Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

A Curious Mystery

WTFDproject

Moderator
Staff member
Premium Pilot
Joined
Oct 27, 2018
Messages
3,929
Reaction score
3,177
Location
North Carolina Piedmont, USA
I doubt many will benefit from this, but it gave me such a struggle I thought I would post it.
The issue was an intermittent loss of power to the CGo3 camera mounted on Q500 4K

My typical route includes a launch point, about 60 Yards level travel to the center of a clearing, then vertical ascent to begin the actual flight.

The camera would start normally, stabilize, good video, etc. The issue began as a loss of WiFi when I reached the center of the clearing. Camera would still have power, gimbal still stabilized. Thought the common location was odd, but assumed the issue was a really bad WiFi board in the camera. Changed out the WiFi board, and tried it again.

Same distance, lost WiFi exactly the same place, except now, the entire camera was dead, no gimbal, no anything.

Turn the drone off, back on, camera comes back, try again. Repeated several times, became convinced it was related to either distance, location, or the time between motor start and camera failure. Wasted a good bit of time using different routes, locations, even sitting on the ground powered up, doing extended hover a few feet off the ground, etc . Nothing, except when I flew out to the middle of the same spot, the camera would die again. I looked around the problem area, tried to think of what might be an issue, but I quickly ran out of ideas. Nothing has changed out there forever.

So, then I put the camera on another drone. Everything worked fine. Put it on a couple more drones, everything fine.

Put a couple different cameras on the first drone, got the power failure again. Every time. Same Spot. Think I wasn’t getting REALLY convinced it had something to do with the location? Yeah, but I didn’t have any more ideas to pursue related to location. So I decided to look at the drone, instead.

I had already checked the 4K contact voltage, always good, tried the pigtail, no change. So I took the top off the drone to look at the wiring. At first, all readings were good. Good voltage, good contact, nothing wrong at the solder joints, nothing wrong at the connectors. Then I happened to move the wires a little bit, and all of a sudden, NO CONTACT on the hot wire. Checked solder joints again, contact pins, no issue, but when the wire was flexed a little, it lost continuity. Used a pin tip probe to inspect this obviously perfect wire, and found I had contact in one direction, but not the other. A little more investigation, and found the issue. The wire inside the insulation had a break in it. When the wire was pulled apart, there was a small section where there was almost no actual wire, and it was pretty much separated. I assume that a combination of battery voltage and wire temperature was causing the intermittent contact/failure. Replaced the wire, put it all back together, and all is great for several flights.

What did I learn from this? No matter how strong a coincidence may seem, it still may be just a coincidence. You need to look elsewhere. There was no actual issue related to the location.

I also suspect someone with @ChuckBridges background might have seen a wire or two with a manufacturing defect before, and maybe solved the mystery a lot faster. Am I right?
 
Last edited:
I doubt many will benefit from this, but it gave me such a struggle I thought I would post it.
The issue was an intermittent loss of power to the CGo3 camera mounted on Q500 4K

My typical route includes a launch point, about 60 Yards level travel to the center of a clearing, then vertical ascent to begin the actual flight.

The camera would start normally, stabilize, good video, etc. The issue began as a loss of WiFi when I reached the center of the clearing. Camera would still have power, gimbal still stabilized. Thought the common location was odd, but assumed the issue was a really bad WiFi board in the camera. Changed out the WiFi board, and tried it again.

Same distance, lost WiFi exactly the same place, except now, the entire camera was dead, no gimbal, no anything.

Turn the drone off, back on, camera comes back, try again. Repeated several times, became convinced it was related to either distance, location, or the time between motor start and camera failure. Wasted a good bit of time using different routes, locations, even sitting on the ground powered up, doing extended hover a few feet off the ground, etc . Nothing, except when I flew out to the middle of the same spot, the camera would die again. I looked around the problem area, tried to think of what might be an issue, but I quickly ran out of ideas. Nothing has changed out there forever.

