Archive for April, 2007
Comments (1)
Add Comment
-
Views: 482
Meet RUSSIAN HEROINE II. -
Views: 380
Since we don't have public display permission, the only views of R.H.II you get are these nuggets of heroine beauty. -
Views: 329
Rare view of a clean airfoil. -
Views: 321
Rare view of a clean airfoil. -
Views: 344
New battery strap to keep it from shifting sideways.
-
Views: 290
R.H.II demo flight -
Views: 277
-
Views: 280
Went back to the hanging antenna and radio quality was greatly improved. The hanging antenna swings around a bit and has a shorter lifetime, but it's the only way. -
Views: 275
-
Views: 276
Interestingly, when she crashed and the motor stalled, the battery voltage dropped way down, causing the algorithm to shut it off completely. -
Views: 363
Takeoff & landing occurred near 2nd base today. -
Views: 511
Flat broke, unemployed programmer, recharged with emotional support from R.H.II.. -
Views: 403
-
Views: 397
1st successful battery save. With the 3.3Ah and an agressive battery safing algorithm, she lost altitude pretty fast, but it beat a motor shutdown. -
Views: 270
2nd successful battery save. This was more a subtle algorithm than the 3.3Ah algorithm and it clearly dipped below the minimum voltage. The 4Ah clearly has less current capacity. -
Views: 279
Radio contact was just good enough to stay in the air, but not much else. -
Views: 310
Now the mane event. Glitching in the autopilot enable switch has truncated this recording of attitude. -
Views: 309
With glitching defeated, we have a complete 17 min attitude plot. -
Views: 278
Clearly we still have issues with position detection except for altitude. -
Views: 286
3rd successful battery save. Aggressive algorithm worked perfectly and gave another smooooth Launchpad McQuack landing.
-
Views: 432
The first, reasonably intelligable Kalman filter output showing tilt and heading output in degrees. Obviously a certain tilt is sensed for stationary hovering. It looks like vibration is being reduced. -
Views: 301
A demonstration showing agreement between the $10 lensatic compass and the $600 Vicacopter compass. -
Views: 317
If they looked misaligned, it was because the mag was misaligned.
-
Views: 390
The first attempts at protecting components from crashes. -
Views: 294
It's quieter with this on. A real fuselage and perfect gears and she might reach whisper mode. -
Views: 275
The first plot of battery telemetry with the 3.3Ah. At 58000 it should start reducing maximum throttle such that it stays above 58000. The maximum throttle setting should only reset after a power cycle. -
Views: 451
This flight had relatively few radio dropouts. -
Views: 281
With the 4Ah it crashed at 60000, well above the empirical minimum of 58000. The 4Ah also had a serious glitch, maybe from a landing. The crash abatement algorithm never had a chance. -
Views: 269
This flight had massive radio dropouts.
-
Views: 413
A rare view of how the magnetometer is stuffed in custom shaped foam to handle its Z axis board. -
Views: 300
And our worthless attempt to rotate the mag so it would cut through the Earth's magnetic flux at a better angle. -
Views: 316
Capacitor was damaged in the last crash. Crash from a dead battery. The main spur gear is compressing down on the board in these crashes. -
Views: 326
The capacitor could not be recovered, so it's time to do what we do with all dead capacitors. -
Views: 333
Inside, you'll find a tightly wrapped spool of tissue paper soaked in aerosol smelling fluid. -
Views: 328
Unwrap the spool and you get a very long sheet of aluminum. -
Views: 288
Two long sheets of aluminum separated by an insulator, soaked in electrolyte. -
Views: 311
Finally the heart of the capacitor, the aluminum sheets. Hard to believe they fit in such a small container.
-
Views: 425
Actual orientations where magnetic north passes up through the rotor shaft. Mag X & Y are 0 and mag Z is full negative. -
Views: 331
Did 4 possible orientations to show it isn't hard iron but the country that's the problem. -
Views: 300
There's no hard iron in the room. You can put a lensatic compass anywhere and get the same orientation. -
Views: 320
Lensatic compasses have weighted magnets to counteract the force of U. Know. Where.. -
Views: 333
Another section of dumpy apartment to prove there's no hard iron. -
Views: 294
-
Views: 394
The first intelligable gyro plot from the state propagation matrix. -
Views: 264
Unfiltered gyro and unfiltered accel data superimposed. Clearly they're agreeing. -
Views: 303
With the rainy season back, it's time to waterproof Vicacopter. -
Views: 286
-
Views: 271
-
Views: 279
-
Views: 255
Someone forgot to upload new decals 4 U to admire. -
Views: 266
Rock formations in Arizona.
-
Views: 362
Another radio packet rate plot. Much better results this time. Had the antenna outside all the other wires. Maybe time for a Dean whip antenna. -
Views: 273
Processing a flight recording on the Opteron, the Kalman now shows more activity, but not much correlation. -
Views: 286
Processing the Kalman on the Gumstix with Vicacopter sitting on the ground, it only does 2Hz. -
Views: 283
Moving yaw, roll, and pitch by hand, we have crosstalk between roll & pitch and nothing for yaw still, but it's progress. -
Views: 317
Leaving her on constantly, with the serial console on, seems to create a "virtual R.H.II presence". -
Views: 320
"Virtual R.H.II presence"
-
Views: 301
Underside logo looks a bit overdone. -
Views: 301
Determined look -
Views: 256
Another underside -
Views: 250
Got tired of looking at bare wood. That's why you're staring at an underside decal. -
Views: 276
Maybe a smaller logo like what we have on the video station, but it won't be visible at high altitude. -
Views: 278
-
Views: 276
Debugging position for the PIC erase failures. -
Views: 288
MCLR now has a 2s delay, but it still has flash losses with the bootloader. Erasing the bootloader seems to improve matters.
-
Views: 363
The 4/12 Hillary attack. -
Views: 295
She's so mad, she can talk with her mouth closed. -
Views: 249
Section of the Gyro data dealing with pirouettes. They're nice and tight now. -
Views: 247
Mag for the entire campaign. Got definite headings but Z axis is pretty worthless. -
Views: 231
Section of accel data concerned with translation. -
Views: 264
Radio packets/second. May have to mount the GWS on a backing.
-
Views: 381
The angle that most looks like Her. -
Views: 281
Another angle that looks like Her. -
Views: 250
This angle doesn't look like Her. -
Views: 266
R. A. at night. -
Views: 246
R. A. in daylight. -
Views: 254
In the indoor angle, her mane looks too short. -
Views: 302
It's a shame Hillary Clinton is going to destroy all this. -
Views: 297
Since the DVD is already loaded, have some lions from the same trip.
-
Views: 395
Taking more static abatement measures. -
Views: 297
Spur gear grinding. -
Views: 281
Header destruction. -
Views: 309
After repairs, wanted to point out that Amazon looks remarkably like Russian Heroine II. Well, everything looks like R.H.II. -
Views: 315
Downloading flight recordings with the ESC armed. Do you feel lucky? -
Views: 276
ADXL330 data is greatly improved over the LIS3. Translation maneuvers are clearly visible. -
Views: 282
Gyroscope data had to be compressed to get the accelerometer data on the same A/D converters, but it still looks useful.