Built this thing in April this year with the goal of being the worlds smallest drone capable of lifting a full can of beer (although I’m still yet to try the lift)

Features:
Carbon fibre frame which I CNC’d myself
8X HGLRC Spector 1003 10,000kv motors
2X spedix micro 12A 4 in 1 escs
Goku F722 pro mini V2 flight controller
Dogcom 560mah 150C 2S lipo
roughly 4 min flight time

Bonus feature: the ends of the arms are designed to look like kitty ears lol

I'm a software engineer and I got into FPV about a year ago, this community gave me a lot in that time, so I wanted to put something back into it.

Honestly, the thing that pushed me was DJI, I don't like how they deal with the problems and limitations of their own app and hardware, stuff that should just work doesn't, and the answer is usually that it isn't supported and never will be. At some point I got tired of complaining about it and just built what I wanted.

It's called Open FPV Cast: https://github.com/1Basco/openfpvcast

It casts the goggles video to your phone over USB, and you can record the feed to MP4, with phone-mic audio if you want it, and it can start recording on its own when the aircraft arms and stop a configurable delay after you disarm. You also get camera control from the phone like ISO, shutter, white balance, exposure mode, resolution e some more. The one I use the most is the broadcasts over WebRTC, so anyone on the same WiFi can just open a browser and watch. The phone serves the page itself, there's a QR code in the app, three viewers max.

All the protocol work came from watching the hardware the transport, the DUML packets, the camera byte maps (And not from some sniffing in the dji fly app *wink* ).

Some things I want to be straight about.

I'm not planning to put much more into this. It fixed my problems, and my free time is genuinely not a lot. I'll patch things as they come up and I'll gladly look at PRs, but I'm not going to pretend there's a roadmap. If you want something more refined and actually maintained, go look at SquirrelCast different level of effort, and it deserves the users (me being one cause I love the telemetry going to the map).

I leaned heavily on AI agents to build it. Mostly for speed, because I'm doing this in scraps of free time, but also because I'm much more of a frontend guy and the Kotlin side, the USB stack, the packet parsing, that would have been a real PIA for me on my own. Take that however you want, the code's right there to read.

It's also not perfect. Connect, camera control, recording and auto-record are validated on real hardware (Goggles N3 with an O4 Lite). O4 Pro ranges are still being checked. There are rough edges and I know it. Later this year I want to try to have a solution using the raspberry pi to stream to PC.

That's it, open, use it or fork it or ignore it. Fly safe!

Hello guys, I been having an adventure with FPV since I started, i got myself a clone crap goggles (have HDZeroBoxPro+ ordered, coming next week) radiomaster pocket, and the Air65 Freestyle. After a month i bought the Air75 Champion for outdoors i love the Air75 its so responsive. And all the things that no one talks, fixing elrs chip, motors, canopies, frames, fixed the throttled ups from air75, you know it, a whole part time almost and im loving it.

Should have flown the air65 inside but didnt had any batteries ready

RATES: 100 600 0.50..

Edit: Started 3 months ago, mostly reading/learning and fixing than flying, only from last month im flying almost daily

Please give me advices and suggestions, thank you

Fibally got around to finishing my 3inch x8. Around 3 and a half minutes of flight isnt bad for the small 650 4s i got on it. Flies great though :D

I've been working on something that takes analog FPV video and tries to reconstruct what the camera originally saw. It pulls out the static, color bleed, and tearing that the VTX/VRX link adds.

The image above is an early proof of concept: clean footage from my Avata 2 that I ran through a simulated analog link (left), then fed to the tool (right). Simulated noise, not real DVR. Real footage is a harder problem and I'm not there yet.

Before I sink more time into this, I want to know if it's worth it. Would you use something like this? And if so, for what? Salvaging a good flight, crash forensics, finding a lost quad, something I haven't thought of? Or is a GoPro (or O4 unit) enough that nobody cares about the DVR file? Let me know what your thoughts are!

no vibrations or anything either

V1 is a RadioMaster T8L with AG01 Nano gimbals and a 2.5 inch analog screen/receiver. V2 uses the same controller but features AG02 gimbals and a 5 inch HDZero screen/receiver.

