Sunday, August 30, 2015

Flight Session 38 - Nova-1 Return to Flight

A little windy but a nice day to fly anyway. I ran 2 batteries through repaired Nova-1. After re-calibration, it still flies ok. Barometer/altimeter is still un-precise (floats within 1-meter span variance) so, nothing new there. For this flight, the only external payload is the HK Quanum Retracts.

Might post a short YouTube video later. I took it with my iPhone5, while holding Taranis in other hand (kinda scary for me). Nova was in Loiter mode (solid accept for altitude ... can't be helped).

Flew in most modes. Looks pretty cool flying fast and low with retracts-up. Practiced manual landings which went fine. The larger stance is nice.

The small prop-plane carrying the banner is a new thing. It circled a few times and his return leg was directly overhead field. I stayed alert to where he was and also didn't fly too high.

There was a whole flock of Scissor-Tail Fly-Catchers low patrolling and feeding from field. They were very interested in the new "large insect" on the field.

I tried a completely automatic landing but it flipped over. I knew better but thought maybe last week's rough-landing helped it somehow (knocked it into submission :-). As usual ... it lands softly, but doesn't know it. The motors are still spinning fast, blowing against ground, and it eventually tips itself over :( With flakey altimeter, only hope to prevent a prop-stike is to switch to Zero-Throttle Stabilize at last moment (as it first touches ground). Re-enforced puck held-up fine on tip-and-flip.

Wednesday, August 26, 2015

Bought another HK Quanum Nova Quadcopter

While HobbyKing had them on sale last week (at extreme discount) I went ahead and picked-up a second PnF Quanum Nova. While I originally bought it for spare parts, I'm thinking I might set-it-up in an alternate configuration, and fly it for a while (until I need the parts). I ordered from USA East Warehouse (in Arkansas) and it arrived 5 business days.

While my original Nova was one of the first off the assembly-line a year ago, this one has more updated-revision components.

Flight Controller: APM_252 v5.0 (now with pre-installed Molex Pico-Blade sockets for Telemetry, Compass, and Power-Module)
Power Distribution Board: v1.7
ESCs: 4x v1.6 (no sockets to PDB connection)
Compass: Original small one, hot-glued inside puck
GPS: BeStar BS-280 (different than first Nova, but likely still comparable to an uBlox 6-series)

Looking forward to finally having a Power-Module on an aircraft and I'm hoping the barometer on this FC is more precise than my old one.

Compass socket came pre-bent (pulled-away from board) so I hope it's ok. Is still comes with special ArduCopter v3.1.2 pre-loaded. I pulled boards apart to check I2C port and general soldering. Not great but checks with ohm-meter. I installed black-tape insulation between USB-port case and lower PCB before re-assembling. General soldering on ESCs and stuff looks ok. I will fire it up soon and test log-memory, barometer, and general functions.
The supplied silver and black prop-nuts seem to be from a different vendor than old ones. They have more of a matte finish now. One of the new black ones starts to bind if tried to screw all the way down (without a propeller installed). However, it only happens on ONE of the CCW motors (not the other). That seems to point to a slightly out-of-spec motor thread. However, if replace both black ones with older (glossy finish) spares, binding goes away. Definite varying thread tolerances on both the motors and nuts ... might be why some people have "props-flying-off" problems at first.

As a quick test, I hooked up a X8R-RX, loaded some stock params from MissionPlanner, and got it to Arm. Motors seem to spin. Sensors seem to work. Logs (ram memory) are working. BEC is outputting 5.25volts under load. All looks good so I will start my mods and build-up (using tips I have learnt on first one).

Monday, August 24, 2015

Nova-1 Reconfig and ready for "return to flight"

I bought these a while back (in my quest for taller and better landing gear). Since I had to pull the gimbal and camera off for repairs, I thought that now might be a good time to try them.

The Quanum Nova Retracts are 193 grams and 160mm tall. I control them with FrSky Channel-13 and SA-Switch on Taranis. For reference, stock white-plastic landing gear was 90 grams.

Be sure that 2 inner screws don't protrude inside and scratch battery.

