FPV Conspiracies - The perfect flight-controller
• Chris Landa
The past couple of weeks while refining and updating the flight-controller comparison chart I have been thinking about what my perfect flight controller would look like, and I wanted to share my thoughts and see what you think. Oftentimes I have a feeling, that I settle for something that is good enough and simply live with it, because I have no other choice, then I get used to it and do not question it anymore.
First of all: there will not be the “one” perfect flight controller that fits every application, so I will look at it from the perspective of whoops and micros/toothpicks - lets say up to light weight 3”.
Also I could not really decide for one set of specs that will fit both applications, but instead I will first list all the common features I wish my perfect flight controllers might have and then go into details what the differences for whoop and toothpick should be.
Power supply
The flight controller should be able to be powered from 1-3S. The 5V line should be rock solid and clean, 2A is enough. This means there has to be a buck converter that can step up 1S voltage and step down 2 and 3S voltage to 5V.
Preferably the whole power system is filtered well enough without needing a dedicated low ESR cap, but for 2 and 3S I would be OK with adding one.
Apart from the 5V line there should be 3.3V too for the receivers that need it. The pads supplying power for the receiver should also be powered when plugging in via USB.
ESCs
The ESCs should be capable of pushing 15A continuous, should have a low dead time of 5 and should be BlHeli_S compatible using the EFM8 BB21 MCU (the better “H” variant of the BuysBee MCUs).
I am not interested in proprietary BlHeli_32 and instead want to run the open source Bluejay firmware for RPM filtering - also this should keep the price down, since the manufacturer does not have to pay license fees.
MCU & Sensors
An F4 is adequate, F7 is obviously also OK, but I guess an F4 will probably keep the price down.
When it comes to sensors, I only want the bare minimum, Gyro and Acc. Technically I don’t even need the Acc since I fly acro exclusively, but in order to appeal to a broader audience, the Acc is a good idea. And the MPU600 is gyro and acc in one package anyway, there is no reason not to have it.
What I am missing from most (almost all) whoop FC’s is a blackbox, so my perfect flight-controller should therefor at least have a 4MB flash for the blackbox on board.
OSD should be included too, I don’t think I have to mention that.
Layout
All solder pads should be easily accessible on the edge of the flight controller all on the same side - the one that is considered “up”, preferably properly labeled on the silk screen - in case there is not enough space available for labeling a pinout should be available. Printed. In the box.
Motor pads should be multi purpose, so it should be possible to either solder a plug to them or solder the motor wires directly. Meaning an elongated pad with the hole for the plug in its center.
No plugs should be used for any of the peripherals.
No fluff
LED’s are not required. Barometer is not required. Neither is WiFi nor BT.
Bootloader button is not required either - bridging two pads should be enough, since you can enter bootloader mode via BF CLI anyway, no one is pressing the boot button unless perhaps when flashing a different firmware or something goes south - for those cases bridging the pads is enough.
Although I am a big fan of high integration, a VTX should not be integrated on the board, the boards that have a VTX integrated on the flight-controller itself were all an utter disappointment, I’d say we are not there yet.
Peripherals
2 UARTs need to be fully broken out, including pads for GND and 5V.
- Receiver: Apart from a TX pin, this UART should have an inverted and non inverted RX pin - this way any receiver can be used, also utilizing telemetry. As mentioned in the power supply section, the RX should be powered from USB when plugged in.
- VTX: the second UART is intended for the VTX - for analog only a TX pin is needed, but since we also want to be able to run DJI, this UART should definitely have both pins broken out.
Pads for video in and out should be conveniently located next to the second UART for the VTX.
Buzzer pads should be available too - although for me the motor beacons are enough, I’ve heard some people just cannot hear them.
Accessories
Rubber grommets for mounting the flight-controller, a low ESR cap, motor plugs and a power lead should be included.
Variations
Especially when it comes to form factor, I don’t think that whoop and toothpick can be satisfied with the exact same board, thus I would suggest 3 variations of this board:
- Whoop: USB on the bottom, and indented edges so that the FC will actually fit into whoop frames. Power plug should be a solid pin PH2.0 connector
- Whoop RX: Same as Whoop, but with integrated SPI receiver. I like highly integrated flight controllers and especially when it comes to 1S whoops, I think that an SPI receiver is adequate.
- Toothpick/Micro: This one can have the wider edges and the USB port should be on the side. Power plug should be an XT30 connector. This one should also have a 9V BEC in order to run DJI.
QC, Testing and Support
This is a separate section, because I want to emphasize this. Right now it seems as if there were no reliable whoop/toothpick FCs, they are all basically hit or miss. I want to be confident, that when I take the flight-controller out of the box, it will work flawlessly.
This section can also be seen as some kind of an open letter to manufacturers.
I imagine the discussions in the board rooms of FPV companies must sound something like this:
A: Alright, a week passed since we last released a flight-controller, it is time to release a new one.
B: Sounds like a plan, what should we improve on?
A: First of all, lets do everything as we did the last time, just move some components around, add a plug and slap a v 2.0 label onto it, higher number means better, our customers are suckers anyway.
B: OK, but how about improving the voltage regulator that our customers have been complaining about for the last five iterations on our boards, should we improve on that?
A: Are you insane? Why fix a perfectly broken thing? Customers got used to it and don't question it anymore, also, this feature keeps us in business, if the flight controllers don't break, what are we going to sell?
I want to buy a tested version - meaning: I do not want to be the beta tester and hope that at some point there might be revision X that is somewhat working.
So before release, this flight controller should have at least 2 iterations that are given to at least 50 testers:
- Initial release: Feedback is gathered from all the testers, assessed and implemented
- Revision 1: Testers check if their feedback has actually been implemented
- Revision 2-n: Those revisions are optional - if the first revision fixed all issues, then great, ready for launch, otherwise repeat until product is polished.
Use quality components, don’t use 2nd choice inductors with chipped corners. Make sure no solder blobs are randomly floating around on the board and shorting stuff out - generally make sure that the soldering is done properly.
Obviously, nothing is perfect, so be there when the customer needs you. Don’t just make a facebook group and let the community handle your short comings. Stay around even after you took our money.
Price
I am aware that I am asking for a lot, so this will obviously come with a price. I would be willing to pay up to 75$ for my perfect flight controller. I am paying the premium for QC, proper testing and a small premium, since I will not need to swap flight controllers every couple of weeks just because they are randomly dying on me.
Conclusion
I am surprised that such a thing does not yet exist on the market, what is your opinion? Do you agree with my feature set, what would you add/remove and what kind of money would you be willing to pay?
I think that if a company set out to produce this flight-controller they would have a solid lineup, especially with good QC I could see this lineup to run for years without the need of changing anything.
Chris is a Vienna based software developer. In his spare time he enjoys reviewing tech gear, ripping quads of all sizes and making stuff.
Learn more about Chris, the gear he uses and follow him on social media: