Location Specific Issue Needs Our Collective minds for Solution

[maxwelltub]

Hello,
I posted a couple weeks ago about a issue I had with a NAZA v2 where I almost lost my Y6, twice, due to what appeared to be a compass calibration issue. To briefly summarize the issue,

I got a compass calibration light before take off, tried reaccelerating, rebooting and eventually decided to test fly because I don't use the GPS/ Compass for flight anyway. After a quick hover test decided it was flying well enough and went for it. About 3 minutes in I had two close calls where the aircraft lost control for a short time and nearly tumbled out of the sky. It was a quick failure causing the aircraft to literally tumble and fall, but last less the about 4 seconds at a time. Despite this happening twice was able to land without further incident.

I couldn't find anything wrong in the assistant software. I was flying very close to a large bridge, and large metal container ships. So I chalked it up to microwave interference or magnetic interference from the ships.

I returned to the location today after a month. My location was much farther away from the bridge and ships and had a quick flight blip again but again was able to land safely. I checked 2.4 frequency on the ground and saw no red flags. I would say I was about 1/2 to 3/4 miles from where the original incident occurred. On the day of the original incident I flew for about an hour as I moved down the pier before I ended up at the take off spot where the trouble first presented itself.

Now I am scratching head trying to figure out the problem. I have flown this setup extensively in other areas with zero problems. Both before, and after the original incident.

Could it be 2.4 interference? If so it would not only have to block out my radio but also effectively prevented failsafe mode from working and or giving some other confusing signal to the flight controller. I would think the 2.4 interference would show up on my preflight, but a focused microwave dish/ satellite dish might not display on the ground. I would also think that if the signal was blocking the 2.4 tx then the craft would go into failsafe, which has been tested. Also should note that i was flying with 10 satellites, and again in Atti mode, not very far away.

Could it be GPS? Doubtful because I only fly in Atti and manual on occasion.

Could it be ESCs/ vibration or other hardware? Not likely because the aircraft has worked well in other locations for the last month.

So my best guess is some sort of 2.4 issue. This location is a heavy industrial area, a shipyard. Any thoughts on how to trouble shoot this problem?

 

[SchrodingersCar]

Doesn't ATTI mode still use the compass for orientation? I could be wrong, but that's what I figured was the cause of my ATTI mode disaster...

 

[maxwelltub]

No I believe the compass is located inside the gps puck.

 

[Hexacrafter]

Just a thought.....
Maybe it is not the FC at all....especially with as extreme & apparently violent aircraft actions.
Perhaps a dry solder joint on an ESC/ Motor/ PDB???
The only way to know for sure would be to test with a different FC, but my gut says either bad connection somewhere or just a bad FC..

Just a thought...

 

[maxwelltub]

That was definitely I thought I had as well. I've seen mulitrotors go down from bad solder joints. I've checked all the connections the first time I had the incident and nothing came up. I've probably flown this aircraft 2-4 times a week for the last month and change and literally never had any type of remote issue before or after the incident, until I returned to the location today. Could be strange luck here but the fact the first time I had a error message on the LED make me think its something with this location. I will be flying later this week so I'll report back if there is any issues at other flying locations.

 

[violetwolf]

Not addressing your specific issue. But I live in fear of this as well. Interference at different locations etc.

What I've done is built a beater hex using identical FC (apm) and radio (frsky) so that I can do a test flight or two at new locations before I fly the camera bird.

I've also recently switched to Taranis radios and x8r receivers for the RSSI telemetry so I can monitor the signal strength that the rx sees while airborne.

I've also added a pre-flight range test to my pre-flight checks.

I don't know if this will save my *ss but it gives me peace of mind.

 

[maxwelltub]

Update:
Original hypothesis that this was a location specific issue has been disproven. It must have been just an incredible coincidence that the problem displayed the first two times at the same general location with over a month in between.
The problem displayed again this weekend at a completely different location. My best guess now is flight controller issue. If it was a motor or esc I don't think it would dip and dive. It is a y6 with ample power and I have tested prop/ motor failure and it can fly with 5 motors. I will have to retire this FC, but I am going to put it on a smaller quad test rig and see if I can really narrow down the issue. I'll report any findings if any, my guess is I will never really know what the culprit is.
The Take away Lesson is
These things work great... until they don't. Problems can show up with little to no warning. That's why I always tell people to not fly over crowds. Losing a multirotor is one thing, I've accepted that risk, hurting someone is not. Set a good example by practicing safe flying.

