My question is this: is the Y6 actually supposed to compensate in cases of motor/prop/esc failure, or is this configuration really no more reliable than "normal" multicopter configurations?
I've been flying my Y6 regularly for nearly a year now. I've upgraded to the 2014 frame, upgraded to Pixhawk, upgraded to 4S batteries... it's been a great machine and I've been very impressed with it. And although other multicopters seemed lighter/faster/more-agile, I thought the Y6 was better because it was "safer".
Yesterday, mid-hover (in Loiter mode) the motors sounded wrong. It looked like the bottom motor on the tail slowed down, then nearly instantly the whole thing flipped upside down and drove itself into the ground. When I tested the motors afterward the back-bottom one didn't respond at all... it appears that it failed mid-flight.
It looks like the damage is limited to the GPS mast, a few stand-offs, several propellers, and at least one prop-adapter that was smashed into a rock during this upside-down dive. It could have been worse, but I truly believed that it was supposed to gracefully recover from a failure like this.
I chose to invest in the Y6 over other configurations with the understanding that the stacked motors/props provide redundant safety in the case of a motor/prop/esc failure. Is this really the case?
Replies
I just posted to Chris' post on previous page about this exact issue. I recommend a safety wire/interlock to prevent hubs from being spun off. Also recommended they develop and sell right and left-hand prop shafts to ensure the nut/hub torques agains the prop in the event a prop touches the ground while the motor is still turning. However, the simpler process you mentioned mybe the quickest and easiest solution to this present problem. Thanks for the idea.
Yeah, I always wondered if the cone shape of prop nuts actually serves a purpose and have used hex lock nuts on occasion as well. Seems like everyone with a pro or semi-pro bent is using three-hole prop attachments these days anyway.
This exact thing happened to my Y6, only it took three crashes to diagnose the problem. It appears that the Y6 will fly with a motor out under SOME circumstances but not ALL, and it does not indicate that a motor is out unless you happen to hear or visualize the problem on run-up or in flight. In my case the logs showed that if the remaining motor on the affected arm becomes saturated by an attitude command or a throttle-up, a crash is likely to ensue.
I chose the Y6 explicitly for the redundancy that 3DR so proudly advertises. It was a huge let down. I have carefully analyzed my crash logfiles with the help of one of the APM Copter programmers. I contacted 3DR with the thought that they would value real-life field logs from failure events. If I were a manufacturer I would crave this type of information. They were extraordinarily uninterested. Hovering around in place and killing a motor to "prove" redundancy is fine, but flying around with a motor out and not being aware of it is a completely different story.
My Y6 is my video workhorse but I would never fly it over person or property with the thought that it could confidently maintain safe flight with a motor out. And I would never buy one again.
I expect Andrew Chapman on drones-discuss would be very interested in your logs. He's written a nice python log analysis package, but I don't think it has motor failure detection yet. Of course, it's for analysis afterwards and can't run realtime on the FMU.
Hi Amos
I have had a similar experience with a Y6. It had height (60m) and was in Auto mode when it occurred.
I suspected a motor / ESC failure but through diagnosing the log files, discovered that - even though the GPS was reporting (via Telemetry) status 3 (3D lock) 11 satellites locked and flying the proper course, the GPS had lost power. This caused the copter to flip and crash at the next waypoint.
So my comment is: have you fully checked your log files. The busted ESC could be a result of the crash - the real culprit might be something else.
Nigel Brown
I've seen videos and tales of it working as it should. Time and altitude will not help if it just flips over. You're not going to recover from flipping over no matter what. It should be able to maintain position, altitude, and attitude with the loss of any one motor/prop.
If the props are mounted right, it can handle the loss of one motor/prop fine. We have that happen pretty regularly and sometimes don't even notice (when it's on an automission). We've even seen it keep flying with two failed (that can work if they're symmetrical, with one off on top and one off on bottom)
However, it does need to be tuned well you need a little altitude for it to recover. In your case, are you sure the problem wasn't something else, such as all the wiring on one arm (both motors) coming loose temporarily?
Chris,
Saying it happens "pretty regularly" is a little alarming in itself. But it's the part about sometimes not noticing the failure that really gets my attention. When and how do you notice the problem then? That specific event would seem to be crucial to the operator as an indicator to return to launch and perform a close inspection of the craft.
I will enthusiastically attest to the Y6's ability to hand the loss of one motor in a hover and steady forward flight but clearly not under certain climb, turn, and RTL situations which will saturate the remaining arm-pair motor and likely result in a crash. As the pilot of a factory-programmed RTF Y6 that crashed multiple times due to an intermittent motor out, diagnosed by Randy Mackay on the basis of logfiles, and confirmed by ground testing, I was needless to say pretty disappointed in the motor redundancy premise.
The concept of redundancy is excellent, but to be tenable it must either be absolutely trustworthy or it should come with some caveat as to how to recognize and contend with a motor failure.
Ed, it happens when we're with new users who don't tighten their props properly. We do a lot of training sessions with newcomers, so we see every kind of mistake you can imagine.
We can usually just see a prop fly off, but sometimes on an auto mission it's not noticed until the next flight.
In the next generation copters, we'll have closed-loop motor controllers that can detect such failures automatically and enter failsafe modes, but with standard ESCs that's hard to do reliably. I suppose you can think of it as a continuum: A hex is more reliable than a quad; a Y6 is more reliable than a hex; an Oct is more reliable than a Y6 and a X8 is the most reliable of all.
Chris,
Would it be smart for 3DR to develop prop shafts that have left and right-handed treads for counter and clockwise motors? Maybe a safety wire type or interlock to ensure props don't spin off inadvertently.
I can tell you of a few situations where I tipped my Y6 forward catching the grass with a prop while the motor was spinning and the hub and prop unscrewed and were flung off.
Having a improved prop locking system will help ensure those types of errors don't happen. Just my 2 cents on the topic.
-
1
-
2
-
3
of 3 Next