I've been fascinated by drones for quite some time, but have recently had the opportunity to join a university engineering team and we're looking to develop some sort of drone component. We've had a few great ideas so far, but we want to do some more research by asking actual thought leaders and seasoned hobbyists.
I want to hear what really pains you about drones. What would make your life easier? How can we build something to fix your problems? Also, I'm very open to having a Skype call!
Replies
I don't think that is correct Jake. The NamPilot is getting 10km with 500mw on 2.4Ghz
http://diydrones.com/profiles/blogs/video-and-datalink-antenna-for-...
"Tests were conducted with the ground tracking antenna at 10km from the aircraft position.
Video TX power was 27dBM @ 2.432GHz, into the Skew Planar Wheel. Datalink TX was a 868MHz xBee Pro- range was tested at 25milliwat TX power ( can be set to 1mw, 25mw, 100mw, 200mw and 300mw - 1mw at 10km did not work)
Video was perfect picture with the aircraft at any azimuth or pitch angle ( combined) except for the Aircraft directly facing the ground antenna, at 0deg pitch - the A/C antenna is shielded by the avionics, the on-board IR gimballed IR camera, and the Lipo batteries. Tilting the A/C up or down 5 degrees brought the video back with perfect picture."
He just has a tracking helical antenna on the ground. 1.2Ghz should go even further with less power as the free path losses are less for 1.2Ghz
I'll just state flat out... You cannot get 5 miles from 300mW. That's not even within the realm of reason.
No manufacturer or distance calculator anywhere even tries to claim that, and they use theoretical best-case scenarios.
Indeed, if you are not achieving just that, you most likely do not have proper antennae
But 1.2 doesnt go very well thru a few trees. Multi km range is not the problem its objects that get in the way on any freq.
I should also add that we need a proper ADSB implementation. It needs to be reasonably priced, lightweight, and not consuming a ton of power.
Nobody will design a transponder for drones since the FAA requirements don't fit well. Drones can't afford to carry that much weight and transmit at that high of power levels.
Somebody needs to implement a next-gen system and let the regulations evolve around the physical requirements of micro-sized air flight.
We should have the capability to use a ground based ADSB instead of an airborne transponder. So the ground station receives ADSB traffic to avoid other aircraft (avoid), then transmits it's own telemetry via the ground transponder OR cellular/internet (warn other craft).
Your ground station could do this with no additional weight or power consumption in the air.
It may match the function but probably won't be ADSB. The AOPA is even fighting the mandate right now because of the certification and expense for small aircraft.
That's coming Jake
So far hobbyist drone development has been mostly a hodge podge bottom up affair that operates with the premise lets take what we can scrounge from existing hobbies and off the shelf stuff and see what we can make from it.
As a result evolution has had a lot more in common with natural selection (on a similar time scale) than a properly directed top down design with specific goals in mind.
This is most obvious in multicopters which seem to have started with the basic thought lets see what happens if we take four PC power supply fans and turn them upside down.
As it turned out, the basic concept wasn't bad, but all evolution has been sort of a why don't we try this better looking part and see if it helps sort of a thing.
Sort of the intrinsic opposite of the application of scientific method.
Everything is borrowed, under designed and at least somewhat inappropriate for what would be considered optimal.
Recently the economics of scale have brought a more commercial attempt at integration so they can sell more of them, but the basic designs are still extremely flawed.
Multirotor hover condition optimized prop design would be very different than it is now if it were done properly.
It really means propeller diameter, shape, contour would be optimized for each size and weight copter and they would be a lot different than they are now.
I sometimes wonder if there is anybody involved in this entire venture who can really use a proper propeller design program.
Our motors are all slight redesigns from normal horizontal thrusting airplane motors.
Among other things, the primary load in a multicopter is straight up (axial load) and nobody is using actual thrust bearings in any of their motors.
KDE uses 3 Conrad type bearings usually including at least one large one, but Conrad's are designed primarily for radial load not axial.
The radio systems we are using are all designed to actually fly RC planes without computer controls and as such use PWM channel outputs whereas we should really be using a computer communication language and transmitters / ground stations with a lot more interesting and accessible controls.
Some of the tablet - WiFi control systems provide some help, but hugely compromise where manual control is needed (a virtual stick is not a stick.)
Maybe the New Parrot BeBops controller is a bit more in the right direction.
That said, we have forged ahead hugely using "stone knives and bear skins".
It will get better (except for the FAA of course).
Best,
Gary
-
4
-
5
-
6
-
7
-
8
of 8 Next