I've setup my helicopter (Trex 600 with self written autopilot ) so I could position it via RC then tell it to freeze in place.
Today I had fun with that out at the Friends of Amateur rocketry site.
Its been a long 20 hour day and I'm tired. I have to catch a plane in the morning so I won't get a chance to do this later.
I'm uploading the view from the helicopter.It should be on my youtube channel in about an hour.
Please note in the video the windsock in the background and the stability of the helicopter .
Comments
Not bad for a "side" project.
RTK GPS can be quite impressive when guiding a rotorcraft, but you do need good satellite visibility (especially for L1 only) so basically it only really allows you to get close to something from above or skirt rooftops, etc. L1/L2 RTK with good IMU integration is much more robust, but also more $.
I know your flight was with autonomous GPS, but if you do play around with RTK I'd suggest that vibration probably isn't the problem. Look at the signal to noise ratio for each satellite. My guess would be noise from the ESC/motor combo is probably knocking down the SNR enough to make L1 RTK very difficult. You will see this especially with power surges (high blade pitch inputs). Short satellite dropouts may not affect the CA code solution much at all, but will blow up a RTK solution.
Hi Pbreed,
I have a few samples of the ST LYPR540AH (3-axis gyro in one chip) and LSM303DLH (3-axis accel and 3-axis mag in one chip). I could send them to you for tests if you like?
Let me know by PM, I'd like to see the results!
thanks
I have 3 ADXRS450 gyros which are the vibration resistant ones (1 SOIC Z axis package, and 2 side mountable packages) if you would like them. After I bought them I realized that these however wouldn't be too good for a IMU as they have a low update rate.
These are looking excellent:
http://www.analog.com/en/mems-sensors/gyroscopes/adxrs642/products/...
I was on a plane all afternoon/evening..
Question answering will be eratic till Friday.
GPS:
I'm presently using a hemisphere crescent L1 GPS in unaided mode.
So that flight was done with no RTK or any external aiding.
I also used GPS for altitude. That helicopter has no sonar or pressure sensor.
I do use a quality GPS antenna and pay attention to RF shiielding the GPS receiver.
I have tried doing RTK with the the crescent L1 GPS using crescents as both base and rover.
I never had much luck with any L1 GPS in RTK mode in a high vibration environment.
I've also flow with a Trimble BD950 L1/L@ survey grade (read $5K) GPS.
With the 4 pound survey grade antenna strapped to a pod on the front of the helicopter L1/L2 RTK using a pair of trimbles was amazing, accuracy to a few cm at 20hz. With a smaller L1/l2 antenna mounted on the tail
(like the picture above) The accuracy was not quite as good and I got cycle slips. This may because the antenna had 24db gain and the trimbles wanted 33db.
I also tried the Trimble reciever with the Omistar HP service and the accuracy was almost as good as the full local RTK system with significantly less hassles. The trimble uses position differeence for velocity in RTK mode. Thuse cycle slips had a big effect on the velocity reporting. Since I use GPS velocity as the key input to my navigation software it makes for erratic flight. The best solution was had when flying two GPS receivers, the L2 RTK for position and the L1 for velocity. Maybe it was just with that much stuff strapped to the helicopter flight times were so short that it could not wander off ;-)
IMU
When I started this proect in late 2006 the microstrain IMU I choose was the only solution under 10K that would output a complete solution with integrated magnatometers.
If I was starting today I'd use analog devices digital gyros and the DCM or quatrino algorithms so much discussed on this site. . Its a bit pricy, but the analog devices has a nice 9DOF IMU sensor package integrated in a nice little cube. They have also recently released vibration resistant individual gyros, and I expect to see these in a vibration resistant 9 DOF cube some time in the future.
Next goals:
Redo the enectonics to be smaller and itegrated in one small package.
Test with more modern gyros. (The present IMU only outputs at 75Hz)\
Switch to a flybarless head and use the IMU to directly manage the tail rotor ratther than have
a conventional helicopter gyro in the loop.
The helicopter is a side project of my rocket work, so the rocket stuff comes first.
This means thaqt the helicopter changes described above will probably take 6 months or more.
I'll try to post updates and progress here.
One of the first steps will be to layout an IMU test board that has 3 or 4 different brands of gyro, IE the analog devices high vibration ones the analog devices cube and maybe some low cost invensense or ST gyros/accelerometers for comparison. If anyone here has a preference for a specific brand or type of gyro they would like to see tested list your favorites. I will be doing this in the next 4 weeks, and will first fly in the rocket.
Another question, What RTK base where you using, and i asume you where using the MaxStream to send the RTCM/CMR data. I have read a bit of your blog but didnt see any reference to this.
I had RTK GPS on the copter.
This flight was done with normal L1 GPS.
-
1
-
2
of 2 Next