So, then I put the camera on another drone. Everything worked fine. Put it on a couple more drones, everything fine.

Put a couple different cameras on the first drone, got the power failure again. Every time. Same Spot. Think I wasn’t getting REALLY convinced it had something to do with the location? Yeah, but I didn’t have any more ideas to pursue related to location. So I decided to look at the drone, instead.

I had already checked the 4K contact voltage, always good, tried the pigtail, no change. So I took the top off the drone to look at the wiring. At first, all readings were good. Good voltage, good contact, nothing wrong at the solder joints, nothing wrong at the connectors. Then I happened to move the wires a little bit, and all of a sudden, NO CONTACT on the hot wire. Checked solder joints again, contact pins, no issue, but when the wire was flexed a little, it lost continuity. Used a pin tip probe to inspect this obviously perfect wire, and found I had contact in one direction, but not the other. A little more investigation, and found the issue. The wire inside the insulation had a break in it. When the wire was pulled apart, there was a small section where there was almost no actual wire, and it was pretty much separated. I assume that a combination of battery voltage and wire temperature was causing the intermittent contact/failure. Replaced the wire, put it all back together, and all is great for several flights.

What did I learn from this? No matter how strong a coincidence may seem, it still may be just a coincidence. You need to look elsewhere. There was no actual issue related to the location.

I also suspect someone with @ChuckBridges background might have seen a wire or two with a manufacturing defect before, and maybe solved the mystery a lot faster. Am I right?
Pinched or just a bad wire?
 
Bad wire. There was a section that had a lot less actual wire than the rest, and it was deteriorating from use. I assume it would have failed completely very shortly, but it was still making enough contact when cool and when battery was fresh.
 
Was the wire at a flex point where the break occurred? We used to see a lot of that on our through hole insertion heads for IC’s.
 
No. It was ~ 1/2 inch from a soldered joint, but was laid straight, and does not normally experience movement. That doesn't mean it has never been flexed, but it's location does not imply a significant challenge during installation or maintenance. The Bird's been crashed a time or two (that's how I got it), and It has had repairs in the area that would involve wire movement. However, the apparent lack of wire mass makes me suspect more just an oddball defect than mechanical damage.
 
I have seen weird wire problems before, but they are not that common. it is possible, since it was near an end that water (even condensation can affect this) traveled down the wire to the weakest point in the copper. It then does it's voodoo there. In outside plant (Wires on poles and in pedestals), we block water movement with a gel that solidifies. Makes repairs in that area tough, but keeps water from traveling. In the old days of PCM (Pulse Code Modulation, when fiber was too expensive to put everywhere), we would actually pressurize the cable from the CO. We would monitor the pressure on the cable and anything with lower pressure had a hole. This was fixed right away.

Sorry, I get long winded sometimes. Yes, this does happen. There can be multiple causes but manufacturing is usually not the problem. The process involves wrapping multiple small threads to make up the wire. It is pretty failsafe. The wire might have been flexed multiple times before it was installed.

Have a great day!
 
ooOOFFF. You just reminded me of something. This was a water bird. I got it after the previous owner dunked it in the Gulf of Mexico. Not much original on it except the plastic and the wires, including the camera wires. I assumed the dark color of the remaining wire was heat, but maybe heat and salt? I got it about a year or so ago, and don't know how long it had been since the dunking.
I think you just solved the mystery. And I don't know what else might have failed next, if not for your post. This bird is grounded until it gets a complete rewiring.
Love this Forum.
 
