know better than I do. But on Falconsat-3, the clock runs fast. I forget...a couple minutes per month? What I think I learned---if we leave the bird alone, and just let the clock keep going, even though it's fast--that's best. Just leave it run fast... I correlate commanding a change in time--back in time, mind you---with having to reload the bird! I suspect it's because we don't stop all the filesystems software....we do shut the BBS, but all the time stamps, etc. are now "in the future" from the spacecraft reset clock....anyhow, right now, FS-3 has been up continuously for over 2 years...maybe 2.5 by now. It just keeps going and going.... :)
73, Mark N8MH
On Mon, May 18, 2020 at 10:26 AM Andrew Glasbrenner via LTM ltm@amsat.org wrote:
Internal clock. It drifted a little bit, but we commanded it back correct when it got too far off.
On May 18, 2020, at 10:22 AM, Burns Fisher (AMSAT) wb1fj@fisher.cc wrote:
Thanks, Drew. Did AO-51 have a fairly exact UTC on board? Do you know how it got it? GPS? 73,
Burns Fisher, WB1FJ *AMSAT(R) Flight Software*
On Mon, May 18, 2020 at 10:20 AM Andrew Glasbrenner < glasbrenner@mindspring.com> wrote:
On AO-51, Jim White WD0E wrote a scheduler that ran on the satellite. We created schedules that executed mode changes at desired times (at 0000Z 5/19 execute Mode 3) where mode 3 was V/s repeater, or something similar. It worked very well. I could load schedules at my leisure and mode changes would be exact times, not “during the first pass over EL88 on Tuesday night”.
73, Drew
*From:* LTM ltm-bounces@amsat.org *On Behalf Of *Burns Fisher (AMSAT) via LTM *Sent:* Monday, May 18, 2020 9:38 AM *To:* Linear Transponder Module ltm@amsat.org *Subject:* [LTM] Scheduled automatic changes on LTM-1 and Golf?
I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ
*AMSAT**(R) Flight Software*
LTM mailing list LTM@amsat.org https://www.amsat.org/mailman/listinfo/ltm
Thanks, Mark. I understand the "setting the time backwards" issue for a "real OS" and that's a good reminder. We don't really have a real OS on the main flight computer of Golf, but there IS on the SDR computer. In any case, good to remember that even if the clock is reset via GPS, we still have to worry about time going backwards. (BTW, this is why the main s/c time is kept as "resets+seconds" so that it always increases.)
73,
Burns Fisher, WB1FJ *AMSAT(R) Flight Software*
On Mon, May 18, 2020 at 11:35 AM Mark L. Hammond marklhammond@gmail.com wrote:
I concur with Drew's recollections. We uploaded the schedule file to the satellite.
Now I diverge---Remember, AO-51 (and Falconsat-3, a close cousin...) ran/run very elaborate OS and file systems, to support the Pacsat broad cast protocol BBS. That has advantages and disadvantages, as you know better than I do. But on Falconsat-3, the clock runs fast. I forget...a couple minutes per month? What I think I learned---if we leave the bird alone, and just let the clock keep going, even though it's fast--that's best. Just leave it run fast... I correlate commanding a change in time--back in time, mind you---with having to reload the bird! I suspect it's because we don't stop all the filesystems software....we do shut the BBS, but all the time stamps, etc. are now "in the future" from the spacecraft reset clock....anyhow, right now, FS-3 has been up continuously for over 2 years...maybe 2.5 by now. It just keeps going and going.... :)
73, Mark N8MH
On Mon, May 18, 2020 at 10:26 AM Andrew Glasbrenner via LTM ltm@amsat.org wrote:
Internal clock. It drifted a little bit, but we commanded it back correct when it got too far off.
On May 18, 2020, at 10:22 AM, Burns Fisher (AMSAT) wb1fj@fisher.cc wrote:
Thanks, Drew. Did AO-51 have a fairly exact UTC on board? Do you know how it got it? GPS? 73,
Burns Fisher, WB1FJ *AMSAT(R) Flight Software*
On Mon, May 18, 2020 at 10:20 AM Andrew Glasbrenner < glasbrenner@mindspring.com> wrote:
On AO-51, Jim White WD0E wrote a scheduler that ran on the satellite. We created schedules that executed mode changes at desired times (at 0000Z 5/19 execute Mode 3) where mode 3 was V/s repeater, or something similar. It worked very well. I could load schedules at my leisure and mode changes would be exact times, not “during the first pass over EL88 on Tuesday night”.
73, Drew
*From:* LTM ltm-bounces@amsat.org *On Behalf Of *Burns Fisher (AMSAT) via LTM *Sent:* Monday, May 18, 2020 9:38 AM *To:* Linear Transponder Module ltm@amsat.org *Subject:* [LTM] Scheduled automatic changes on LTM-1 and Golf?
I may have mentioned this before, but saw no answer. But the behavior of HO-107, and someone mentioning the schedule for AO-27 got me thinking about it again.
This question is largely for the ops team:
Suppose that the LTM-1 and Golf-Tee had the capability of commanding a schedule of operations. Let's say you could upload one or more sequences of commands labelled n. What would a schedule look like? Would it be "every n hours execute sequence x" realizing that 'n' is not terribly exact and would drift against UTC. Would it be "do sequence x entering eclipse and sequence y exiting eclipse? Would it be "Execute sequence x when I send an "execute sequence x" command?
I think those are all possible, at least if the sequences are simple. No promises--there is a lot of software to do to make Golf-Tee and LTM-1 work at all, but if I get any feedback on these general ideas, I can write something up to prioritize and potentially keep in our back pocket for implementation as a "may" rather than a "must".
73,
Burns Fisher, WB1FJ
*AMSAT**(R) Flight Software*
LTM mailing list LTM@amsat.org https://www.amsat.org/mailman/listinfo/ltm
-- Mark L. Hammond [N8MH]
participants (3)
-
Andrew Glasbrenner -
Burns Fisher (AMSAT) -
Mark L. Hammond