I ‘needed’ to experience a tricopter, and I chose a Hobbyking X900 frame kit.
The 1245SF props arrived and replacement HK922 servo, both were fitted.
The tricopter is very hard to fly with 1245SF despite lots of time tuning pitch and roll, and level PID parameters. It appears that the propellers are just too much and that control loop stability is hard to achieve at lower motor speed.
So, the 1145 propellers were refitted and fine tuning of the saved PID parameters commenced.
The replacement servo works fine, stable stand alone with and without load, and mechanically smooth through its entire range (no sign of any stiff spots in the gears as in the previous one).
The project is coming to an end, the tail mechanism is an obvious vulnerability in this implementation though it seems to work quite well… more in the final report.
Tests on the tricopter using a data logger to capture motor speed, current and pack voltage gives an insight into platform operation.
Above, data capture from the tricopter on 3S. The datalogger adds 50g mass.
The plot shows the device hovering at about 5000rpm and 10.8A at 11V. Two WOT ascents were made to check acceleration and motor recovery during closed throttle descents. From idle, it takes about 500ms to accelerate to hover rpm. This test was repeated and the motor reliably restarted every time and the tumbling craft stabilised quickly. A set of 12×4.7SF props has been ordered to try at 3S to more fully load the motors.
Similar tests were conducted also on 4S, and peak current on WOT was 80A on 15.5V for 8500rpm. Operation on 4S would provide sufficient lift to carry a camera payload, but WOT operation exceeds the motor’s continuous current rating.
The tricopter project lumbers on, delayed by faulty parts.
The recommended tail servo, a Hobbyking HK 922 Metal Gear Digital Servo, failed. It had a rough spot at about 70% of its range from new and which became worse quite quickly until the motor had insufficient torque to overcome it.
There is a risk of damage when flashing ESCs. It accrues from the fact that ESCs have a three legged H bridge and if a high and low FET are turned on simultaneously, damaging currents may flow. In fact, this can be an issue if the FETs are on together for just microseconds on each PWM cycle. Loading the wrong hex module is a recipe for disaster, it may turn on FETs in an unexpected way.
My first tricopter build is based on a frame from Hobbyking. they say:
Designed from the ground-up by our own engineering team, the X900 Tricopter is a culmination of months of design, testing and material sourcing to provide you with the perfect mix of quality, performance and value.
Oh well, with hype like that, the reality can only fall short!
The size is 820mm diagonal between motor shafts (not 900mm as the type suggests).
Key elements of the configuration are:
- Hobbyking X900 tricopter frame;
- Turnigy D3530-14 1000kv Brushless Motor;
- Hobbywing Skywalker 4S 40A ESC, loaded with BLHeli v11.0;
- HK 11×4.5 SF two blade propellers;
- Hextronic MultiWii 328P Flight Controller w/FTDI & DSM2 Port;
- FRSKY V8FR-II HV receiver;
- Battery monitor;
- Turnigy Discovery Beeper
- Zippy 3000mAh 3S 30C LiPo battery;
- Turnigy 9XR/OpenTx transmitter.
The frame has had its problems.
The springs used for the “shock absorbing landing legs” are low grade and straighten out when stretched by the legs, see above. Fundamentally, the leg design is flawed, the spring fouls the recommended tail servo and it fouls the mounting screws at the end of the tricopter arms.
The replaced plated steel springs were placed under the inboard arm mounting nut and washer and onto one of the pins in the leg assembly, see above. They don’t found anything in this configuration. The shock absorbing leg folds up so that the servo bracket etc take the bump when the tricopter lands… the most delicate part of the whole craft is unprotected. A temporary measure is some foam zip tied to the legs, but the whole shock absorbing leg is proving to be a bit of a worthless gimmick.
The kit lacked appropriate screws to fix the servo to its bracket and the servo crank to the ball link crank, in my case some 2mm hex head screws and nuts were used.
The one part of the frame that seems well done is the tail servo bearing and motor support. The system is free of backlash, and control is stable.
The ESCs were unwrapped, cables and JST-1.0mm attached to the C2 pads as permanent programming cables and longer motor wires fitted, re-wrapped and loaded with BLHeli Multi v11.0. The ESC was given a bench test on some challenging motors at 4S, and it was very responsive with no hint of sync problems.
The FC was loaded with Mutliwii 2.3 configured for a tricopter.
Initial flights have been good, the craft has plenty of power on a flat battery, is quite responsive for such long arms, and quite stable though tuning work continues. Expectation is that a 4S battery will allow carrying camera payload should that transpire.
More when it is tuned!
The above source was built with COMP_PWM for an Afro30 ESC and bench tested on three motors that have had sync issues in some previous FW versions:
- Hobbyking DT-750;
- Turnigy 4822-690kV; and
- Turnigy 2730-1500.
On 4S and using a servo tester, the motors were tested for sync loss as indicated by the red LED on the Afro30.No attempts were made to tune timing, default advance was used.
|AWG||Cond dia (mm)||CSA (mm^2)||R/km (Ω)||Imax|
Derived in part from data published by Wiremax.