NASA UP CLOSE
Was lucky to have a personal tour of NASA's UAV program, in the restricted area of Moffet field. Quite a contrast to the cattle herding of the KSC tour bus. Seeing a NASA division from the point of view of a NASA employee & eating with the people in the employee cafeteria this time was a thrill. After struggling to photograph NASA Ames in many a blog post from Fremont, here we were actually in the restricted area.
Honestly wasn't happy with any of our photos of the affair. Maybe it was because we were dying inside & weren't completely functional. Maybe it was because we weren't shown any vehicles we didn't see before & certainly no flying.
The NASA UAV program uses tried & true, basic airframes, tried & true flight control laws, & focuses on the instrumentation & high level algorithms. The most exotic work is on automating full size electric gliders, but these R tried & true systems.
There's no sense of making software do what software's got no business doing.
There was no 3DOF IMU, no Marcy Foam, no tri rotor, no spin copter, no unidirectional sonar, no flybarless copter, no EDF. They also maintain their vehicles with the same protocols as manned aircraft, hence no crashes.
We really finished the Jack Crossfire blog in Air Force Heroine's stomping ground. It's hard to get any more immersed in Her world than that, but it was beyond our means to meet Air Force Heroine, no matter how hard God pushed.
THE FINAL FLIGHT
Stepped up our final altitude to 90m. Wanted to be sure we got something on this last campaign. 90m ended up just out of range for XBee Pros, but got lucky on 3 flights. She yawed around during a radio dropout & that ended flight #4.
Have the video. You're going straight up to 90m above Mountain View, panning around the area, & landing straight down on computer control. At 90m it feels like real flying & not RC. That's followed by the final flight in which radio contact dropped below limits, but it had better lighting while it lasted.
Putting our last flight 90m above Mountain View was the Everest, the grand finale, the piece de resistance, real pain. We're done.
The Jack Crossfire story continues on http://heroineworshiper.blogspot.com/. Learn all about Major Marcy, read off topic blog posts, see the world like U've never seen it, read tomorrow's news today.
Was lucky to have a personal tour of NASA's UAV program, in the restricted area of Moffet field. Quite a contrast to the cattle herding of the KSC tour bus. Seeing a NASA division from the point of view of a NASA employee & eating with the people in the employee cafeteria this time was a thrill. After struggling to photograph NASA Ames in many a blog post from Fremont, here we were actually in the restricted area.
Honestly wasn't happy with any of our photos of the affair. Maybe it was because we were dying inside & weren't completely functional. Maybe it was because we weren't shown any vehicles we didn't see before & certainly no flying.
The NASA UAV program uses tried & true, basic airframes, tried & true flight control laws, & focuses on the instrumentation & high level algorithms. The most exotic work is on automating full size electric gliders, but these R tried & true systems.
There's no sense of making software do what software's got no business doing.
There was no 3DOF IMU, no Marcy Foam, no tri rotor, no spin copter, no unidirectional sonar, no flybarless copter, no EDF. They also maintain their vehicles with the same protocols as manned aircraft, hence no crashes.
We really finished the Jack Crossfire blog in Air Force Heroine's stomping ground. It's hard to get any more immersed in Her world than that, but it was beyond our means to meet Air Force Heroine, no matter how hard God pushed.
THE FINAL FLIGHT
Stepped up our final altitude to 90m. Wanted to be sure we got something on this last campaign. 90m ended up just out of range for XBee Pros, but got lucky on 3 flights. She yawed around during a radio dropout & that ended flight #4.
| The last flight of Jack Crossfire (8 min 14 sec) |
| Last flight in still photos (2 min 12 sec) |
Have the video. You're going straight up to 90m above Mountain View, panning around the area, & landing straight down on computer control. At 90m it feels like real flying & not RC. That's followed by the final flight in which radio contact dropped below limits, but it had better lighting while it lasted.
Putting our last flight 90m above Mountain View was the Everest, the grand finale, the piece de resistance, real pain. We're done.
The Jack Crossfire story continues on http://heroineworshiper.blogspot.com/. Learn all about Major Marcy, read off topic blog posts, see the world like U've never seen it, read tomorrow's news today.
Alright kids. Wednesday we're doing a 60m aerial shoot of Mountain View & then we're done flying.
Flying UAV's on the golf course is just killing us. We're not recovering from Air Force Heroine. We're dying just on routine golf course photo shoots that used to be our vacation. We need a new gig. Enjoy the Marcy Foam, 3DOF IMU, VicaCopter, & spin copter. The blog will move to another time slot.
Now the last of the flying blog notes.
Tough video to watch. Tough video to watch.
Exposed a lot of problems with flag payloads in the wind. Did U see how the tail rotor was so far over it shredded the radio cable? Did U see how horizontal velocity was maximized to keep the flag from pulling it away? Very psychotic. At least the Marcy Foam got improved performance.
spin copter closeout
Lots of questions arose on spin copter. Does the fuselage have too much momentum to slow down for half a revolution? Would the possible speed change give much horizontal velocity? What good is it if you're not also going to have a pushbroom camera?
spaceship blog post
Spaceships don't have accelerometers & magnetometers for sensing attitude. They only use gyros & photograph the stars to reset their IMU's. That makes VicaCopter essentially a spaceship.
