WKM mid-air reboot... All motors stop.

W

wbboydcofc

Member
Hello all,

Have been flying MRCs for a while now, and have been enjoying silently reading the forums, but now I need some feedback.

I have an otherwise perfectly stable S800/WKM/Zen rig that displayed the most odd and frightening behavior on a shoot yesterday evening. I have not had any issues with white flashes, but I do suspect what happened was the result of vibration induced garbage in the IMU. Firmware is 5.16 and gains are 160/160/140/150. I'm going to look at improving the IMU vibration isolation when time allows, but I wanted to toss this out there to see if anyone has seen this behavior before. We've all seen the s800/wkm crash videos, but I have not seen any that include a mid-air recovery like this. The first link is the glitch, and the second link is what the flying leading up to it looked like. There were no beeps when this happened, and if you pause the video during the free fall, you can see that the green light on the bottom of the ESC is still on. Any ideas or feedback would be greatly appreciated,

Thanks,

Brad


Fall
http://www.youtube.com/watch?v=RmCHfNbDDhI


Rest of flight
http://www.youtube.com/watch?v=-NwpkH0S6PM
 

olof

olof

Osprey
If you have the V2 Iosd then it should be possible to figure out what happened. V2 has a data logger built in.

What shape was your Lipo/s in?

Spectrum TX/RX radios sometimes have brown outs, for many different reasons including low voltage.

It could be a bad connection or loose wire.
 

W

wbboydcofc

Member
Unfortunately no iOSD, but I suspect that will be changing soon. The lipos (2 x 5000mah thunder power) are in great shape, and the voltage right after this happened was 24.3. Under load, this would still be in the 23 range.

I do use JR9503/DX7i txs and AR9010/AR8000 rxs (both with sats), so I think you may be on to something. There were no beeps that would indicate rearming, and failsafe would keep the throttle channel at mid-stick. I guess it's possible that a brownout could give a zero throttle signal briefly then go back to actual stick position.

Thanks!
 

gtranquilla

gtranquilla

RadioActive
The fact that the S800 remained level with the horizon suggests that the controller was not rebooting as the gyros etc. are still giving it orientation control. It is common in the spring, summer, fall for thermals aka updrafts to occur. Once your MR gets to the top of a thermal it can suddenly fall over the edge into a downdraft. I used to hang glide so I experienced it several times.
 

W

wbboydcofc

Member
Actually, I think the short duration of the power loss, and the fact that it was all motors at once is the only reason it seemed to maintain its attitude. Well, I suppose that's not entirely true... The zen made it look a LOT smoother than it really was. What you see flying through the top of the frame is one of the motors, and the camera was pitched down. Had I been much lower, there would have been no margin for recovery.
After Olof's comments, an rx brownout makes total sense. I've observed somewhat similar behavior on my 550, though it was from a much greater distance. Totally different feeling watching this thing fall from the sky though!
 

D

Dewster

Member
wbboydcofc said:
Actually, I think the short duration of the power loss, and the fact that it was all motors at once is the only reason it seemed to maintain its attitude. Well, I suppose that's not entirely true... The zen made it look a LOT smoother than it really was. What you see flying through the top of the frame is one of the motors, and the camera was pitched down. Had I been much lower, there would have been no margin for recovery.
After Olof's comments, an rx brownout makes total sense. I've observed somewhat similar behavior on my 550, though it was from a much greater distance. Totally different feeling watching this thing fall from the sky though!
Click to expand...

man...that's a lot of money in the air. The Zenmuse did a good job of keeping the camera steady despite your falling craft. Is your Zenmuse powered by the flight battery or its own? Do you think you were pushing max amps??

There's a sickening feeling when you watch your craft behave like that. All of the what ifs start going through your mind and you lose trust in the equipment. It is amazing how it can recover once power is supplied. If it took as long as some computers to boot up you would be through.
 

olof

olof

Osprey
Spectrum TX/RX brownouts are strange and do often lead to loss of control. Most Spectrum RX's have a light that that blinks if there has a been a loss of signal. It is a good idea to check this after each landing. I know it is too late now. But check in the future.

One strange way to get brownouts on Spectrums is if you turn off power to the RX w/o shutting off the TX as well. I have had this happen on RC planes and I have seen a devastating crash caused by "lockout" on a Spectrum RX. Yesterday I had a Cub with AR6214 RX (with sat) loose rudder control after taxying a while (it worked fine at first), I discovered right before take off trying to line up. A reboot of both RX and TX fixed this. And the flight went fine.

I use Futabas on my MR's they seem more solid.

