Date: 10/1/10 [[BR]]
Location: Tri Valley [[BR]]
Aircraft: Odin & Loki [[BR]]
Pilot: Garrison Hoe [[BR]]
Flights: 5 data flights (2 w/out data); 3 Odin, 2 Loki [[BR]]

[source:/trunk/FlightData/2010_10_01 Link to Flight Data]

Garrison and I went flying this morning. We headed to Jensen field, but they are in the process of taking down the hangars there, so we went over to Tri Valley. There was plenty of guys there, and they were happy to let us fly. They asked a lot of questions, but we still got a lot done. They mentioned it would cost around $4000 to remove the runway asphalt.

We took both ultra sticks out, with 3 objectives:

   1. Test new microbotics airdata system
   2. Test a fault detection filter, where the two faults injected were a yaw rate bias of 2 deg/sec and asymmetric flap deployment to 10 deg.
   3. Test the ASDX pressure sensors again, with revised pressure zero procedure.

We completed 7 flights and met all objectives; notes as follows for each flight number:

Odin Flight x: MB air data system (ADS). did not observe proper ADS operation during flight. Landed and found the power connector for the ADS was not securely engaged.

Odin Flight xx: MB ADS. Proper operation of ADS observed. Noted airspeed seemed to have low resolution, but little noise. Partway through the flight the RxMux appeared to stop working (pilot still had control when handoff switch was thrown). We landed immediately, and found the channel 5 (handoff switch) wire had worked loose from the R/C receiver, so the RxMux had no signal.

NOTE: No data from first two flights due the files being overwritten (operator error).

Odin Flight 11: FDI w/baseline FCL: did not observe fault indication on GCS. Controller handled fault ok, slight roll bias.

Odin Flight 12: FDI w/LQR FCL: did not observe fault indication on GCS. Controller handled fault ok, no roll bias this time. Noted occasional loss of GPS LOCK, hypothesis is excess heat (micronav uses ublox GPS).

Loki Flight 04: new avionics w/ASDX pressure sensors: ias/alt looked reasonable. no drift noted. noisy. GPS performed good.

Loki Flight 05: repeat of Loki Flight 04.

Odin Flight 13: repeat of Odin Flight xx, this time data was properly saved.


Actions & notes:

   1. Upon examination of the MB ADS data, both airspeed and altitude have a resolution of one meter. This seems like an error in the code. David, please examine the code to see if there is a cause for this. FIXED.
   2. The ASDX pressure sensors did not exhibit the same drift and abnormal readings observed on 9/16. Previous issues may have been related to the incorrect zeroing procedure. Noise is still significant, but otherwise these sensors worked ok. The I2C pressure sensor route may still be a good low-cost option with appropriate sensor ranges.
   3. New GPS (and micronav) had no problem getting lock. Issues on 9/16 may have been due to low cloud cover.
   4. Garrison had to adjust the rudder sensitivity due to the ground handling on asphalt. He also noticed the aircraft handled different with the additional weight (sensor + another battery) of the air data system. This setup is only temporary.
   5. The paved runway at Tri Valley is very nice.They also have plenty of picnic tables and workbenches, which really makes it easier to work with the airplanes.

GPS Notes from data analysis:
Odin flight 11 3 (mnav): GPS nolock:2, .32, 2.28 secs [[BR]]
Odin flight 12 4 (mnav): GPS nolock:36, mean 2.68 sec; in flight: 2, .3, .28 sec [[BR]]
Loki flight 04 5 (newav): GPS checksum error: 20, .98 sec each. (no loss of lock) [[BR]]
Loki flight 05 6 (newav): GPS checksum error: 18, .98 sec each. (no loss of lock) [[BR]]
Odin flight 13 7 (mnav): GPS nolock:3, .34, .3, .3 secs [[BR]]

Hypothesis is that mnav (ublox) GPS is susceptible to heat. New GPS (EM406A) had no loss of lock, but did have many checksum errors. The length of the checksum errors was always .98 seconds (note GPS read at 1Hz). The time between each checksum error was often 6 seconds exactly, and was ALWAYS an integer multiple of 6. This indicates something systemic is happening, rather than a random error like a loose wire.

Long range weather says more warm sunny days next week!  We'll try to go fly again next week.

Back to FlightReports