MAV blog post
There is a plan to develop 2 systems for a future airframe:
Ground based navigation in polar coordinates:
2 sonar receivers & 1 IR transmitter on the ground
1 sonar transmitter & 2 IR receivers on the airframe
Sense altitude & radius by measuring sonar delay
Sense heading with IR receivers & gyro
Solenoid thrust vectoring:
mount tail motor on spring
Control with direct drive magnet.
problem with plywood fuselage & magnetometer
So basically, the plywood warped when it was installed & may change shape some more in flight. The result is the magnetometer & IMU aren't aligned & she goes into spirals. Spirals R caused by the compass reading deviating because the attitude estimate is wrong. What U need is a $150 tilt compensated compass from $parkfun. Fuggedaboutit.
USING LWNEURALNET RECURRENTLY
So with lwneuralnet, you call straight net_compute's for every step in time. Use a master network & unique inputs for every step. U need to store all the neuron outputs & network inputs for every net_compute step. Network inputs are the first layer outputs.
Then U call net_compute_output_error exactly once on the last iteration. This sets the errors on the last outputs.
Now net_train_batch to begin the training pass. This must be modified. backward_pass must start on the neuron outputs from the last time step & repeat working backward in time. Modify backward_pass to recall the neuron outputs for each time step, copy the first layer's errors in the current time step to the last layer's errors in the previous time step. backward_pass needs to store errors for all time steps instead of accumulating.
adjust_deltas_batch needs to go through all time steps & recall the errors & outputs for each time step. The deltas R properly accumulated in the master network.
Next, call adjust_weights_batch to train the master network. This averages the deltas over the number of time steps.
NOTE ABOUT LEAD COMPENSATION
The velocity term in your PID equations provides rate damping but not lead compensation. To get lead compensation, U need to add velocity to the position. The fully damped & lead compensated equation has a velocity feedback & position + velocity feedback.
Flying UAV's on the golf course is just killing us. We're not recovering from Air Force Heroine. We're dying just on routine golf course photo shoots that used to be our vacation. We need a new gig. Enjoy the Marcy Foam, 3DOF IMU, VicaCopter, & spin copter. The blog will move to another time slot.
Now the last of the flying blog notes.
| Marcy Foam vs. flag in the wind (1 min 56 sec) |
Tough video to watch. Tough video to watch.
Exposed a lot of problems with flag payloads in the wind. Did U see how the tail rotor was so far over it shredded the radio cable? Did U see how horizontal velocity was maximized to keep the flag from pulling it away? Very psychotic. At least the Marcy Foam got improved performance.
spin copter closeout
Lots of questions arose on spin copter. Does the fuselage have too much momentum to slow down for half a revolution? Would the possible speed change give much horizontal velocity? What good is it if you're not also going to have a pushbroom camera?
spaceship blog post
Spaceships don't have accelerometers & magnetometers for sensing attitude. They only use gyros & photograph the stars to reset their IMU's. That makes VicaCopter essentially a spaceship.
MAV blog post
There is a plan to develop 2 systems for a future airframe:
Ground based navigation in polar coordinates:
2 sonar receivers & 1 IR transmitter on the ground
1 sonar transmitter & 2 IR receivers on the airframe
Sense altitude & radius by measuring sonar delay
Sense heading with IR receivers & gyro
Solenoid thrust vectoring:
mount tail motor on spring
Control with direct drive magnet.
problem with plywood fuselage & magnetometer
So basically, the plywood warped when it was installed & may change shape some more in flight. The result is the magnetometer & IMU aren't aligned & she goes into spirals. Spirals R caused by the compass reading deviating because the attitude estimate is wrong. What U need is a $150 tilt compensated compass from $parkfun. Fuggedaboutit.
USING LWNEURALNET RECURRENTLY
So with lwneuralnet, you call straight net_compute's for every step in time. Use a master network & unique inputs for every step. U need to store all the neuron outputs & network inputs for every net_compute step. Network inputs are the first layer outputs.
Then U call net_compute_output_error exactly once on the last iteration. This sets the errors on the last outputs.
Now net_train_batch to begin the training pass. This must be modified. backward_pass must start on the neuron outputs from the last time step & repeat working backward in time. Modify backward_pass to recall the neuron outputs for each time step, copy the first layer's errors in the current time step to the last layer's errors in the previous time step. backward_pass needs to store errors for all time steps instead of accumulating.
adjust_deltas_batch needs to go through all time steps & recall the errors & outputs for each time step. The deltas R properly accumulated in the master network.
Next, call adjust_weights_batch to train the master network. This averages the deltas over the number of time steps.