For anyone installing modded landing gear (from other quads) it might help to examine how HobbyKing (for the Nova Retracts) used the box-like or U-shaped channel pieces to build a solid and flat mounting area (on the unusually shaped bolting/mounting area).

Quad Recalibrations:
As previously posted, the gyro calibration was knocked out-of-wack. I recalibrated in normal fashion and bench testing (like in Mission-Planner's HUD) reveals its working properly again (good to see gyro wasn't permanently damaged). I also recalibrated compass for good measure (-34, -208, 127).

Retracts Notes (updated 9-7-2015):
They worked fine for initial flights. Later (on bench) the left retract failed to lower ONCE. Pilots in HobbyKing reviews said it is likely end-point micro-switches. Strange because these types of micro-switches are usually rated at "thousands of clicks". I'm hoping the switch just got over-actuated and the plastic tip got slightly stuck inside the switch-case (once, maybe the last time it was "touched"). Or, maybe not pressed hard-enough once. I think we have all had a run-in with a flakey switch at one-time-or-the-other. Sometimes, more actuations loosen-up the switch and they get more reliable.

While extended, through side-slot, I applied a rice-grain sized dab of  (dielectric) Automotive Silicone Bulb Grease to the tip of the switch. I let some also get on screw-drive and internal mechanism. Lowered retracts and repeated for other end. Tested 10 times with no incorrect operation. I hope that fixes it and there are no surprises later (especially at flying field).

Sunday, August 23, 2015

Damage Evaluation and Repairs

Best I can tell, there is no other damage (no body/frame, internal, or other accessory damage). I will update this post as repairs progress. With gyro reading tilted like that, I expected to find FC dislodged, but no ... the Zeal held and FC is still secure. So, if FC ever gets heavily jolted, it's likely going to need a re-calibration.

Gimbal Repairs:

Using advice from RCGroups user SureShot_ , flattened 45 degree bend in gimbal frame. I used hammer, soft mallet, and anvil. I did not apply any heat. I was afraid it would crack during repairs, but it worked fine. Very soft metal and likely why it bent so easily in the first place. Take your time and make small adjustments until finished.

Gimbal controller looks ok. It was working a bit after crash. Turns out the bent frame was only touching the negative/ground input terminal on the back-side of the board (lucked-out there).

Saturday, August 15, 2015

Flight Session #37 - Flying to 120 meters - Poor Landing

Same field and not very cloudy, but hdop could only achieve 1.8 . Later found out it might have been geomagnetic storms (poor KP-Index).

Battery #1 - Manual flying in various modes. Practiced landings. A little FPV flying. Battery very hot after landing.

Battery #2 - Took it up to 120 meters (393 feet) in Loiter. Quad did great and seemed to hold a fixed position. Never been this high before (about 90 meters was old max). Tower/DP3 warnings sounded as I got close to 400 foot limit. At apogee, I slowly did a 360 to take-in the view.

I started my way down. At about 60 meters, LowVoltageAlarm starts to sound even though I was only about 6 minutes into flight, so I slightly accelerated my descent. At about 30 meters, the quad started to wobble. Apparently I was descending too fast and the Nova had entered a Vortex Ring State (VRS). I gave it more power but the VRS seemed to be affecting the thrust and it was falling too fast. The Tower/DP3 "Imminent Ground Collision Warning" was going off now even though I had the throttle at max now.

The Nova bounced off the ground pretty hard. The gimbal and Mobius got slammed into the ground. I could see part of the landing gear had broken off (laying on ground) but quad was still flying, gimbal was still working, and camera still recording. I flew it a few more seconds to get my bearings and landed it (remember, LVA was going off this whole time).

Edit: After flight analysis reveals I made at least 2 piloting mistakes.
1. Descending too quickly (for a quad with this payload and so little net thrust).
2. After entering VRS, I continued to descend straight down.

So far, it looks like the damage is just:
- Broken landing gear.
- Gimbal lower aluminum plate bent at 45-degree angle. Not sure how gimbal is still working and board didn't short-out.
- Quick MP check (getting logs) reveals Nova's accel/gyro calib. is now way off (severely lilted).

Not really a crash, but 2nd most damage done in an afternoon.