"Team Possible" status - Gypsy & Lil' Joe are on their way!
2 views
Skip to first unread message
Brian Davis
Jul 18, 2008, 4:03:53 PM7/18/08
to HALE TEAMS
Well, given that the first deadline is 21 Jul, I sent my two payloads
out today - they are supposed to reach Reno on Monday. The hardware
has been tested as best I can - with "Lil' Joe", this involved driving
down the road with a parachute fluttering out the driver's side
window, but at significantly lower speeds and much higher air
densities than it will experience in practice. For "Gypsy", it was
tested in the freezer (heaters and temperature sensors functioned, SW
could control the system), as well as probably close to 20 hours of
hanging in my entryway, making sure the pre-programmed script was free
of bugs (amazingly, got it in one, but then went on to test the HW
still more and refine the script).
The first draft (last draft?) of the unpacking, preparation, and
launch instructions for both payloads are up in the files of this
group. Additionally, I have some images already up in Brickshelf with
more coming, & hopefully some brief video of packing the payloads to
help explain the instructions. Please let me know what you think, and
where I'm sorely lacking in documentation.
I was really struggling for a way to close the payload easily, as well
as attach my "remove before flight" trigger. While despondently
looking over the fantastic documentation on the LUXPAC site (nice job
guys!), I realized that velcro was just about the ideal solution for
both my problems, saving weight and making the payload very easy to
work with - Thank you Claude & company!
A brief (well, brief for me) summary of the payloads...
Lil' Joe: designed to test the idea of a free-falling "skydiving" NXT
payload. the NXT & a SPOT sit at the bottom of a tall squarish payload
shell, with the "rear" being nothing more than a flat piece of foam
with a tailfin that extends beyond the edges of the payload body. It
will take 3-axis timestamped acceleration data rapidly, compressing
and storing it to a file which is periodically erased. When a free-
fall condition is detected (less than 0.25 G's for more than 0.5
seconds), it will stop erasing this file and just add to it, thereby
capturing the actual drop conditions as well as the following free-
fall. Based on how long it's been since it was turned on, it will
calculate a conservative drop time based on how long it has been
ascending (it's important not to turn it on hours before the flight
for this reason... rats, something else I need to add to the
instructions). After a brief free-fall period, it will release the
tailfin cap, which will be pulled behind the main body to act like a
deployment drogue for the main parachute. If all goes well, the main
'chute will deploy and Lil' Joe will float (OK, semi-plummet at 1000
ft/min or faster) to a soft (OK, pretty hard) landing (semi-crash) in
the friendly terra-firma (make that "scorchingly hot and rocky
desert") below. Here's hoping.
Gypsy: designed as a flexible camera platform, the payload will
control a camera, switching the camera on and off to conserve
batteries, as well as switching between video & still photo modes. The
third motor will control the pitch of the entire payload, allowing the
camera to be pointed anywhere between straight up and straight down.
This will allow pictures up and down the payload string as well as
look-down views of the Earth and other variety. The camera platform
will by controlled by a pre-written "script", a file with a series of
commands ("i" = take a still sequence of N photos, "v0" = start/stop a
video, "v1" = turn camera on/off, "p" = pitch platform to a setting of
N degrees) followed by a wait interval ("wait X seconds until looking
for a new command"). When free-fall is detected, the program should
switch immediately to a second script file for commands to follow
during descent. Interfacing LEGO with the camera has proved... tricky,
but hopefully I'll at least get some video and pictures.
The other function of Gypsy is datalogging the mission - it
incorporates a HiTechnic accelerometer, a custom high-resolution
pressure sensor (with included thermal sensor for temperature
compensation, also prototyped by HiTechnic), a stock LEGO sound
sensor, and a HiTechnic prototyping board that controls four other
items: a temperature sensor on the NXT battery compartment, a
temperature sensor near the camera battery, a CDS photosensor that
sticks through the back of the payload shell, and a reed switch to
allow the NXT to turn the heaters on and off (the protoboard could
actually handle much more than this, but it's all I had time for). The
NXT will cycle the heaters on any time either NXT or camera
temperature sensor drop too low, and cycle the heaters off when they
are both warm again (the heaters are driven by 4 AAA NiMH batteries).
Every 5 seconds the NXT will log the pressure, temperature, and light
level as reported by the CDS cell. Every 20 seconds (if memory serves)
it will log a series of engineering measurements, including the other
two temperatures, the NXT battery level, the current position of the
pitch motor, and if the heaters are currently commanded on or off.
Every 30 seconds, it will log the average sound level, peak sound
level, and peak net acceleration during the previous 30 sec interval.