NOTE ABOUT LEAD COMPENSATION
The velocity term in your PID equations provides rate damping but not lead compensation. To get lead compensation, U need to add velocity to the position. The fully damped & lead compensated equation has a velocity feedback & position + velocity feedback.
Tried to figure out why we didn't get a single trace of interest in 1.5 months of real world heroine worshiping after such an easy introduction with Russian Heroine. It's a bit of a mystery how they figure out from just a couple sentences to put us out of their life. We have 2 ideas.
We have successfully gotten 1 date out of an American in our lifetime. Just 1. The Americans were all psychotic.
The Bay Area is the most difficult environment for men our age. Heroines can take their pick.
Apparently Air Force Heroine got our 2 emails & had a malfunction because She removed Her meetup.com membership immediately after #2. Definitely a screwball right up there with Ego Monger. Actually, nothing tops Ego Monger.
How do U go from Facebook friend to Facebook rejection & then meetup.com removal? That accurately summarizes our experience with Americans.
We could chase after military heroines forever. The limitation is we enjoy eating, not crying, & sleeping. U can't bank on princess charming any more than heroines can bank on prince charming. That's why men avoid heroine warriors & stick with teachers. We learn it so well, some of us get mean about it.
Maybe we bothered with the psychotics because they looked like us, talked like us & had similar experiences. Maybe we're psychotic. Time to avoid them completely if you're sticking with heroine worshiping.
How bout some flying to forget about the Air Force.
Next, it's more aerial views of U. Know. Where. We've stepped up the autopilot climb & descent rates. They're terrifying but more efficient.
Well, 3 people removed their meetup membership & we still don't believe in Air Force heroines giving up their freedom on account of outside forces, so that supports the Bay Area being extremely competitive over a personal vendetta.
We have successfully gotten 1 date out of an American in our lifetime. Just 1. The Americans were all psychotic.
The Bay Area is the most difficult environment for men our age. Heroines can take their pick.
Apparently Air Force Heroine got our 2 emails & had a malfunction because She removed Her meetup.com membership immediately after #2. Definitely a screwball right up there with Ego Monger. Actually, nothing tops Ego Monger.
How do U go from Facebook friend to Facebook rejection & then meetup.com removal? That accurately summarizes our experience with Americans.
We could chase after military heroines forever. The limitation is we enjoy eating, not crying, & sleeping. U can't bank on princess charming any more than heroines can bank on prince charming. That's why men avoid heroine warriors & stick with teachers. We learn it so well, some of us get mean about it.
Maybe we bothered with the psychotics because they looked like us, talked like us & had similar experiences. Maybe we're psychotic. Time to avoid them completely if you're sticking with heroine worshiping.
How bout some flying to forget about the Air Force.
| Above Pleasant Hill (3 min 30 sec) |
Next, it's more aerial views of U. Know. Where. We've stepped up the autopilot climb & descent rates. They're terrifying but more efficient.
Well, 3 people removed their meetup membership & we still don't believe in Air Force heroines giving up their freedom on account of outside forces, so that supports the Bay Area being extremely competitive over a personal vendetta.
It took a few months & a lot of chips, but the Dobsonian finally worked. Finally grounded the circuit board to the structure & that seems to have worked. The PCM modulation eliminated all glitches.
The movement isn't as precise as hoped. The only benefit we're probably getting is some stiffness to the movement when U want to manually move it. Can take 3 shots of the moon & track it with the RC controller in the time of another 3 shots.
There's no shaking at all.
1 other thing. Here's the movie of yesterday's still photos. 1 flight was stable enough to get a movie out of it. Unfortunately, it wasn't the 3deg/sec one.
The movement isn't as precise as hoped. The only benefit we're probably getting is some stiffness to the movement when U want to manually move it. Can take 3 shots of the moon & track it with the RC controller in the time of another 3 shots.
There's no shaking at all.
1 other thing. Here's the movie of yesterday's still photos. 1 flight was stable enough to get a movie out of it. Unfortunately, it wasn't the 3deg/sec one.
| Hi res still photo movie (0 min 39 sec) |
The answer is yes. There is an exposed shutter signal on the A480's main board U can solder very carefully. With that out of the way, our first high resolution photos from space in over a year came down.
Programmed 2 panoramas at 30m. 6deg/sec & 3deg/sec. Still programming & praying to get the right shot. It's tempting to get one of these http://www.sparkfun.com/commerce/pro...oducts_id=9334 UART cameras. These could squeeze a video downlink on the 2.4Ghz band with no risk of jamming GPS like 900Mhz.
On tomorrow's manifest, it's more photography of ... U Know. Where.
Programmed 2 panoramas at 30m. 6deg/sec & 3deg/sec. Still programming & praying to get the right shot. It's tempting to get one of these http://www.sparkfun.com/commerce/pro...oducts_id=9334 UART cameras. These could squeeze a video downlink on the 2.4Ghz band with no risk of jamming GPS like 900Mhz.
On tomorrow's manifest, it's more photography of ... U Know. Where.
