Is it really possible to build a practical camera drone with HD video, a reasonable flight time, GPS-based smart features and a two-axis gimbal like the much heavier DJI offerings — whilst staying under 250g?

I think it is and I have already started work on the design of such a project. Tell me what you’d like to see in the sub-250g DIY camera drone project and help shape the final result.

Please also visit Mario’s excellent review of the Runcam Split Mini at:


  1. A bit too ambitious Bruce, I think a small foam flying wing with inav or similar might be the way to go, main problem is the stall speed so no hovering footage possible!

  2. Sub 250… my race kwads end up around 290 grams. Freestyle kwads end up 390 with a Run Cam 3 or session on top. Now that's with props and battery straps, I mean strap a battery on and fly weight. Building a 6" freestyle this week and I do t expect it be any lighter than the 390 ready to fly except battery. My 3" racer is lighter than 250 but not with a 1000 mah 4s on it.

    I just don't think you are going to get good film product outside with something this small. Wind is going to push it hard. Of course your not flying rate so that might help. I will watch and hopefully be pleasantly surprised.


  4. I've been gathering the bits to build slightly larger camera platform based on a HobbyKing Trifecta frame using a split. The tricky bit has been the gimballing of the camera. I have a carbon fibre gimble designed for servos, I'll be interested to see what you do.

  5. I have been trying for some time now to get the hunk-of-junk Onagofly to work, mainly for the said reason in this video. I have had some limited success, including getting it bound and flying with an actual transmitter. Where I've stalled on the project is tuning– I can't get a decent tune on the thing for the life of me. I think this is due to the PX4 FC is designed for large drones, and the Onago is a micro size. If there is a way to increase the PID loop time on an APM, that might do the trick, but for now (especially since it's battery bit the dust) it sits on my shelf. If you're curious, you can look at my project here:

  6. Really great idea 🙂 I would love for it to be able to have autopilot with point of interest, for self filming extreme sports

  7. Thanks for linking to my split mini review, totally no problem if you use some shots!
    I have it on the diatone gt90 now. 156gr drone but not tge best quality, more focused on the fun aspect while still being able to share so videos In hd.
    In facebook i saw paul nurk chasing another drone. He filmed this with gopro fusion. Looks pretty good but will be too heavy for the sub 250gr build.
    Good idea btw, i think there will be a lot of products below that 250. I expect some very powerful little racers… greets, Mario

  8. Bruce, this is a great idea. I'm a real estate photographer and would love to have something under 250 grams. If it could do still photos as well that would be very helpful.
    Its funny to think of pulling out my goggles instead of a tablet to do the aerial shots.

  9. Definitely following this! I've found I'm more interested in aerial camera platforms than racing. I'll likely try to duplicate your results in the end.

  10. This will be interesting Bruce. I've battled with this challenge for about 18 months for a camera drone to take on motorcycle adventures. My family finally got sick of my constant build experiments and bought me a DJI Spark for X-Mas!!
    I've used a 250 frame with GPS, magnetometer, barometer etc. with servos (too jerky), 2-axis gimbals, ending up heavy and sluggish with short battery life. Long story short – most recent was a 180 with GPS (Omibus F3 AIO), 600mW VTX, FlySky FSiA10 (for camera control), 1000 mAh battery and mounted a Firefly 8S (with optical stabilisation @ 1080P) without battery. It's works okay. Dry weight is 309g, battery jumps to 390 and camera for a total of 442g.
    Getting it all under 250g with 3-D control, good optics and video stabilisation? A bloody miracle!!

  11. Awesome project, and can't wait to see it through to the end! I've also been looking at converting my 6in tricopter to a camera drone (definitely gonna go above 250 grams), but I never thought of using a split with a gimbal! And I was planning to put 2 cameras, 1 FPV traditionally mounted and 1 HD cam, but I was thinking of figuring out how to switch between cameras remotely and view via FPV goggles, and that's where I got stuck in the design process.

  12. Go with a 250 size quad with 6 inch props for more efficiency, and don't worry so much about weight. 250 size equipment is cheaper anyway. Get that working, then move on to the challenge of 250 grams.

  13. Hi Bruce, the biggest problem with a DIY camera drone is a good steady position hold function. I have tried INav with an F3 controller but still it’s not good. Crack that first and all the other stuff will fall into place, cheers John.

  14. Mobius and Hawkeye Q6 cams can be split from their mainboards using extension flex cables if you want 4K raw footage (Q6 has built-in electronic stabilization). 51 pins with 0.3mm separation is available for the Hawkeye through various Chinese sites(I just ordered one) and the Moebious cable is officially available (credit to ROBOFORCERX2000

    Hawkeye Q6 running on 1S 3.7v you will need a voltage regulator. Cobra 2204's are still around in the high-effiency 1960kv version. I wonder if Arduinos are fast enough to act as flight controllers or even for real-time image processing, I hope people come with better input than I do. Also see they have some interesting ideas and regarding battery life, all good comes to he who waits:
    I have a spare Mavic gimbal+cam and a dissected Dobby if you want that sort of donations, they are basically higly advanced mobile phone cams on flex cables with no available diagrams so I scrapped them.

  15. I have a scrapper made from a Diatone grasshopper and is 230g with a 1000mah 4S on it. I've used 1306 3100kv motors with 4045 twin blade props. I get eight minutes or so just flying about normally and about 5 mins flying it hard.

  16. 10 minutes lfight time is not unrealistic, I've been testing a 220 sized FPV quad I slowly built with cheap/on sale parts, AUW 339g with 1500 3S battery. I did some line of sight flying with lots of punch outs to test its power and longevity.
    After flying for 10 minutes I landed and checked the battery, it wasn't completely drained.
    Will be watching these videos with interest.

  17. I am so happy and ready for this project. You have helped me learn so much and now you are starting my literal dream project.
    Thank you so very very much.

  18. Back on the path of why I initially followed. Interesting build videos, problem solving and how it works lectures. Can't wait to follow the build!

  19. A few years ago i did some testing with the zmx 1804 2400kv motors with 7038 slowfly props on 2S. Its a super efficent setup and lightwight too. Maybe this is a point to start….!

  20. Hey bruce, the 250g limit doesn't apply in NZ does it? I have recently read part 101 and cant find a reference to a lower limit anywhere

  21. Bruce, I've been pondering and tinkering away at the same idea, so I heartily welcome your proposal. As for the camera, I completely dismantled an Aosenma CG035 and it turns out that the HD camera and recorder together (from the v2, egg-shaped gimbal) only weigh around 10 grams.

  22. This project! It's been kicking around in the back of my head for the last couple of months. Really looking forward to seeing how you go about this, especially how you approach making a suitable micro gimbal.

  23. I just did an Atom v3 with cobra 1407s a f4 with 3 inch props and it's working well .I really want one of these .Thank you for the video Mr.Bruce.

  24. Bruce, great idea. I would like to see that proper metadata for the images is created for georefrencing so that I could use them with photogrammetry software.

Leave a Reply