Any time a camera command is executed, this will also be timestamped
and logging into the data file. If the platform pitch is changed, the
NXT will also acquire more detailed sound information - the ambient
(or background) sound level, the sound level 0.5 seconds into the
pitch change, and the sound level produced by a 100% volume
"Beep" (middle A, 440 Hz) produced by the NXT during an otherwise
quite period. The logging rates are not fixed, but can be changed by
the program at any time (for instance, upon free-fall the logging of
the ambient conditions of pressure is increased to a higher rate).
If anyone wants to see the script for the Gypsy mission, I can post it
here. I'm really not sure how much of it will happen, as it depends on
the camera working per specification during the whole mission... which
has only happened in about 60% of my trials. Next time I may give in
and cut the camera housing open.
Oh, and both payloads have the required important minifigs. Lil' Joe
has Flex (from the Alpha Team) carefully secured in a virtual cage of
LEGO, in hopes that he stays in (after parachute deployment, Lil' Joe
is open to the thin but very cold air up there... I wish I had a
temperature sensor for it). Riding on Gypsy are my sig-fig Larry
(wearing armor and sitting by the HT Protoboard), a happy Rebel X-wing
pilot, and a "space pirate" with breathing gear submitted (risked?) by
my son.
--
Brian "who says I write too much?" Davis
out today - they are supposed to reach Reno on Monday. The hardware
has been tested as best I can - with "Lil' Joe", this involved driving
down the road with a parachute fluttering out the driver's side
window, but at significantly lower speeds and much higher air
densities than it will experience in practice. For "Gypsy", it was
tested in the freezer (heaters and temperature sensors functioned, SW
could control the system), as well as probably close to 20 hours of
hanging in my entryway, making sure the pre-programmed script was free
of bugs (amazingly, got it in one, but then went on to test the HW
still more and refine the script).
The first draft (last draft?) of the unpacking, preparation, and
launch instructions for both payloads are up in the files of this
group. Additionally, I have some images already up in Brickshelf with
more coming, & hopefully some brief video of packing the payloads to
help explain the instructions. Please let me know what you think, and
where I'm sorely lacking in documentation.
I was really struggling for a way to close the payload easily, as well
as attach my "remove before flight" trigger. While despondently
looking over the fantastic documentation on the LUXPAC site (nice job
guys!), I realized that velcro was just about the ideal solution for
both my problems, saving weight and making the payload very easy to
work with - Thank you Claude & company!
A brief (well, brief for me) summary of the payloads...
Lil' Joe: designed to test the idea of a free-falling "skydiving" NXT
payload. the NXT & a SPOT sit at the bottom of a tall squarish payload
shell, with the "rear" being nothing more than a flat piece of foam
with a tailfin that extends beyond the edges of the payload body. It
will take 3-axis timestamped acceleration data rapidly, compressing
and storing it to a file which is periodically erased. When a free-
fall condition is detected (less than 0.25 G's for more than 0.5
seconds), it will stop erasing this file and just add to it, thereby
capturing the actual drop conditions as well as the following free-
fall. Based on how long it's been since it was turned on, it will
calculate a conservative drop time based on how long it has been
ascending (it's important not to turn it on hours before the flight
for this reason... rats, something else I need to add to the
instructions). After a brief free-fall period, it will release the
tailfin cap, which will be pulled behind the main body to act like a
deployment drogue for the main parachute. If all goes well, the main
'chute will deploy and Lil' Joe will float (OK, semi-plummet at 1000
ft/min or faster) to a soft (OK, pretty hard) landing (semi-crash) in
the friendly terra-firma (make that "scorchingly hot and rocky
desert") below. Here's hoping.
Gypsy: designed as a flexible camera platform, the payload will
control a camera, switching the camera on and off to conserve
batteries, as well as switching between video & still photo modes. The
third motor will control the pitch of the entire payload, allowing the
camera to be pointed anywhere between straight up and straight down.
This will allow pictures up and down the payload string as well as
look-down views of the Earth and other variety. The camera platform
will by controlled by a pre-written "script", a file with a series of
commands ("i" = take a still sequence of N photos, "v0" = start/stop a
video, "v1" = turn camera on/off, "p" = pitch platform to a setting of
N degrees) followed by a wait interval ("wait X seconds until looking
for a new command"). When free-fall is detected, the program should
switch immediately to a second script file for commands to follow
during descent. Interfacing LEGO with the camera has proved... tricky,
but hopefully I'll at least get some video and pictures.
The other function of Gypsy is datalogging the mission - it
incorporates a HiTechnic accelerometer, a custom high-resolution
pressure sensor (with included thermal sensor for temperature
compensation, also prototyped by HiTechnic), a stock LEGO sound
sensor, and a HiTechnic prototyping board that controls four other
items: a temperature sensor on the NXT battery compartment, a
temperature sensor near the camera battery, a CDS photosensor that
sticks through the back of the payload shell, and a reed switch to
allow the NXT to turn the heaters on and off (the protoboard could
actually handle much more than this, but it's all I had time for). The
NXT will cycle the heaters on any time either NXT or camera
temperature sensor drop too low, and cycle the heaters off when they
are both warm again (the heaters are driven by 4 AAA NiMH batteries).