Thinking about a scale model EDF tri rotor, mainly for the artistic value. We modeled Heroine 2200 after buildings, so artistic goals aren't new. The only other benefits could be:
EDF's vibrate less than propellers & could get more stable video.
EDF's won't have dissymetry of lift & roll oscillation in forward flight.
EDF's may be more stable because of the smaller diameter.
An enclosed fuselage would be weather proof & transportation proof.
Because of the lack of space, the scale tri rotor would cannibalize Vika 2's superstructure & become a sonar platform/GPS + pressure. Around that, we would bolt on a fuselage & EDF's.
Jack Crossfire congress won't fund without promoting heroine pilots.
Other features:
Retractable landing gear.
Compound curve fuselage.
Same CG as current tri rotors with small front area for camera.
Tilting camera pod but no panning.
- tri rotors look cooler when the entire fuselage pans
How do U attach an EDF to the end of a CF rod?
- dual CF rods
How do U vector an EDF on the end of a CF rod?
- probably with a CF rod triangle that rotates around a central rod.
Probably solid foam fuselage with square cutouts for electronics.
EDF's vibrate less than propellers & could get more stable video.
EDF's won't have dissymetry of lift & roll oscillation in forward flight.
EDF's may be more stable because of the smaller diameter.
An enclosed fuselage would be weather proof & transportation proof.
Because of the lack of space, the scale tri rotor would cannibalize Vika 2's superstructure & become a sonar platform/GPS + pressure. Around that, we would bolt on a fuselage & EDF's.
Jack Crossfire congress won't fund without promoting heroine pilots.
Other features:
Retractable landing gear.
Compound curve fuselage.
Same CG as current tri rotors with small front area for camera.
Tilting camera pod but no panning.
- tri rotors look cooler when the entire fuselage pans
How do U attach an EDF to the end of a CF rod?
- dual CF rods
How do U vector an EDF on the end of a CF rod?
- probably with a CF rod triangle that rotates around a central rod.
Probably solid foam fuselage with square cutouts for electronics.
Marcy 1 misery
The Marcy 1 board was no better at 915Mhz than the Vika 1 board. It turns out the MRF49XA is a packet based radio. U don't write 1 byte to it at a time. U have to write a preamble, start code & an entire packet in 1 SPI transaction. The receiver detects the start codes & through a very complicated procedure, the microcontroller reads packets out. Each step has many layers of options.
You need to do some real nasty logic level manipulations outside of SPI to get it to clock the data. It's a nightmare chip to work with, requires too many connections & we recommend the more expensive Atmel part.
Dead dobsonian
The dobsonian mount didn't go so well either. Replaced 3 chips & 3 chips burned out. That was basically the inventory that would have given us LCROSS. Now we're looking at a manually guided mount.
A telescope mount is an incredibly hostile environment for electronics. You've got high powered H bridges, humidity, large moving surfaces, large static potentials. Unless your electronics R in a hermetically sealed aluminum container & every part is grounded like crazy, you're going nowhere.
Canon crisis
So we're leaning towards a balsa + spring shutter button jig for the Canon. It's the only way & it's still better than that servo horn jig everyone uses.
The Marcy 1 board was no better at 915Mhz than the Vika 1 board. It turns out the MRF49XA is a packet based radio. U don't write 1 byte to it at a time. U have to write a preamble, start code & an entire packet in 1 SPI transaction. The receiver detects the start codes & through a very complicated procedure, the microcontroller reads packets out. Each step has many layers of options.
You need to do some real nasty logic level manipulations outside of SPI to get it to clock the data. It's a nightmare chip to work with, requires too many connections & we recommend the more expensive Atmel part.
Dead dobsonian
The dobsonian mount didn't go so well either. Replaced 3 chips & 3 chips burned out. That was basically the inventory that would have given us LCROSS. Now we're looking at a manually guided mount.
A telescope mount is an incredibly hostile environment for electronics. You've got high powered H bridges, humidity, large moving surfaces, large static potentials. Unless your electronics R in a hermetically sealed aluminum container & every part is grounded like crazy, you're going nowhere.
Canon crisis
So we're leaning towards a balsa + spring shutter button jig for the Canon. It's the only way & it's still better than that servo horn jig everyone uses.
Reversing the Titanic
PCM conversion for the Sanwa works. Sadly, it's throttle stick broke permanently & we're down to 1 stick. Our next task is to optimize the bandwidth & test it in the field.
Reversing the Titanic
We have directed the Jack Crossfire congress to slowly abandon heroine worshiping & seek traditional gender roles, over a certain amount of time, some point in the future. There will be no more Air Force heroine worshiping, eventually. Embracing patriarchy will be a long, slow transition, but hopefully less painful overall than the alternative.
U think U could do it? Domesticate another human f@#$cking being? Turn a human being into one of those lifeless housewives who depend on men for survival & have no other means of sustaining themselves for thousands of miles?
Air Force got us taking a hard look at our gender views for the first time ever & less at manipulating the world. Over the years, our investment in heroinewarrior.com, heroine UAV's, heroine this & heroine that, & improvements in the technology for manipulating information probably brainwashed us.
