Joined Oct 2011
Fatshark Dominator HeadTracker Hack/Mod
My intention for the fatshark dominator was to use it in an augmented reality video game/training system I'm developing at the University of Washington's Department of Rehabilitation Medicine.
I figured out the pinout for the 4-pin mini din connector on the bottom of the headset and decoded the headtracker's PPM signal using a microcontroller (Picaxe), with the intention of sending the head angles to a PC via bluetooth. The headtracker, while nice, only provides +-45 degrees of head pitch and yaw, and is quite slow on recentering (no roll angles provided). This is OK for pan tilt but not quite so good for clinical measurement. I wanted a full fledged IMU to track head attitude, so I turned to my favorite electronic retailer SparkFun and bought the "9DOF sensor stick" (http://www.sparkfun.com/products/10724), which has 3 axis accelerometer, magnetometer, and gyroscope.
Amazingly, this IMU fits inside the same compartment as the branded fatshark tracker and only needs two pins to communicate, brilliantly supplied by the 4 pin mini din. The other two pins are ground and power for the headset. The fatshark has an onboard regulated 5V supply to power the IMU, but it's got a nasty ripple so I supplied my own 3.3V regulator for the IMU power. I have a partial pinout for the header sockets inside the head tracker compartment as well.
The fatshark is connected to an additional IMU which attaches to a user's chest. The chest pack has two microcontrollers (one for each IMU), a bluetooth module, programming jacks, and power electronics. A separate beltpack has a 3S LiPo supply with low voltage monitor and breakouts for balancing/charging.
I mounted a 2.4GHz transmitter and video camera on the fatshark using a part made in a 3D printer. The transmitter is powered from the fatshark's DC barrel jack connector (had to bypass a couple diodes to do this). So, I can send video data from the head to a PC, augment it with 3D overlay, and transmit back to the fatshark @ 5.8GHz, all the while collecting attitude from the head and chest simultaneously (using quaternions to avoid gimbal lock) for interaction with the 3D overlay (i.e. game mechanics).
IMUs work fabulously after calibration. I'm using a sensor fusion algorithm from a paper titled "An efficient orientation filter for inertial and inertialmagnetic sensor arrays." And I compiled a C++ shared library from Matlab to run optimization routines to calibrate the magnetomers/accelerometers.
Here a few images of the system. The dude in the photo is not me You can see how nicely the SparkFun sensor array sits inside the compartment, it's like it was meant to be! Compartment closes fully and everything, totally seamless. The black box with the gold jack pack is the transmit/receive for the video signals from/to the computer.
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|Discussion||Question about my Fatshark goggles and headtracker||indycobra||FPV Talk||1||Apr 01, 2011 10:19 PM|
|Sold||Sold: FatShark video goggles and TrackR1 HeadTracker||Pepelatz||Aircraft - General - Miscellaneous (FS/W)||3||Nov 23, 2009 01:00 PM|
|Discussion||FS: Fatshark Goggles and TrackR1 Headtracker||Pepelatz||UAV - Unmanned Aerial Vehicles||1||Nov 22, 2009 09:56 PM|
|Discussion||FS: Fatshark Goggles and TrackR1 Headtracker||Pepelatz||FPV Talk||0||Nov 21, 2009 09:19 PM|
|Discussion||FS: Fatshark Goggles and TrackR1 Headtracker||Pepelatz||Aerial Photography||0||Nov 21, 2009 09:17 PM|