So heres the scoop. I had a well flying quadcopter using an APM 2.6. I decided to move to pixhawk. Bought a brand new unit from 3dr. Wired everything up and installed 3.2 Took it out to fly today and wasn't able to get it off the ground. The throttle and motors seemed to be acting weird and it sounded almost like a grinding sound. Keep in mind these motors flew through two batteries just the other day on my APM 15min flights each without a hiccup. The only thing thats changed since then is the flight controller. I brought it home and flipped the props and went to work trying to debug. I did the motor test in mission planner and ran each motor up to 15% throttle. Everything seemed fine, no weird grinding. So I got my camera out and started recording while I armed and tried the throttle. Again I was getting the grinding noise, then I noticed that above 50% throttle it seemed like no more actual throttle was being applied but the grinding noise got louder. I blipped the throttle a few times and sure enough it would seem thrust was coming up to the 50% point and then it was just making a louder grinding noise but not giving anymore thrust beyond that. SUDDENLY all the motors went to full power. No grinding noise anymore. I closed the throttle but nothing happened it was stuck wide open. I tried to disarm but it wasn't responding, so i took cover behind the bed as you can see in the video. The whole time I am trying to disarm. Eventually unsure of what my best move was I decided to turn off my transmitter. A moment later the motors returned to idle and I was able to disarm through my laptop (which I might have been able to do all along but in the rush I didnt think of it).
To put it lightly, what the hell is going on.
Video here, you can hear the grinding initially (warning loud)
Photo of the carnage below.
As I look back over that video. You can see the copter slides off my excellent tupperware support. Could that be the cause of the throttle lock? The board sensed it was leaned over and sent full power to that side trying to right itself? If thats the case than the full power isn't an issue, but the grinding noise still is.
If you look at the attached screen grab from MP you'll see that your throttle out (blue line) was tracking very closely RCIN #3 (red line). Channel 3 RCIN is what the Pixhawk is seeing coming directly from your RC Rx as the throttle signal. So I'd say you have an issue with either your RC Rx or its connection to the Pixhawk. The Pixhawk was doing what it was being told. You need to figure out why it was being told to do it. Rf interference maybe? Bad RC Rx? Correct wiring?
Another thing you should look at is that when you first armed the log shows it was in RTL mode. Double check your flight modes. Something doesn't look quite right there.
RTL is my default when I lose transmitter. So that makes sense. I also saw that line. Thats surprising that it would be sending a full throttle input as I haven't had any issues with this after using it for the past 6 months. None of this however explains the grinding noise.
Have a good look at your motors to see if any metal scraps or screws have fallen in them. Twice I have found screws in my motors after working on my bench. Remember, the magnets in those motors are very strong. It really is easy for a screw to fall in or for them to pick something up off your bench. Just an idea. But grinding sounds awful. Hope you find the cause.
Just watched the video. I don't think that sound is mechanical grinding. Hopefully someone else can be of more help. I would recommend that you reconsider your testing. After a trip to the emergency room for a couple of stitches I now have a hard rule of 'no props' on the copter when powered up inside.
Ya I dont think the grinding is mechanical. The motors spin freely by hand. When I did the motor tests to 15% throttle they all spun and sounded fine. Ive never heard that grinding sound before I got the pixhawk.
No props is a good rule. However I have no where outside to test it. I live in an apartment and have to drive 20min or so to the field. So basic testing just has to be done like this inside. In the past i've held it down with some bungee straps. With the new retracts I didnt use the straps... i'll be using straps from now on.
Good to hear you are taking precautions. There was a recent posting here where a copter took off inside that was quite a tale. Scary stuff.
I often times joke with my friends that maybe we need to be like lion tamers and have someone at our side with a shotgun in case things go wrong. :-)
So I did another test this time with lots of bungees. The motors are making a real obvious grinding noise. But it sounds like a digital grinding to me, like maybe the throttle output from the pixhawk isn't clean or something? I run up the throttle here and you can hear the obvious grinding. Then I do the motor tests again individually. Each motor makes the grinding noise when set to 25% throttle and tested. When I spin the motors by hand they spin completely smoothly and make no grinding sound, so im thinking it must be electrical/digital. These motors/escs were working perfectly fine with my previous APM setup. The only thing I have changed is swapping the APM for the Pixhawk and I also swapped my previous power module with the 3dr one that came with my pixhawk. Otherwise the motors, escs, power distribution is all the same.
Could the propellers be loose and rubbing against the motor shaft?
Well, when you have strange issues like this, take the props off and then do your testing. I couldn't see whether all motors are actually turning when the sound occurs and it's hard to tell whether all motors behave this way or not. If the pixhawk is the only thing that changed, it sounds like the receiver was playing well with the apm but not with the pixhawk.
I'd do the following tests to see what you can learn:
- remove the props
- remove the signal wires from all ESC's from the pixhawk
- turn on everything and track how the pixhawk is reading your throttle input. Then mess with the wires between the rx and the pixhawk a bit. If the connection there is a bit faulty, it can explain what's happening. You shouldn't see any glitches there.
- Then plug in each ESC one by one and apply throttle just a little bit until the motor starts spinning. See how well it tracks. if the grinding noise occurs, then you can investigate the issue better. Without the props, do not apply 100% throttle or you can damage the motor by overspeed.
- Then plug all ESC's in and see how you can make the speed of the motors track the throttle. Apply some bursts inbetween. Once again, do not apply 100% throttle.
No, they are on tight.
The motors don't have the issue when they are unloaded. So the props have to be on. All motors behave this way. There is only one wire between the pixhawk and receiver since its ppm only. From other things i'm reading about motor sync issues I think my next step is to try adding a ground for the signal wire.