 

[Hexacrafter]

Max...
Thanks for the update...I always wonder how some of these things turn out...
Also...
I agree entirely with the lesson....
Even the BEST I know in the business have the rare equipment failure...
This is why I always stress the need to have an emergency plan in place....and to practice a emergency divert... so it becomes something you are ready for....
On the shoots I have been on, a plan for a "ditching" / "safe" location is always planned...
If any aircraft control remains, the plan is in place to get as far away from people as possible....
The aircraft and camera are insignificant when compared to injuring people.....
Thanks again for the follow up....

 

[maxwelltub]

Update,
I believe the issue can be narrowed down to a motor or the FC. I am replacing the Naza with an old WKM. I also replaced a motor that was giving me a strange spool up. It looked as if one of the 3 phase wires had come loose. It was secure in the esc, but perhaps loose in the motor. I replaced the motor, (in the field) and took the bell off the motor in question when I got home. Its a sunnny sky motor and too small to rally see the connections to the magnets. But when I jostles the leads I can get motor to sputter on spool up. My guess is that this was the issue, although as I said before I've tested the platform with a simulated motor failure and didn't have this response. Its the not knowing for sure thats driving me nuts.

 

[fltundra]

The Take away Lesson is
These things work great... until they don't. Problems can show up with little to no warning. That's why I always tell people to not fly over crowds. Losing a multirotor is one thing, I've accepted that risk, hurting someone is not. Set a good example by practicing safe flying.

The lesson should be to avoid all DJI flight controllers.

 

[maxwelltub]

that certainly is the easy response, however I am not a hardware or electrical engineer. So its hard for me to know whether its actually dji has a higher fail rate, or just a higher market share. If someone could take a part two similar price point FC and tell me why one is better that would be awesome. Although I agree and I've shed the dji FC's on my larger platforms, but this one is built around the gopro zenmuse which is hands down the best gopro gimbal.

 

[fltundra]

So its hard for me to know whether its actually dji has a higher fail rate, or just a higher market share. If someone could take a part two similar price point FC and tell me why one is better that would be awesome.

It's real easy, just do i google search for each flight controller. It's not hard to see a pattern with a few, as in the "A2".
Two years ago I spent three months researching FC's and the final Decision maker for me at that time was Kloner aka "Steve" ditching the A2. I spoke with him 2 months ago and he was still flying the SuperX as I am without issue. He did say they had one SuperX rig that failed, but it wasn't one he put together.

 

[maxwelltub]

Yes thats what I mean when I say higher market share. If super x was used as much as a Naza or A2, would we see similar fail rates? I don't know, and when i researched this failure of FC, (if it was indeed the FC) it was very hard to find objective information, Only rants about DJI. I've used a WKM for many years with Zero issue. So the fact that Kloner uses a Super X with Zero issue is only anecdotal evidence.

 

[fltundra]

Yes thats what I mean when I say higher market share. If super x was used as much as a Naza or A2, would we see similar fail rates? I don't know, and when i researched this failure of FC, (if it was indeed the FC) it was very hard to find objective information, Only rants about DJI. I've used a WKM for many years with Zero issue. So the fact that Kloner uses a Super X with Zero issue is only anecdotal evidence.

There is many many users flying the X without issue. They just don't visit the forums. There flying, not repairing or rebuilding.

 

[maxwelltub]

I guess what I am really trying to get across, and we are going down a rabbit hole here, which is totally cool, but what I am getting at is objective FC comparison. What you and I and many people do is fall victim to confirmation biases and anecdotal evidence. So when you say there are plenty of people flying without issue and we don't hear about it, you could make the same statement about people who fly DJI who don't report when everything is working. I don't have the numbers but I would imagine that Super X sells a very small fraction compared to the amount of DJI FC in the air.
There is a word for another type of research error/ bias which I can't recall but it works like this. Say I have a hypothesis, for example "a location if messing up my FC". If I go out to prove or support this hypothesis and it turns out to be unsupported what can I do? I can file that report away in my own desk and no one will ever see it, which is probably very very common or I can continue the thread to say my hypothesis is not supported lets look at XYZ next, which is probably less common.
Now lets say I went out and my hypothesis was found to have more support to prove it. What would I do, I would run back here to report it. This can cause people to very unknowingly sque the information available. There are many hundreds of threads that just end with no solution to be found. Or no confirmation that the hypothesis was supported or unsupported.
So just for numbers here let's say for every 1 Super X controller there are 20 Nazas. (in real life it is probably more than 1/100) If they have the same fail rate naza complaints will be reported 20 times more often.
You see where I am going here? I'm not saying one is better than the other, I am saying I would like to see an electrical engineer (preferably a 3rd party one who doesn't work for the company) take apart both and say why one is better than the other. I think flight control companies could benefit tremendously from this sort of transparency.

Maaaaaaaybe we could do a kickstarter campaign to fund something like this. The great FC shoot out, similar to what Zacuto did a few years ago with cameras. Anyone interested in taking this on with me? @Bartman?