I was going to refresh my old FPV 250 racer built two years ago with an "old" Naze board and an "old" baseflight firmware. Not surprisingly, the whole firmware landscape has changed dramatically and I had to investigate quite a bit to bring me up to date with current states of affairs (lots of new firmware versions, new flight controller hardware).
As opposed to software (r)evolutions, flight controllers hardware did not evolve as much (apart from processing power increase, nothing revolutionnary). One noticeable novelty on hardware level is the relatively new open source "SP racing F3" board. Made a quick tech review of this board in the video which, as I understand it, became the state of the art flight controller for "seriously pro" FPV racers.
This video also shares some brief historical & evolution about hardware/firmware for FPV racing (MultiiWii evolution branch).
Cheers,
Hugues
Comments
Well you might want to chill down, it was me (Maxzor) who thought the comments were deleted, actually I entered clumsy mode cuz comments are on this thread http://www.rcgroups.com/forums/showthread.php?t=2540753 :)
Still need measurable results ...
I went visit briefly the rcgroups threads and saw how Rob has had his messages deleted because he tried to swim against the wave of these hysterical guys wanting always more PWM update/PID loop frequencies. Rob's arguments are sound and make me wondering where is the scientific data showing it gives a measurable result ? Did anyone ever did measure results improvements in what ? Speed ? Stability ? something else ?
I think aswell this is exiting, although, as Rob Lefebvre mentioned on the RCGroup Raceflight thread (http://www.rcgroups.com/forums/showthread.php?t=2533601) months ago, there has been no scientific measurement on actual progress made in control with the increase in various update rates. This is a major step to confirm it is an interesting project...
About Za Pf comment :
I will look further at dRonin, I come from the Cleanflight stack.
The closed-source software is KISS, it claimed to have developed a bidirectional protocol on its 32-bit ESCs.
BlHeli_S 14.6 is said to have come to the level of control of KISS ESCs. https://www.youtube.com/watch?v=lGVSJMp2tWM Hardware PWM
Other major developments? Maybe to me tiltrotor racers, and the possibly incoming dualboot option APM-Raceflight - PX4/Raceflight on the PixRacer that is in the mail for me... Blackbox logs ofc if you don't know it http://www.rcgroups.com/forums/showthread.php?t=2649495
@Maxime C, as far as openpilot forks go, librepilot and dronin are the ones to be looking at. the boards I know that came with tau labs code have since switched to dronin (the brainfpv in particular, an f4 flight controller. its successor will be out in august)
www.dronin.org
@Hugues, there are a few new hardware designs for ESCs now; a few 32bit ESCs (closed source firmware however), and some new 8 bit ones with hardware PWM, running a new flavor of blheli firmware
@Maxime, these are exciting times coming with such leading edge developments. Thx for sharing your knowledge on the topic, I believe it will interest lots of people on diydrones.
Are there to your knowledge other major developments for FPV racers we should be aware of (except what wasalready mentioned here) ?
I do think not much of it at the moment, having tried only betaflight.
I think the development of Raceflight is more vivid than Betaflight at the moment, possibly only more bragging on social medias. Lost of professional racers adopted it. In terms of frequency grind both are close as hell, betaflight now supports multishot.
What I personally think is the most promising about Raceflight is the mid/long-term view of its developers about the industry : they have in head a long awaited bidirectional FC-ESC protocol for closed flight loop move, evolving from PWM and PID controller. They have been doing I believe great work in diminishing control latency so far, but if what they state is correct the real progress is yet to come, in control accuracy.
What do you think of raceflight versus Betaflight ?
Btw Betaflight is not really a beta anymore... I upgraded all my racer builds to it (longing for Raceflight), and the flight performance discrepancy is huge, as far as yaw control, rotational speed, update rate...
Thx Gary & Chris for the comments.
I think you're right, smallest UAVs offer lost of advantages: cheaper to build (you could imagine in a not so distant future to use disposable drones), less strict or no regulations, less dangerous, totally usable for video/photography with new technologies (especially new software stabilization techniques), could be used in swarms to cover greater areas for a given autonomy, with increasing high bandwidth low latency communications links, more processing could be off-loaded to ground computers in real time making the drones even lighter. And there are probably more than that...
Thank you for bringing us this Hugues,
The FPV racing world has started a whole new group of enthusiasts completely separate from our group and there has been very little crossover of information between them.
Personally, I believe that the new really small high performance FPV racing systems will also end up having a great future in video and photography as well as the camera / gimbal systems get smaller.
The fact they are getting so small and light actually already puts some of them below the FAA's minimum weight registration requirement and it also greatly decreases real and perceived liability.
So there is a lot of incentive to go this direction.
Based on this I predict our paths will grow increasingly closer together and it is definitely worth our while to know what is going on there as well.
Best Regards,
Gary
-
1
-
2
of 2 Next