  • Like
Reactions: Docdor and Grumpy
Glad to help. Since I retired, and until my Granddaughter can speak, who else am I going to share my knowledge with. It try with my wife, but I start speaking and all she hears is... wa wa wa wak. I feel like an adult depicted in Charlie Brown. You know what they say, use it or lose it. I use to say, knowledge in cold storage is lost knowledge. I love sharing my experiences and helping others.:cool:
 
Hi there,

The "coincidence" error is something diagnosticians run into all the time, whether medical or electronic. That's the way the brain works. If you put on a piece of music, and then run something with fixed periodicity, like a metronome or the windshield wipers, most of the time you'll get the impression that they are rhythmically synced up when obviously they are not!!! That's why coincidence may be useful to narrow down the problem, but is never conclusive. Great work on troubleshooting the problem: you did all the right things and finally figured it out by the process of elimination. Usually a bad solder joint is more common, but anything that has been in salt water is highly suspect. Do you still have the original wire? If so it would be interesting if you very gently cut off the insulation in the problem area and then inspect under magnification of about 5 - 15x. You should clearly see the evidence of erosion. We always use a low power dissection microscope when troubleshooting faulty circuitry or PCBs. You just cannot see in enough detail without something like that.

One possibility, rather than replacing all the wiring, would be to inspect each wire separately using an ohm meter. Alligator clip the probes right to the solder joints if possible (otherwise have someone else maintain probe contact manually), then pull, twist and mechanically stress the wire in all directions. If there is no break in the ohm reading (all modern DVMs that I am familiar with beep audibly when there is contact), then the wire is OK. Might save you a lot of re-wiring time. There is always the issue that you might make a mistake in the re-soldering (so called "cold" solder joints are a common problem in which there may be electrical contact, but not enough mechanical contact to keep the joint from failing.) Then you'll be back to troubleshooting all over again!! As the saying goes, "if it ain't broke, don't fix it".

You may also want to closely inspect (again under some magnification) the circuit boards themselves. If you see any bluish spots, or any kind of powerdy deposit that comes off easily when poked with a toothpick, then out of an abundance of caution, I would suggest you disqualify that whole circuit board. I have heard that if you bake the problem board in a 350 degrees F. oven for about 8-10 minutes, the solder will re-liquify, re-seal all the joints simultaneously, and then solidify again when you remove from the oven. I have never had to resort to this myself, but I have a friend who swears by this method and has used it quite successfully to fix supposedly "dead" mother boards.

If you do any soldering, here is a great tutorial on how to do it properly: {{{ How To Solder: A Complete Beginners Guide - Makerspaces.com }}} and {{{ Electronics Repair - Soldering Guide }}}.

The latter link was authored by a fellow who spent about 30 years instructing NASA technicians on proper soldering technique. Can't ask for a better tutor than that :) I guess if you are building rocket ships, the last thing you want is a bad solder joint failing on you when the bird is in orbit. Rocket ships are a bit more costly than drones :)

One more thing, the best way to tell if a solder joint is OK or not is through visual inspection. The second link above discusses what to look for.
 
  • Like
Reactions: WTFDproject

@homedoc,
WOW. Great post.

I had already discarded the old wire before the real problem was revealed, and could not find it again. That was one of the first things I did, after realizing what had happened. I would love to have taken a better look at it.

I've used the 350 deg oven for larger projects, such as laptop motherboards, and it does work pretty well. For small stuff, like our mainboards, I use a hotplate to pre-warm the boards, and a hot air rework station mounted on a frame over the hotplate to do the actual re-melt. That works really well, and it allows for actually working with / replacing individual SMDs when required.

Reviving salt water damaged boards (ESCs, mainboards, receivers, GPS) has proven difficult for me. I have been successful in replacing board mounted components, repairing damaged tracing, and getting them back in the air, but they always seem to develop new corrosion over time. Even if I replace solder, I can't seem to fully eliminate the salt. I've tried several commercial cleaning fluids in an ultrasonic cleaner, and some homemade concoctions as well, but have not found anything that will really get down into salt saturated solder. Now days, I just replace the boards if they have had salt water contact. This particular bird already has a new mainboard and ESCs. I'm not sure about the receiver and the GPS, but I don't see any corrosion, even under a microscope, so I probably changed them at some point also. (I hope)
 
Last edited:

Members online

Forum statistics

Threads
20,954
Messages
241,584
Members
27,284
Latest member
csandoval