You've been fed an information banquet courtesy of Cinelerra, Goo Tube, Google, & everyone who stands to make money by showing U whatever U want to see.
1 month on heroine worshiping assignment & U know what? It's just like the 1950's. It's real conservative out there. Most heroines don't work.
Forget about what anyone says in public. Heroines say they want to be jet pilots & men say they want to marry housewives. All humans learn the fake sideshow to say the opposite of what they want by age 13 because it's required to get dates.
Statistically, it's a no brainer. The human population's current growth rate couldn't exist if 99.99% of the heroines your age weren't focused on domestication & raising children. U need to choose between being morally right or emotionally supported, because U can't have both. Tough questions. Tough choices.
Forget about falling in love & just focus on emotional support. Providing & domesticating sux, U can't look in the mirror, but U need to force yourself to do it like everyone else. It's a conscious effort all those other men in your daily traffic jam are constantly puking, regurgitating & forcing themselves to do every day. They figured out emotional support beats heroine worshiping decades ago.
Reversing the Titanic
We have directed the Jack Crossfire congress to slowly abandon heroine worshiping & seek traditional gender roles, over a certain amount of time, some point in the future. There will be no more Air Force heroine worshiping, eventually. Embracing patriarchy will be a long, slow transition, but hopefully less painful overall than the alternative.
U think U could do it? Domesticate another human f@#$cking being? Turn a human being into one of those lifeless housewives who depend on men for survival & have no other means of sustaining themselves for thousands of miles?
Air Force got us taking a hard look at our gender views for the first time ever & less at manipulating the world. Over the years, our investment in heroinewarrior.com, heroine UAV's, heroine this & heroine that, & improvements in the technology for manipulating information probably brainwashed us.
You've been fed an information banquet courtesy of Cinelerra, Goo Tube, Google, & everyone who stands to make money by showing U whatever U want to see.
1 month on heroine worshiping assignment & U know what? It's just like the 1950's. It's real conservative out there. Most heroines don't work.
Forget about what anyone says in public. Heroines say they want to be jet pilots & men say they want to marry housewives. All humans learn the fake sideshow to say the opposite of what they want by age 13 because it's required to get dates.
Statistically, it's a no brainer. The human population's current growth rate couldn't exist if 99.99% of the heroines your age weren't focused on domestication & raising children. U need to choose between being morally right or emotionally supported, because U can't have both. Tough questions. Tough choices.
Forget about falling in love & just focus on emotional support. Providing & domesticating sux, U can't look in the mirror, but U need to force yourself to do it like everyone else. It's a conscious effort all those other men in your daily traffic jam are constantly puking, regurgitating & forcing themselves to do every day. They figured out emotional support beats heroine worshiping decades ago.
So there isn't going to be CHDK support for the Canon A480. Your best option is to somehow port just what you need from an older CHDK to it or solder the shutter button.
Been flying & flying & flying to evaluate Marcy Foam. In the first head to head test of Marcy Foam on the barometer in the same conditions, we definitely had fewer collective glitches with the foam & the altimeter calibration was much better with the foam. Unfortunately, the graph still doesn't show any obvious difference.
Also seeing evidence of improved attitude hold in the night time video. It's not the foam of your dreams & the start of a new life, but it's an improvement.
It was a long fight, but flying on the golf course is bearable again. It's not the picnic it was before the Air Force incident, but we're not crying out there anymore & we're walking almost as fast as we once did.
Also seeing evidence of improved attitude hold in the night time video. It's not the foam of your dreams & the start of a new life, but it's an improvement.
It was a long fight, but flying on the golf course is bearable again. It's not the picnic it was before the Air Force incident, but we're not crying out there anymore & we're walking almost as fast as we once did.
The truth about the Air Force
According to The Goog, the Air Force could know more about us than we do. She could quite easily find out all about the Jack Crossfire blog, all our Goo Tube videos, our gender views, every day job we ever had, every love affair we had, Cinelerra, Heroine 2200, Trapezoidian mount, VicaCopter, Marcy Foam, spin copter, everything. She doesn't need friend status to read our Facebook profile.
Air Force Heroine is a friggin' NRO manager & she works in the National Reconnaissance Office Mission Control Center "blue cube" in Sunnyvale. So not only is she still active in the Air Force, she is VERY active.
She probably didn't use national assets on us, but probably saw a quick computer model predicting everything that we're going to do for the rest of our life & decided we're an idiot. She'll probably move to Vandenburger in 2011 when Onizuka Air Force Station is closed, or is it really closing?
The NRO can predict if you'll ever get emotional support again & from whom, where you're going to be in a year, how long U have to live, how you're going to die, when your next crash is going to be.
Maybe our entire lives R being controlled by an NRO computer model from 1980. The NRO predicted U would have backwards gender views, decided U had to go, arranged to hire Air Force Heroine, predicted U would fall in love... Scary thought.