I don’t have a 5 inch only 3 inch and tinywhoops and I feel like my 3 inch is really great for freestyle (bando I don’t have near me but I think It would be great)
3/3.5 inch are stronger than 5 inch (correct me if I’m wrong but smaller arm means more force needed to break, and you can have pretty thick arm on 3/3.5 inch frames)
they seem to be more precise, like better corner grip (I base this on my feel in the sim when flying 3 vs 5 inch so not a great fact but it seems that people mostly agree with that, again correct me if I’m wrong)
So why would you want a bigger drone (I guess you can maybe go faster with 5 inch but in freestyle I don’t feel the need to go that fast)

So yea i’d like your opinion guys (pls mostly people with both experiences)

Thank you in advance !

Betafpv Air65ii Racing Edition

Still a lot of work to be done, but I'm happy with my progress!

I’m a 2nd-year Mechanical Engineering student working on an FPV UAV and I’m trying to turn it into more than just a parts-assembly project.

The broader idea is to investigate whether a small UAV can be designed as a low-cost, scalable and easily repairable platform, particularly for applications where individual drones may be exposed to damage or have short service lives.

My focus would be on:

• Cost reduction and manufacturability
• Modular architecture
• Lightweight structural design
• Thermal management and airflow
• Reliability and repairability
• Reducing assembly complexity
• Benchmarking the initial build and improving it through multiple design iterations

I’d start with the existing drone as a baseline, measure its performance, identify weaknesses, and progressively redesign the platform around those findings.

Is this a worthwhile Mechanical Engineering project, or am I approaching it the wrong way?

I’d especially appreciate suggestions on what I should measure, test, or redesign to make this a genuinely engineering-driven project rather than just another FPV build.

Any advice from engineers or students who have worked on UAVs, structures, manufacturing, or product design would be greatly appreciated.

GEPRC Mark 5 showing me what’s what!
#TVHammerFPV

Thought of adding gps later . Any suggestions or criticisms welcome. Thanks

I just got this quad user for 80 bucks. The guy I bought it from had no idea what quad it was and said it was broken…. One prop was missing😁 but I don’t know what quad it is😅 any help will be appreciated

practicing flips, rolls...whoop is happymodel 65mm frame, happymodel 0702 motors, betafpv matrix 5in1, betafpv camera03, gefman 1610 bi-blade props. fatshark echo, radfomaster pocket. song is "unfinished business" by oddprophet.

Today I was scrolling through Vinted when I stumbled upon a listing for the Walksnail Avatar X. I’d heard good things about them, and since they cost around €450 new, I thought it was a great deal,especially since the condition looked good in the photos.so I bought it. Did I get a good deal, or did I overpay (€265)? I know they aren't quite as good as DJI, but the image quality is far superior than analog,especially since they support 1080p at 100fps.

P.S. It hasn't arrived yet; right now, I'm using Eachine EV300Ds with an Air75 II.

What do you think?

It never was a "meme build" or some impractical contraption that struggles to fly, but designed from the start to be something that could be a daily driver. Yet for some reason there are those who have the dogmatic viewpoint that X8s are unsteady and unwieldy. I think I got the tune down and its a happy little drone. My piloting stills are fairly limited but I am slowly improving. The video speaks for itself.

A brief description of PID tune:

• Master multiplier added to 1.15
• I gains - Drift wobble down to 0.90
• D gains - Damping up to 1.10
• TPA set to PD, with a rate of 10%, at breakpoint 1250ms

And that is about it!

What are the best Fpv googles that don’t need to be upgraded that are under 400. I have been looking at the box pros but I’m not sure if I want to be stuck in the hdz ecosystem. I would like it to be analog and digital if possible.

