That’s slighly worrying, though hopefully H_SV_TEST should indicate if any servo was damaged during power-up, thus avoiding any in-flight issue.
Seems like a neat solution. Would the FC draw any power from the USB port once the main power is connected? A small BEC attached to the balance plug of the main battery might be a solution – plug in the BEC just before connecting the main leads.
All autopilots I know of have isolation or at least some measure of protection between VBat and USB power, so it should be no problem to power them simultaneously.
OK, but I thought the whole idea was to power the processor (via the USB port) before supplying power to the servos, so that the servos can’t respond to any transient signals generated during boot-up.
Right, so at some point, both will have power applied…
Hi,
I haven’t read this whole thread but the initial problem is similar to one I had with my hexarotor. On arm the two servos I had attached would shutter back and forth. The same thing would happen on disarm. I tried adding chokes, capacitors, and twisting wires.
In the end the solution was switching from an analog to a digitial servo!
Hope this helps.
Thanks, These are digital servos already.
Iv tried powering everything but the FC and the servos never move. As far as I can tell its caused by some sort of signal from the FC.
@abenn1 I power my FCs from USB and the battery in various order of plugging in all the time. Not just this one that has servos. So far it has not caused a problem and my understanding is they are all designed to be able to do this. Iv read others do it to get GPS fix since that is normally powered from a 4.5v pad that gets power from USB and battery too.
I’m not a fan of having to plug in the micro USB every flight though. Its in an awkward spot and those plugs do wear out. I setup a WiFi adapter on this aircraft to enable plugging in less even.
This seems like a problem that really should not have existed even this long. When I search for it though I find post in plane that date way back to people having the same issues and no solution. This happens with the receiver and transmitter on and with it off. I have yet to find anything that consistently causes it though.
For now I am just working around the problem as I have another very similar build to start working but with different FC hardware so I am curious if it will have the same issues.
I’ve used these for various applications, and they would be a good option to alleviate some of the inconvenience of this interim fix.
I suspect what you are seeing is the difference in GPIO settings between the bootloader and the actual firmware. A change was made relatively recently (4.3 I think) to make sure the unused pins in the bootloader are pulled low rather than high, but you would have to update your bootloader to take advantage of this.
@andyp1per, excellent! I’ve never had a problem with servos/PWM outputs on boot, but I definitely despise the brief high state of GPIO pins used for relays while the bootloader does its magic.
That looks handy. But they say it connects to a “5v balance plug”. There is no such thing as a 5v balance plug on any of my batteries, so does it incorporate a BEC to get 5v from, say, two LiPo cells?
It connects to 2-6S balance plugs and provides 5V out. The description is probably lost in translation.
Was/is the issue specific to a certain FC brand or a specific board?
I’m unsure right now. I posted that info for me but have not seen any others jump in and say they have the same problem with the same board.
I have a H743-Slim that is going in to my next build. It will hook up in a similar way with external power rail. My plan is to just use the work arounds suggested on the problem heli for now until I can build and try the next one.
Thanks spova. I’ve found a UK source for the 5v gizmos that Yuri_Rage linked, and have ordered a couple.
I updated the bootloaded and so far have not had the issue again. Though I don’t have a lot of cycles on it yet. Ill report back if it resurfaces though.
Super - glad to hear it!
And just happened again. What servos it happens on is not always the same and this time it was just the tail.
And this has now caused a catastrophic crash. Lost the tail control of my helicopter because the servo going haywire on startup cracked a linkage but did not pull out all the way so I didn’t catch it before take off. Once I hit too much yaw torque it pulled through the cracked linkage eye and the heli went in to a spin.
What are you fixed wing pilots doing to work around this? Is everyone just using a secondary power source to boot the FC before powering servos? This seems like a pretty critical issue. I don’t have any problems with servos moving before boot on inav.
I recall there being another safety that stops servos from moving but heli setup suggests turning that off.
Im using chibios, Would possible changing to nuttx help?
I also have the brd_safety off but I’m wondering if there is a way to have this on during boot up and then turn off automatically. Seems like servo output should just be disabled on startup anyway though.
It turns out that the GPIO setting does not take effect in the place that most people put it. I can do you an updated version if you want to test.