Air Force Heroine is a friggin' NRO manager & she works in the National Reconnaissance Office Mission Control Center "blue cube" in Sunnyvale. So not only is she still active in the Air Force, she is VERY active.
She probably didn't use national assets on us, but probably saw a quick computer model predicting everything that we're going to do for the rest of our life & decided we're an idiot. She'll probably move to Vandenburger in 2011 when Onizuka Air Force Station is closed, or is it really closing?
The NRO can predict if you'll ever get emotional support again & from whom, where you're going to be in a year, how long U have to live, how you're going to die, when your next crash is going to be.
Maybe our entire lives R being controlled by an NRO computer model from 1980. The NRO predicted U would have backwards gender views, decided U had to go, arranged to hire Air Force Heroine, predicted U would fall in love... Scary thought.
| Video with Marcy Foam treated IMU (2 min 56 sec) |
These shots were with the AHRS at 1/8 bandwidth, rate damping at 1/4 bandwidth, Marcy Foam IMU treatment. We have interleaved raw footage with stabilized footage. The image stabilizer is definitely getting more mileage.
Got the Canon A480 from Best Buy, loaded it with AA's, bolted it on, & flew. Not the most efficient imaging system, but nothing beats Canon point & shoots. It's a bit better at battery usage than the A560. Not as much zoom as the A560.
The 3DOF IMU has a problem where the resets cause sudden jerks in attitude when the PID controller is saturated. Probably need continuous fading instead of sudden resets.
Stepping up GPS derived heading & getting better results. The algorithm seems sound.
Where was today's video shot? W.H.E.R.E. D.O. U. T.H.I.N.K.
Whew. U wouldn't think after the $130 Canon A560 in 2008, anyone could screw up $100 cameras, but the Kodak C160 is a 9/11 terrorist attack on cameras. U can get better quality from pinhole cameras. The lens has a light leak so if it points perpendicular to the sun, it gives a red blob. It shoots 640x240 video & upscales it.
Stick with what U know is the best & get Canon. Don't be surprised when people straggle in from camera hell asking U what super expensive pro camera got such perfect shots from your UAV.
Well, got some video of the Marcy Foam IMU & Marcy Foam barometer anyways. Had some propeller damage & that definitely shook it up. Got some pretty nasty heading drift. Barometer with a foam standoff is working.
A 47000uF supercap on the uBlox seems to have fixed its random glitches for the moment.
Video on the Kodak C160. Initially there was no low pass filtering. As this campaign wore on, the propellers got damaged by yet another tree & we had to decrease bandwidth to handle more vibration. Because heading drift seems to be a problem, may have separate bandwidth for heading & tilt.
Stick with what U know is the best & get Canon. Don't be surprised when people straggle in from camera hell asking U what super expensive pro camera got such perfect shots from your UAV.
Well, got some video of the Marcy Foam IMU & Marcy Foam barometer anyways. Had some propeller damage & that definitely shook it up. Got some pretty nasty heading drift. Barometer with a foam standoff is working.
A 47000uF supercap on the uBlox seems to have fixed its random glitches for the moment.
| Marcy Foam tests (3 min 24 sec) |
Video on the Kodak C160. Initially there was no low pass filtering. As this campaign wore on, the propellers got damaged by yet another tree & we had to decrease bandwidth to handle more vibration. Because heading drift seems to be a problem, may have separate bandwidth for heading & tilt.
For an organization that hates us, the Air Force has been giving us a lot of satellites. Got 16, the most ever. Unfortunately, the uBlox has a slowly evolving problem. Travel at high speed for a long distance, stop & it often loses acquisition. It takes several autopilot restarts before it regains acquisition.
Good news: got 16 minutes out of the 4Ah Tenergy, the longest flight since the Corona.
THE TRUTH ABOUT BARS
Went into a bar for the first time ever. Compared to bars 30 years ago, modern bars R like Mormon stake centers. No smoking because smoking is illegal. Hardly any drinking because the slightest drunk driving is illegal.
Still a tough place. Solid muscle goons guarding the doors. Over dressed, 7' tall men scanning their abused wives every move. Everyone yelling as loud as possible. Housewife wannabes & wife beating men. All of it just 1 block from the dumpy apartment. Tough places, bars.
BARS & PILOTS
Realized we were way out of our league with air force UAV pilots. We're not even in the same universe of toughness. Good idea: designing UAV's. Bad idea: trying to get emotional support from UAV pilots. They're into men like Ryan Alexander Jenkins, Michael Ratley, Scott Peterson, Hans Reiser, not flat broke low paid programmers.
Good news: got 16 minutes out of the 4Ah Tenergy, the longest flight since the Corona.
THE TRUTH ABOUT BARS
Went into a bar for the first time ever. Compared to bars 30 years ago, modern bars R like Mormon stake centers. No smoking because smoking is illegal. Hardly any drinking because the slightest drunk driving is illegal.