Every 5 seconds the NXT will log the pressure, temperature, and light
level as reported by the CDS cell. Every 20 seconds (if memory serves)
it will log a series of engineering measurements, including the other
two temperatures, the NXT battery level, the current position of the
pitch motor, and if the heaters are currently commanded on or off.
Every 30 seconds, it will log the average sound level, peak sound
level, and peak net acceleration during the previous 30 sec interval.
Any time a camera command is executed, this will also be timestamped
and logging into the data file. If the platform pitch is changed, the
NXT will also acquire more detailed sound information - the ambient
(or background) sound level, the sound level 0.5 seconds into the
pitch change, and the sound level produced by a 100% volume
"Beep" (middle A, 440 Hz) produced by the NXT during an otherwise
quite period. The logging rates are not fixed, but can be changed by
the program at any time (for instance, upon free-fall the logging of
the ambient conditions of pressure is increased to a higher rate).
If anyone wants to see the script for the Gypsy mission, I can post it
here. I'm really not sure how much of it will happen, as it depends on
the camera working per specification during the whole mission... which
has only happened in about 60% of my trials. Next time I may give in
and cut the camera housing open.
Oh, and both payloads have the required important minifigs. Lil' Joe
has Flex (from the Alpha Team) carefully secured in a virtual cage of
LEGO, in hopes that he stays in (after parachute deployment, Lil' Joe
is open to the thin but very cold air up there... I wish I had a
temperature sensor for it). Riding on Gypsy are my sig-fig Larry
(wearing armor and sitting by the HT Protoboard), a happy Rebel X-wing
pilot, and a "space pirate" with breathing gear submitted (risked?) by
my son.
--
Brian "who says I write too much?" Davis
Eric
Jul 18, 2008, 10:52:40 PM7/18/08
to HALE TEAMS
Great!! Can't wait to drop it :)
Brian,
We are hoping you and some others would be willing to blog (live)
about the mission. We'll have a satellite phone and can feed you blow
by blow info.
If we are within cell phone range, we'll also email photos from our
iPhones.
Eric
Brian,
We are hoping you and some others would be willing to blog (live)
about the mission. We'll have a satellite phone and can feed you blow
by blow info.
If we are within cell phone range, we'll also email photos from our
iPhones.
Eric
Eric
Jul 18, 2008, 11:04:14 PM7/18/08
to HALE TEAMS
Just a quick update: so far we've received payloads from Luxembourg
and Boston. Taiwan is on its way and now so is Brian's. We will do a
final payload count on Monday and start the full mission preparation.
We will assign a "highly trained monkey" to be responsible for each
payload. He/she will test each payload (don't worry, we'll ask for
permission first) to make sure he/she know exactly what's needed to
get it running. We'll also load fresh Energizer batteries and top off
any NiMh. We're starting to get excited!!!
We also are hoping to post a short page on each payload (or a link to
an existing webpage) online early next week. If you have a link that
or a page that you can email me, that would be great.
and Boston. Taiwan is on its way and now so is Brian's. We will do a
final payload count on Monday and start the full mission preparation.
We will assign a "highly trained monkey" to be responsible for each
payload. He/she will test each payload (don't worry, we'll ask for
permission first) to make sure he/she know exactly what's needed to
get it running. We'll also load fresh Energizer batteries and top off
any NiMh. We're starting to get excited!!!
We also are hoping to post a short page on each payload (or a link to
an existing webpage) online early next week. If you have a link that
or a page that you can email me, that would be great.
Brian Davis
Jul 23, 2008, 8:20:19 PM7/23/08
to HALE TEAMS
A little bit of follow-up: while Eric has already modified the Lil'
Joe payload for easier preparation and launch, I did put together a
short video to help out. Anyway, here it is:
http://www.youtube.com/watch?v=EI1hvtSx4n8
I'll probably not have time to do one for Gypsy, but I'll see - I'm
heading out of town this weekend (I'll have my iPhone, Eric, if
needed).
--
Brian Davis
Joe payload for easier preparation and launch, I did put together a
short video to help out. Anyway, here it is:
http://www.youtube.com/watch?v=EI1hvtSx4n8
I'll probably not have time to do one for Gypsy, but I'll see - I'm
heading out of town this weekend (I'll have my iPhone, Eric, if
needed).
--
Brian Davis
Reply all
Reply to author
Forward
0 new messages