DDFP Tail on a Trex 500

I've been wanting to experiment with a Direct Drive Fixed Pitch tail for some time on a bigger helicopter to get rid of the fragile and wear prone tail drive, servo, linkages, etc. I got it working on my Trex 500 using a Tiger 3510-630 spinning a 9 x 4 APC prop. Have tested it in 20-25 mph wind gusts with no blowouts.

I eventually want to put this system on my Trex 600. The tail response was a bit quick at first and been tuning it to get it smooth enough for UAV type flight, flying autopilot controlled mapping/survey flights with a 600 helicopter carrying a MAPIR Survey2 camera. The flight controller used on this 500DFC is a Pixhawk running AC3.4.6. There's a couple bugs in the code for tail rotor active state vs FC armed state that I'll look into as time permits. But I now have it working quite smoothly on the 500.

This would also be quite handy for scale builds where the tail rotor is located in the top of a vertical stab, or similar. Without using the quite messy angled tail drive kits for scale builds. The weight difference between the DDFP and standard tail is only 4 grams. I got rid of the tail unit, belt, drive gear, autorotation gear, servo, linkages, tail pitch mechanism, etc.. The setup I'm using on this 500 is way overkill and will give yaw rotation rate of ~700 degrees per second in the anti-torque direction, and about 400 degrees/sec in the torque direction. The maximum available thrust of the tail drive should match the torque direction to yield about the same 400 degrees/sec yaw rotation in the anti-torque direction. But I used some stuff I had laying around the shop to build it just to test it, and get it working.

This type of tail is quite common on little helicopters like the Blade MCPx. Not done too often on 500's and up.

Here's a vid that shows how well I got it working on the 500

 
OK, so I made a few changes to the code and built new firmware with the changes in it and loaded it into the flight controller for today's project. One of the advantages of working with an open-source system. If it doesn't work right you can fix it and submit the changes to the project so they can be incorporated into the stable releases of the firmware.


The changes to the code are on my github fork:
https://github.com/ChristopherOlson/ardupilot/commit/212811f525c403c985b58dba315ac6a281693faa
 

F106DeltaDart

Elite member
Nice work, looks like it holds very well! Would be especially useful in trying to model canted tail rotors similar to the Blackhawk or CH-53.
 
It's quite a bit more efficient than the stock tail too. It takes way less power to run that brushless and a 9" prop than it did with the belt and gears spinning the tail at 10,000 rpm. The DDFP only runs 2,500-2,600 rpm in forward flight and about 3,500 rpm in hover. The heli has a 4S battery and it pulls about 17 amps in forward flight at 15-20 mph. It pulled 18-19 amps with the stock tail. Without the belt on the tail it takes almost 2 minutes just to get the main rotor to stop now after shutdown.