Still a tough place. Solid muscle goons guarding the doors. Over dressed, 7' tall men scanning their abused wives every move. Everyone yelling as loud as possible. Housewife wannabes & wife beating men. All of it just 1 block from the dumpy apartment. Tough places, bars.
BARS & PILOTS
Realized we were way out of our league with air force UAV pilots. We're not even in the same universe of toughness. Good idea: designing UAV's. Bad idea: trying to get emotional support from UAV pilots. They're into men like Ryan Alexander Jenkins, Michael Ratley, Scott Peterson, Hans Reiser, not flat broke low paid programmers.
So the Marcy Foam on the barometer was a failure. It actually degraded performance. Altitude hold is much better today because of software improvements & barometer placement which allow very high bandwidth. There may still be improvement if the foam doesn't touch the barometer. Contact between barometer & any material kills any readings.
Where Marcy Foam blew us away was the IMU. Got 2x more bandwidth on the rate damping & 16x more bandwidth on the AHRS. At 12% bandwidth, the AHRS is no longer filtered & just integrating raw gyro voltage. The amount of drift we're seeing doesn't seem any worse. Attitude hold is incredibly solid even with the ground station latency. It's more like Henrik Copter than Aurora Flight Disasters.
Batteries
Retail batteries like Tenergy, Thunder Power, & Hyperion have performed much better than generic batteries like Maxamps & XCite. The generic discount batteries R all trash. They all puffed & died prematurely. Nowadays with killer inflation & states taxing most online purchases, the retails R no more expensive than the generics.
Got a 3S 4Ah Tenergy from Hobbytown USA for $90 + 10% sales tax. Maxamps charges $135 for trash guaranteed to puff.
Can U believe we once thought Europeans were crazy for paying double digit VAT? Now look who's paying double digit sales tax for a government that works in IOU's. You humans are a disgrace to slavery.
Where Marcy Foam blew us away was the IMU. Got 2x more bandwidth on the rate damping & 16x more bandwidth on the AHRS. At 12% bandwidth, the AHRS is no longer filtered & just integrating raw gyro voltage. The amount of drift we're seeing doesn't seem any worse. Attitude hold is incredibly solid even with the ground station latency. It's more like Henrik Copter than Aurora Flight Disasters.
Batteries
Retail batteries like Tenergy, Thunder Power, & Hyperion have performed much better than generic batteries like Maxamps & XCite. The generic discount batteries R all trash. They all puffed & died prematurely. Nowadays with killer inflation & states taxing most online purchases, the retails R no more expensive than the generics.
Got a 3S 4Ah Tenergy from Hobbytown USA for $90 + 10% sales tax. Maxamps charges $135 for trash guaranteed to puff.
Can U believe we once thought Europeans were crazy for paying double digit VAT? Now look who's paying double digit sales tax for a government that works in IOU's. You humans are a disgrace to slavery.
So had to pull the rip cord & communicate electronically with the Air Force, an impossible task after the Facebook beating. It's much easier to shut people down electronically. Couldn't wait for a face to face encounter because of the hit on productivity so getting ourselves shut down was the only option. Probably won't look at an MQ-9 anymore without crying.
The Management always gives us the craziest, most ridiculous, impossible assignments. Why don't U come up with assignments we can actually solve instead of making us get emotional support out of drop dead gorgeous blond bombshell pilots?
The Management always gives us the craziest, most ridiculous, impossible assignments. Why don't U come up with assignments we can actually solve instead of making us get emotional support out of drop dead gorgeous blond bombshell pilots?
USB resets solved. U can unplug & plug in the ground station in flight & it should immediately resume communication. It rescans instantaneously. A definite must if you're using any USB connection.
Spin copter board is laid out. Upgrading Vika 1 to 915Mhz was not as much of a payoff as building a brand new vehicle on 915Mhz.
Vika 1 has Marcy Foam on the IMU. It's very hard to seat the IMU because the foam is not flat, is still very fragile, & doesn't retain its shape under pressure. Considering using plastic to increase the IMU's footprint. High risk of the foam coming apart in flight. Need better adhesives.
Vika 1's flight computer is dead. Couldn't revive it since the crash. Only explanation is it got damaged in the crash. The only affordable option is to move forward with 915Mhz Spin Copter & once the 915Mhz is proven, build another Vika 1 computer using 915Mhz.
915Mhz is only a guaranteed necessity in a spin copter, but 915Mhz may or may not be suitable for Vika 1. U can't invest anything in Vika 1 until the 915Mhz is tested in a system it's necessary for.
Spin copter board is laid out. Upgrading Vika 1 to 915Mhz was not as much of a payoff as building a brand new vehicle on 915Mhz.
Vika 1 has Marcy Foam on the IMU. It's very hard to seat the IMU because the foam is not flat, is still very fragile, & doesn't retain its shape under pressure. Considering using plastic to increase the IMU's footprint. High risk of the foam coming apart in flight. Need better adhesives.