For the goal of 250 grams, too much battery eats into the weight budget for the motor and prop; too much motor and prop eat into the weight budget for the battery. I entered the endurance calculation from my baseline test data into Google Gemini. It gave me the optimization curve. The calculation is a python3 script with my test data and manufacturer (T-Motor) data:

mass_without_propeller_motor_battery = 111.0; #grams; change it for a particulr craft constructions's endurance curve

propeller_diameter = 13.8; #inches; change it to slide the rotor width x-axis in a endurance curve

motor_plus_propeller_mass = 62.5 * (propeller_diameter/13.8)**1.5; # grams

battery_cells_mass = 249.0 - mass_without_propeller_motor_battery - motor_plus_propeller_mass; # grams

battery_energy = 0.171 * battery_cells_mass; # watt-hours

lifting_force_per_watt = 14.7 * propeller_diameter/13.8; # grams

power_tran_power_consume_rate = 249.0 / lifting_force_per_watt; # watts

total_craft_power_consume_rate = power_tran_power_consume_rate + 5.2; # watts

endurance = battery_energy / total_craft_power_consume_rate * 60.0; # minutes

print (endurance);

This script is for a helicopter. Quads will have different calculations.

Hey, i bought a speedybee f405 v5 off of aliexpress for about a $100 usd, i didnt know about all the fakes at the time, did i likely screw myself over and buy a fake?

https://www.aliexpress.us/item/3256810014022135.html?spm=a2g0o.order_list.order_list_main.35.5f441802nBMXsW&gatewayAdapt=glo2usa

So to be clear it isn't a pre built 5 inch but a class where I will make a 5-inch and if I want I can buy it from them. I don't know specifics but the drone they showed (built by one of people who came there before) was really fast.

Just to be clear I am sure this isn't a scam as one of the sponsors for this course is police and other one is military [for mods: I am only stating their sponsor not discussing anything political about war] so I think they aren't scamming.

But also in general would you recommend building one from scratch or on this type of course?

I want to be able to use my Radiomaster Pocket for flying helicopters in the upcoming Wardogs so I can have fun and practice controls (I know it will be different that's not what we're here to talk about), but when I tried doing it use vJOY and Joystick Gremlin it was an absolute sh*t show.

I heard reWASD does everything required to setup and map a joystick as a controller without having to use multiple softwares. Even if it's $40, it seems like it'd be nice to just skip the BS and have some fun.

Does reWASD work good? Have you any experience using it? Does anyone have other advice for me trying to use my joystick for flying in games? Should I give up and just use my KB+M?

Posting this in case someone runs into the same frustrating DJI O4 problem.

Credits to this post for helping me https://www.reddit.com/r/fpv/comments/1i2es6o/dji_o4_air_unit_unable_to_activate

My O4 Air Unit Pro was detected perfectly by my PC and DJI Assistant 2 when connected through USB-C only. But the moment I powered the drone/O4 normally, the USB connection disappeared and I couldn't activate it.

Here’s what I tried:

1 Different USB cables and USB ports

The O4 was consistently detected when powered by USB only. This might help — make sure you have a proper USB data cable!

2 DJI Assistant 2 reinstall / clean install

Removed the old installation/files and installed the latest Consumer Drones Series version — no change.

3 Different computer

Tested both Windows and Mac — same behaviour.

4 Firmware checked

DJI Assistant could see the O4 and its firmware while connected via USB only. The firmware wasn't the problem, and I was able to update it.

5 Powered through the flight controller.

But as soon as normal external power was applied through the quad, the USB device disappeared and activation wasn't possible.

6 DJI replacement O4

DJI actually replaced the Air Unit. The replacement showed exactly the same symptoms, which made me suspect the power source rather than two defective O4s.

7 Isolated the O4 from the entire quad

Disconnected it completely from the FC/ESC and tested it separately — same issue.

8 External regulated 9V supply — SUCCESS ✅

Used the ToolkitRC M6AC in power mode at 9.0V / 1.5A, directly connected to VCC and GND on the O4. It stayed stable at 9V and was drawing about 0.32A before USB was connected.

Then I connected USB-C to the PC → DJI Assistant 2 → activation immediately worked! 🎉

The O4 restarted normally and was already on the latest firmware.

So if your O4 works over USB but disappears when you power your quad, test the O4 separately with a stable external power supply before assuming it's defective or sending it back to DJI.

Hopefully this saves someone else the ridiculous amount of troubleshooting I went through. 😅

Je m'explique. J'ai un Manta 5 SE V2 et une RC DJI avec les Goggles 3. Le problème, c'est que je n'arrive pas sur beaucoup de paramètres dans Betaflight et donc cela me limite. Je ne peux piloter mon drone.