Flying RC planes we see a lot of RX/TX problems, momentary loss of control of a servo etc. Some of these are because of how the models are built and old crash damage or just stripped servos, bad wiring.

I am always amazed at how reliable my Futaba/DJI combos have been. Of course with $10K in the air I really preflight and triple check everything like TX battery power.

On any given day at our RC field there are several crashes, well over 90% are pilot error but not all. And some are unexplained. The most popular "cause" (excuse) is there was a wind gust.

My point is RC flying is not perfect and things do happen.
 

R

RTRyder

Merlin of Multirotors
Make sure you have the TX/RX and flight controller failsafe setup correctly, if it isn't the end result of a brownout or RX failure can be catastrophic and setting just the throttle failsafe isn't enough. I always setup the radio failsafe to be enough throttle to maintain a hover, aileron and elevator at neutral position, and switches set to GPS position hold and flight controller failsafe/RTH. When setup this way just about any failure of the radio gear will result in the multi hovering for a few seconds waiting for signal to return, followed by RTH if it doesn't. Any other configuration can result in either a flyaway or crash and is something often overlooked in a build due to not knowing how to properly configure the radio gear to do it. With JR/Spektrum, any RX that has a data port has full channel failsafe which is what is required to have the failsafe setup as I described. You must first do the normal bind procedure, setup all channels and endpoints for proper operation, and then do the final full channel failsafe. This is done by setting the TX controls to proper positions, i.e. throttle to midstick, RTH switch on, as well as GPS position hold switch on. Insert the bind plug into the RX, power up the RX, and when the LED starts to flash remove the bind plug. Now turn on the TX with controls in position and bind button held down as you would for the initial bind, this will set the failsafe for all connected channels to where they are currently set.

Cycle the power and with the WKM connected to the Assistant you can watch the TX control positions while powering off the TX, the throttle should go to hover and the flight mode to failsafe, turning it back on should return the control from the TX.

I've tested this procedure with both WKM, Naza, and even my Multiwii flight controllers. Loss of radio signal results in the multi just stopping where it is in the air then resuming flight when the signal returns, I see it frequently when flying FPV a bit further than I should on 2.4. Complete loss of signal should trigger RTH once the timeout limit is reached, just make sure you have the proper altitude configured to clear any obstacles between takeoff point and wherever you're flying to.

Ken
 
Last edited by a moderator:

W

wbboydcofc

Member
Thanks Ken.... TX/RX failsafe are indeed set correctly, with throttle mid-stick, ail/ele/rud neutral, gear down and aux2, which is my flight mode switch on the 9503, set to failsafe/RTH. WKM assistant shows the system goes into failsafe when the TX is turned off, and I have tested it numerous times in the past, and it worked each and every time. TX control returns when the tx is turned back on. Can a RX brownout cause random channel outputs, but not trigger a failsafe condition? It is almost as if the rx went into SmartSafe positions rather than the preset failsafe. Smartsafe would have all controls neutral and throttle down. I have since checked operation through the WKM assistant, and the preset failsafe positions are working. Starting to lose faith in that RX.
 


R

RTRyder

Merlin of Multirotors
wbboydcofc said:
Thanks Ken.... TX/RX failsafe are indeed set correctly, with throttle mid-stick, ail/ele/rud neutral, gear down and aux2, which is my flight mode switch on the 9503, set to failsafe/RTH. WKM assistant shows the system goes into failsafe when the TX is turned off, and I have tested it numerous times in the past, and it worked each and every time. TX control returns when the tx is turned back on. Can a RX brownout cause random channel outputs, but not trigger a failsafe condition? It is almost as if the rx went into SmartSafe positions rather than the preset failsafe. Smartsafe would have all controls neutral and throttle down. I have since checked operation through the WKM assistant, and the preset failsafe positions are working. Starting to lose faith in that RX.
Click to expand...

Anything is possible, I would think a brownout is one of those situations where you never really know exactly how the RX will respond. How much voltage did it lose and for how long? No way to really know but if you were to do some controlled testing with a variable power supply to see what happens at different voltage levels you might get an answer to what will trigger a failsafe vs. what does not and how the RX responds.

I've never had an issue at close range, knock on wood, but I have had signal "brownouts" flying FPV which is a very different thing as the RX still has power, just no signal from the TX and if it stays that way long enough it will kick into failsafe via the F/C sensing the loss of throttle signal.

For anything that flys out beyond LOS I use UHF equipment in the 430 - 435 mHz range, have yet to have an issue with that setup but I also tend to fly away from built up areas and any known sources of RFI so maybe I've just been lucky to not have problems with either radio due to being "out in the boonies" ;)

Ken
 

You must log in or register to reply here.
Top