Vika 1's flight computer is dead. Couldn't revive it since the crash. Only explanation is it got damaged in the crash. The only affordable option is to move forward with 915Mhz Spin Copter & once the 915Mhz is proven, build another Vika 1 computer using 915Mhz.
915Mhz is only a guaranteed necessity in a spin copter, but 915Mhz may or may not be suitable for Vika 1. U can't invest anything in Vika 1 until the 915Mhz is tested in a system it's necessary for.
The truth about Firefox
While you're waiting for Vika 1 to be rebuilt, 1 commute at a time,
Looking over Firefox's history, air force added U as a friend immediately after your request on 8/11/09 & U thought it was a bug in Facebook so U didn't take it seriously, nor do we take flying on the golf course as seriously as someone else might. How serious can parts sold at "Hobbytown USA" be?
U were an air force friend on 8/12/09. That was probably your happiest day since 1987. The profile was unavailable yet listed as a friend that afternoon. That night, the profile was available.
The morning of 8/13/09, U surely noted U were still an air force buddy as U uploaded the photos from the night's flight, then in the afternoon, the history shows panic reloads, "mean facebook", & frantic searches for new friends.
Your brain has been constantly simulating every possible scenario, trying to prevent future Facebook incidents & trying to will the air force back to life. Simplest explanation is the photos freaked air force out because of your golf course & U didn't take the hobby seriously, but most logical explanation is you're just an idiot.
Probably can't will air force back to life. U can make trash cans fly but U can't reprogram humans, but most likely solution is a face to face, non threatening encounter & not electronic communication. Still banging on that one.
The Facebook interface makes a lot of people avoid each other & makes a lot of options impossible because U have to assume the worst. Things would definitely be different if we never used Facebook & followed the Russian heroine path.
Looking over Firefox's history, air force added U as a friend immediately after your request on 8/11/09 & U thought it was a bug in Facebook so U didn't take it seriously, nor do we take flying on the golf course as seriously as someone else might. How serious can parts sold at "Hobbytown USA" be?
U were an air force friend on 8/12/09. That was probably your happiest day since 1987. The profile was unavailable yet listed as a friend that afternoon. That night, the profile was available.
The morning of 8/13/09, U surely noted U were still an air force buddy as U uploaded the photos from the night's flight, then in the afternoon, the history shows panic reloads, "mean facebook", & frantic searches for new friends.
Your brain has been constantly simulating every possible scenario, trying to prevent future Facebook incidents & trying to will the air force back to life. Simplest explanation is the photos freaked air force out because of your golf course & U didn't take the hobby seriously, but most logical explanation is you're just an idiot.
Probably can't will air force back to life. U can make trash cans fly but U can't reprogram humans, but most likely solution is a face to face, non threatening encounter & not electronic communication. Still banging on that one.
The Facebook interface makes a lot of people avoid each other & makes a lot of options impossible because U have to assume the worst. Things would definitely be different if we never used Facebook & followed the Russian heroine path.
The answer is yes. There is a company called Marcy Adhesives in Ohio. It makes adhesive foam for shipping glass & insulating car windows. It is not inspired by heroine pilots.
Obviously if U tried selling Marcy Foam U'd get busted on a trademark violation, but this is just a stupid blog. If the Marcy Adhesives company actually mass produced our foam, it got used in real combat UAV's, & we were the poster boy, that would be something. Never going to happen, but U can pointlessly dream about it just like U dream about heroine warriors.
Obviously if U tried selling Marcy Foam U'd get busted on a trademark violation, but this is just a stupid blog. If the Marcy Adhesives company actually mass produced our foam, it got used in real combat UAV's, & we were the poster boy, that would be something. Never going to happen, but U can pointlessly dream about it just like U dream about heroine warriors.
Been flying & flying & flying to test the Marcy foam, but seeing nothing conclusive.
Since no-one has ever documented cotton swabs wrapped in fabric softener, the material does serve a unique need, & it was inspired by heroine UAV pilots, we're calling it Marcy Foam. You'll live.
So the pilot shifted his laptop & caused the USB cable to disconnect & reconnect. Linux changed device addresses, ground station lost contact, & loss of signal shutdown from a high altitude followed. Obviously we need a way to handle USB reconnections.
Not sure the Marcy Foam makes any difference, but the altimeter bandwidth got stepped up to 50% & it's very stable on the golf course. Another new goal for Vika 1 is extending flight time by ganging batteries, so damaged batteries need to be replaced.
Since no-one has ever documented cotton swabs wrapped in fabric softener, the material does serve a unique need, & it was inspired by heroine UAV pilots, we're calling it Marcy Foam. You'll live.
So the pilot shifted his laptop & caused the USB cable to disconnect & reconnect. Linux changed device addresses, ground station lost contact, & loss of signal shutdown from a high altitude followed. Obviously we need a way to handle USB reconnections.
Not sure the Marcy Foam makes any difference, but the altimeter bandwidth got stepped up to 50% & it's very stable on the golf course. Another new goal for Vika 1 is extending flight time by ganging batteries, so damaged batteries need to be replaced.






